Women and Seasonality: Coping with Crisis and Calamity
Janice Jiggins
I Introduction
Over the last few years, a great deal of evidence has
been amassed on the impact of seasonal adversities on
women, children and their families. Attempts have
been made to differentiate the varying impacts on
households and, within households, on women and
children in different income classes and to build
dynamic models of the 'screws and ratchets' which
push manageable seasonal stress toward the break-
down limits of livelihood systems.
What is attempted here is an exploration of the
contribution of female production, labour and
domestic domain services to the management of inter-
annual and intra-annual uncertainty, the steps in the
sequence of deterioration under accumulating stress,
and of the options open to women and their children
through and beyond the point of family disintegration,
when managing seasonalities becomes a matter of
individual physical survival.
The evidence of female mortality and morbidity rates,
from some areas at particular times, suggests that the
wastage of females may be countenanced in times of
acute stress as necessary to the survival of social
systems as a whole, however distressing at the level of
family survival. It establishes the extreme end of a
range of situations in which poor rural men and
women act and react to expected inter-annual and
intra-annual fluctuations, interspersed with shocks
whose advent is always latent but whose timing and
severity is unguessable.
The management of uncertainty is inherent in small
producers' and labourers' livelihood systems; not
surprisingly, these are characterised by flexibility, the
maintenance of a range of options to meet expected
fluctuations in resource endowments, entitlements to
food, work and income, climatic variation and the
unreliability of government services. If it is true that
the less flexible the livelihood system, the harder it is to
manage seasonal stress and sudden shock, then it is
IDS Bulletin. 1986, vol 17 no 3. Institute of Development Studmx, Sussex
important to understand how and what different
members of a household contribute to that flexibility.
Such an exploration leads to consideration of how
members of households assess probabilities and how
they express risk preferences. It has been fashionable,
for example, to assert that small producers prefer to
minimise risk by aiming for inter-annual yield stability
around the minimum necessary to meet subsistence
needs. The concentration on yield stability per se may
be diverting attention from a more dynamic calculus
in which household members complement each
others' contributions to livelihood stability across
seasons by maintaining the capacity to transfer
resources in and out of the sub-systems which together
constitute their livelihood.
In some enterprises, one family member might be
happy to make a high risk-high pay off investment if
assured that failure could be covered, or another to
make a high input-low return investment if that return
were deemed essential but could be gained in no other
way. This calculus is likely to change overtime. As in a
commercial
business, both risk preferences and
probability assessments are likely to become more
conservative after a run of bad years, as assets and
room for manoeuvre dwindle and as investments
made in the course of a run of good years have to be
paid for out of shrinking revenues.
As households head into the bottom of the cycle, it
becomes a fine run thing for many of them to maintain
the flexibility to ride out the bad years. The need to
concentrate time and effort on essential high input-
low return activities (such as fetching water from
distant river beds in the dry season), may absorb
household resources to the point of no return;
households here must enter into new livelihood
systems closer
to the point of destitution, or
disintegrate.
It is because men and women make separate if
complementary contributions to the maintenance of
9
livelihood flexibility within the framework of expected
inter-annual and intra-annual uncertainty, that not
only the timing of a sudden shock but the gender of its
victim(s) is important. The death of children from
measles at the beginning of the agricultural season
might provide greater room to manoeuvre to a couple
seeking their daily living from an uncertain and
gender-ascribed wage labour market or, on the other
hand, remove essential labour at a critical moment
from a female household head farming on her own
account. The death of a husband for a relatively well-
off woman in a tenant household might lead to her
forced acceptance
of the
position of unpaid
agricultural worker for her brother-in-law; the death
of the wife, on the other hand, might offer her husband
the opportunity for re-capitalisation of his farming
activities through remarriage and the acquisition of a
second dowry.
The options are various, the strategies complex and, it
seems, as yet we understand very little about how these
operate over time for households in different income
groups or for individual household members. The
following section explores briefly some of the ways in
which women are contributing to the management of
expected seasonal uncertainty and the maintenance of
livelihood flexibility. Sections III and IV attempt a
progression through time in the face of relentless
seasonal adversity.
II Uncertainty and Flexibility
Although neither the timing, distribution nor intensity
of seasonal stresses may be known in advance, their
advent and the range of probable fluctuation are
accepted as normal occurrences by the rural poor.
Among the range of possible responses, seven which
tend to be particular to women are outlined here in
brief.
(i) Switching Tasks and Responsibilities ascribed by
Gender
In many rural societies, specific tasks and areas of
responsibility are ascribed by gender. Where these are
rigid, it might be that households - particularly low
income households - find the management of
seasonalities harder than in societies in which there is
some scope for men and women to take over each
other's tasks and responsibilities
as need and
opportunity arise.
Contrast the
following two
examples. In an area of Tanzania in which only
women cook and carry water, dry season water-
carrying absorbs a great deal of female labour time.
Men welcomed a proposed village water facility
because, they said, 'Water is a big problem for women.
We can sit here all day waiting for food because there
is no women at home' [Wiley 198 1:58].
10
In contrast, a Javanese case study reports greater
scope for a more flexible response to gender-ascribed
tasks and seasonal opportunities: 'Men, for example,
sometimes stay at home to babysit and cook a meal
while adult women and girls are off harvesting, or
trading at the market' [White 1985:132).
In a study of the pastoral Orma along the Tana River
in Kenya, Ensminger (1985) presents data which show
only slight variation in the amount of time spent on or
in the pattern of male and female activities between the
seasons, except that, in the dry season, women do
slightly less work such as cooking and milking and
men spend more time in stock-watering and well-
digging. Although young girls may take on some of the
tasks associated with (male) herding, in general - at a
time of maximum nutritional stress - men's dry
season work increases somewhat whilst women's
leisure time increases. Asking why there are 'relatively
few age/sex cross-overs of labour allocation between
seasons' (page 14), Ensminger finds that her data do
not satisfactorily support explanations based on
reproductive rationality, differential physiological
efficiencies, social reproduction needs nor redistri-
bution.
Indeed, it would seem that it is partly the social
perception of the scope for switching rather than
'objective' assessments of capacity or returns which
determines how flexible households can be
in
assigning seasonal labour tasks. In a study of labour
market behaviour in South India, Ryan and Ghodake
(1980) attempt to relate the effects of season, sex and
socioeconomic status and speculate that differential
labour market opportunities would support the
economic rationality of skewed intra-household food
distribution toward adult males but, as Schofield
(1974) points out, we simply do not know if this
presumed rationality leads to food being seasonally
distributed independent of the task and sex of the
operator:'. . . are women fed more when weeding and
men when ploughing? In this case, commonsense
would suggest that available food is so distributed to
the workers that the non-work force section bears the
brunt of seasonal variation in food supplies'
[Schofield 1974:26).
Where male and female farming are
partially
separated within the household livelihood system, the
answer to the question of the intrahousehold pattern
of income and food distribution
in
relation to
women's labour productivity, as Jones (1982) has
demonstrated for a Cameroonian case, may lie in
calculations of the intrahousehold rate of compen-
sation rather than market opportunity costs.
Another factor may be the degree to which own-
account production is the main livelihood source. One
study in Cajamarcan in the Andes found that in
landless
households depending on non-farming
income-generating activities for the major part of their
livelihood, a 'flexible sexual division of labour [in
agriculture] appears to be required by economic
necessity', whereas in landed households, agriculture
is predominantly a male activity [Deere and de Leal
1982:88].
(ii) Diversifying Household Income Sources
It is common in development studies to see female
income referred to as supplemental and for it to be
subsumed within estimates of household income.
Neither practice seems particularly helpful. For
growing numbers of households headed by women,
women's earnings form the main cash source; in
households where male and female responsibilities are
separated, women are obliged by the terms of their
marriage contract to find the cash needed to fulfill
their assigned responsibilities; amongst the poorest
households, women's earnings may form an equal or
larger share of household income; a greater portion of
the income accruing to women than to men tends to be
spent on household welfare and consumption needs.
For all these reasons, in terms of seasonal analysis, the
sex of the income-earner and the intra-household
distribution of income and responsibility is thus likely
to be more important than total household income as
an indicator of the household's capacity to maintain
itself in the face of seasonal adversity.
A number of studies do, in fact, show that women
make careful judgements of the balance of advantage
between, for example, maintaining food stocks and
converting a portion to beer-brewing and selling as the
agricultural season progresses [see Saul 1981 on
sorghum beer-brewing in Upper Volta] or between
allocating their labour to food production and
processing for domestic consumption or to marketing
[see Kebede 1978 for the balance between enset (the
'false banana') production and the chircharo system of
trading among the Gurage in Ethiopia].
There is,
further, growing evidence of the close
correlation between female income-earning and child-
bearing: the higher women's income, the lower the
number of pregnancies [Evenson 1985:27]. The causal
relationship appears to be mediated through the
monetisation of women's time. If we have evidence
that changes in agriculture lead to an increase in
women's time input with no increase in - or even loss
of control over - their income, then we can expect
that the adverse seasonalities associated with maternal
and child health will, in fact, be exacerbated and may
be contributing to the kinds of family breakdown
outlined in Section IV.
Women in Rwanda combining farming with child care.
(iii) Changing the Intensity and Mix of Multiple
Occupations
There are good records of women manipulating the
intensity of performance and the mix of occupations
associated with their multiple roles in order to cope
with seasonally urgent tasks. In general, it would
appear to be their domestic domain roles which are
squeezed rather than production or income-earning,
though, as one would expect, the balance of net
advantage may be different for women in households
in different income classes [for a Philippines example
see Lilo
1985:85-7]. For example, surveys among
primary school children in the Mochudi District of
Botswana during the ploughing season showed that
nearly one third of primary school children were
caring for themselves without adult help in the month
of February whilst parents were absent at the lands
[Otaala 1980]. Cooking may be reduced to once a day
or every two days during peak farm labour periods or
staples substituted by snack foods which can be eaten
raw [Bantje 1982a, Table 2; Jiggins 1986]. Ryan and
Ghodake [1980] note for four South Indian villages
that it is the hours women work in the domestic
domain or as unpaid farm family labour which tend to
fluctuate seasonally rather than the hours of waged
work.
11
(iv) Household Gardening
The domain of the household garden provides a
further element of flexibility in the livelihoods of those
with
access
to land. Studies from Grenada,
Zimbabwe, West Africa, Jakarta, South East Asia and
Peru emphasise the importance of household gardens
under women's care as a source of early-maturing
varieties of staples to carry families over the hungry
season till main crops mature, as reserve sources of
plant materials should main crops fail, as conservation
sites for special or preferred varieties and as testing
grounds for new varieties or practices [Brierley 1976;
Callear 1982; Eijnatten 1971, Evers 1981, Ninez 1984,
Stoler 1978]. A study in Kalimantan in Indonesia
recorded an average of over 40 different species of
vegetable, spice and fruit crops in household gardens
[Watson 1985:198]. Local cultigens, semi-wild and
protected wild species, together with small stock and
poultry, may add to the diversity and richness,
constituting a complex biological coping mechanism
responsive to intra-annual and inter-annual climatic
and labour time variations, meeting specific seasonal
end-uses which cannot be provided by field crops,
however abundant [Jiggins 1986].
(y) Food Processing, Preservation and Preparation
The choice of crop mix, plant characteristics and
amount of time devoted to
cultivation is not
determined solely by consumption preferences nor are
food purchases determined only by income; they are
intimately associated with the technology available to
women for domestic processing, preservation and
preparation. These technologies in turn may be linked
to the seasonal availability of different types of fuel for
cooking and space heating [Foley et all 9 84:34] and the
differential fuels available to women in different
income strata through the seasons [ibid 36]. Vidyarthi
[1984] shows from data for one Indian village, the use
of dung and firewood by women in bullock-owning
households and an increasing reliance on crop
residues by poorer women, who use spiky millet stems
through the end of the Kharif season in November,
then pigeon pea stems through the end of rabiin April
(which give the best sustained heat of all residues), and
then a weed, Ipomeafitulosa, which gives a smoky heat
and must be gathered, cut and dried for a month
before use, and gathered wood. He estimates that
agricultural residues may form around 40 per cent of
all fuels used by poorer women.
Huss-Ashmore details these links carefully for female-
headed households
in
highland Lesotho [Huss-
Ashmore 1982:156]. In Mokhotlong the type of fuel
used and the time spent getting it vary according to the
seasonal availability of dung. Slow-cooking protein
sources are not used equally through the year but are
depleted during the cold season when the slow burning
12
compacted dung is available. 'During the summer the
population relies heavily on wild vegetable protein
sources, which require more time to locate and gather
but which can be rapidly cooked', using the horse and
cattle dung picked up on the high pastures and kindled
with quick-burning resinous and woody shrubs [ibid
157]. It is fuel seasonalities and not crop availability
which determine which foods are eaten and the food
preparation equipment used at different seasons.
Women also attempt to cope with crop seasonalities
through food processing, to extend the storability and
shelf life of perishables, from simple sun-drying of
leaves and vegetables treated with soda ash, to more
elaborate transformations such as those involved in
the making of Kenke and gari (cassava products) in
West Africa or chuno (potato products) in the High
Andes.
(vi) Social Organisation
An apparently growing phenomenon is the formation
of multi-generational, multi-locational networks of
households headed by women. Only some of these are
the result of family breakdown - women may be
choosing to have children without what they perceive
to be the burdens of marriage [Kerven 1979]. They
appear to be an emergent form of social organisation
designed to spread risk and optimise seasonal
management strategies in areas of high gender-specific
migration, marked seasonality, and marked gender-
specific livelihood opportunities [see Kerven 1979 for
Botswana examples; Phongpaichit 1980 for Thai
examples].
Another strategy in areas where there is a developed
labour market is for women from poor households to
associate in specialist labour gangs to take advantage
of seasonal cropping patterns. They may travel over a
wide area, moving with the season from contract to
contract, with gangs known for their speed and skill
gaining premium rates. In a ten-member Sri Lankan
gang documented in 1979, which moved from the wet
zone to the irrigated dry zone twice a year to carry out
paddy transplanting, six were married women, of
whom two were separated from their husbands
[ESCAP/FAO 1979:28-40]. The four resident hus-
bands worked as casual labourers. The other women
lived with their families, of whom only three had even
a tiny plot of high land for cultivation. Their ages
ranged from 26 to 55 years and they worked as casual
estate labourers
the rest
of
the
year.
Their
transplanting earnings were spent on daily living and
family needs;
their own clothes and jewellery;
furniture and pilgrimages. The high preference for
turning their earnings into an easily convertible store
of value under their own control, as a hedge against a
crisis and calamity, has been noted in many studies
[Jiggins 1983].
Yet another mechanism is to develop semi-formalised
women's groups based on
existing forms and
principles of female association. Yet these might not
be as useful as might at first be supposed in the
maintenance of the poorer members' livelihoods
through seasonal stress. In a study of women's groups
in Kenya, members were asked to identify those who
were 'famine resistant' or 'famine-prone' i.e. who
would or would not be able to stand even a mild
harvest failure or livestock disease. Famine-proneness
turned out to be associated with illiteracy and
household headship [Muzale and Leonard 1985:19].
Resurveyed after a year of drought, the membership
was found to have dropped to those previously
identified as 'famine-resistant'. The famine-prone had
left the groups long before the groups suspended
activities due to the drought and were not expected by
those who were left to return.
'Participation in women's groups at the initial
stages represents a form of long-term investment.
At that stage, the groups do not yield material
benefits for individual use in the family. Joint
welfare funds, friendship, production information
and skills are all the benefits
that group
participation is able to produce at the individual
level at the initial stages. Women operating within
small resource margins in a harsh environment are
not likely to be able to undertake this form of
investment on a continuing basis. If the groups'
policies continue to demand contributions well
into periods of environmental stress, poor women
will be excluded' [Ibid:20].
Yet another mechanism - though possibly the
reference is eccentric - is the practice of what might
be termed 'seasonal polygamy': men contracting
marriage with additional wives at the beginning of the
crop season and divorcing them again afterwards, in
order to optimise household labour resources when
they are needed and to minimise the post-harvest draw
down of household food stocks [Bantje 1982a:l6].
(vii) Gift-Giving
Hidden within rural life is the special advantage that
single, widowed and divorced and separated women
may have to solicit and accept gifts from men in a
relationship which falls short of prostitution in many
respects but which women may skillfully exploit as a
gender-specific coping strategy, even in societies in
which propriety deems it a protective rather than
exploitative relationship. Documentary evidence,
unsurprisingly, is
meagre. One example from
Tanzania records a women's comment: 'We just look
this way and that way for help. You see, lama woman'
[Bantje 1982b:7].
III
Dealing with Relentless Adversity
Given the kind of flexibilities described above, what
gives way as families move into deepening poverty in
the face of relentless adversity, such as several years of
bad harvests? It seems we do not have sufficient
information as yet to write about generalised patterns
of how women adjust (nor of the effects of family
adjustments on women) or to define precisely the
parameters within which they occur. The following
cases from the South Asian region, then, are only
illustrative of the kinds of things which seem likely to
happen.
A study of the sequential responses of deepening
poverty in villages in two areas of Bangladesh, viz.
Comilla and Modhupur, distinguished between
female wage-earning households and those without
female wage-earners and, within the former category,
the position of widowed/divorced/separated women
and married women [Begum 1985:221-41]. At some
point in a run of bad harvest years, in smaliholder
households in which women did not work for wages,
males sought or held non-agricultural wage jobs which
at a pinch could compensate for loss of farm earnings.
In smallholder households where women took wage
work, men had no such alternative job and began to
'... lease out and perhaps sell land. They may also
sell productive assets (e.g. bullocks) or consume
productive inputs (e.g. seed). They may place male
children in permanent jobs where they receive food
and shelter. Finally, women may perform wage
labour. The involvement of rural women in wage
labour seems to be the last step in a series of family
adjustments to economic crises that is taken only
when the alternative is the effective breakdown of
the family unit' [Ibid:232].
Households that had female wage earners also were
more dependent on children's earnings. In particular,
the higher percentage of labour participation of
female children from female wage-earning households
was found to reflect their acute poverty [Ibid:233-4].
There is an indication that the ability to support
livelihoods through gleaning is dependent on the
characteristics of the rice varieties grown. Among a
number of differences between survey sites, the study
pointed to the importance of the rice varieties grown
as an index of the availability to women of harvesting
wages, gleaning and post-harvest
threshing
employment.
'The long strawed broadcast aman rice grown in
Comilla was less uniform in length. Consequently,
some crops remained unharvested in the fields. On
all land but that belonging to the poorest
households, it was a prerogative of the women and
children from poor landless households to glean
13
