Chapter 4. DETERMINANTS OF WOMEN'S WORK AND PARTICIPATION IN THE DIFFERENT PARTS OF THE FOOD CHAIN

4.1 Persistence and change in cultural practices
4.2 Breakdown of cultural practices
4.3 Seasonality
4.4 Socio-economic differentiation
4.5 Macro-level factors of change
4.6 Women and technological change

In the past national population censuses or labour force surveys tended to under-rate women's work in the productive sector and its role in meeting the nutritional needs of the household. Such statistics generally considered women's work as "unpaid family labour" (Palmer 1977, FAO, 1984d, U.N. 1984).

Attempts are being made to better estimate women's role in production, particularly as a result of micro-level studies which have used detailed time/labour analysis to illustrate the extent and importance of women's work in the productive and reproductive sectors. Revised estimates of the contribution of women and men in agriculture in 82 countries found that women constituted over 40 per cent of the labour force in 52 countries. In 24 countries, or about one-third, women make up over 50 per cent of such labour (FAO 1984d).

Conceptual and methodological problems arising out of conventional definitions of "economic activities" still leave a great deal of women's work unaccounted for. Women's labour inputs into such activities as food processing (dehusking, shelling, decobbing, parboiling), storing, preparation for consumption and marketing, vegetable farming, care of livestock, and the cleaning, drying and selling of fish seldom enter national statistics. Neither do their contributions to the cash economy through sale of subsidiary food crops, processed products or crafts.

Another sphere of activities is that which is generally categorized as "domestic services". These activities include the procuring of water and fuel wood, cooking of meals and feeding of the young - all essential parts of the food preparation, distribution and consumption. Women generally provide 90 per cent of the labour inputs in this sphere.

This chapter will deal with some of the key factors that determine the specific forms of women's participation in food chain work and associated activities. The effects on women's food-related work of macro-level changes, commonly occurring in the Third World today, will be described on the basis of current research.

4.1 Persistence and change in cultural practices

One of the most significant determinants of women's participation in food-related work is existing cultural practices in a given society. Traditionally women's work has always been considered as significant for the household's food and nutrition requirements, but culturally women have always been conceived of as dependents and legal minors. This contradiction has often curtailed the full potential of women's participation in the food chain with serious repercussions on the nutritional status of their families and on their own well being. As one United Nations study put it,

While women represent half the global population and one-third of the labour force, they receive only one tenth of the world income and own less than one percent of world property. They are also responsible for two thirds of all working hours (quoted in Folbre, 1985).

a) Land

One of the prerequisites for women's participation in field production is access to land. The research on current norms and practices identifies the following types of access to land:

i) separate women's plot(s) together with unpaid work on family or husband's plot(s)
ii) joint cultivation with husband of the household plot(s)
iii) sole responsibility for cultivation by women heads of households
iv) landlessness.

The practice of women having separate plots is widespread in Sub-Saharan Africa. As a rule married women are allocated a plot or plots by the male head of household on which to cultivate food for the household and surplus for exchange or sale to meet their own needs (Okeyo, 1980; Guyer, 1980; Muchema, 1977).

In many instances women have specific crop responsibilities. Among the Tiv in Nigeria, women perform over 85 per cent of the labour inputs for yams, maize and cowpeas, while men have a greater responsibility for cassava and beniseed. In Zimbabwe, women are solely responsible for sweet potatoes and groundnuts (Muchema, 1977). In Tanzania women have the predominant responsibility for millet, groundnuts and intercropped vegetables (Mascarenhas, 1983).

The practice of separate plots for women has several implications. In the first place, it means that women in Sub-Saharan Africa, are directly involved in field production, and, as will be shown later, to a much greater extent than in the other regions of the developing world. The greater involvement can give women greater ability to provide for the households, but this ability is contingent on several aspects.

In the first place, traditionally, women with own plots could expect some obligatory labour inputs from the male head of household generally in the form of assistance in land preparation. Such inputs are essential especially if they have no means to hire labourers or equipment such as the plough.

Secondly, cultural practices also dictate that women's rights to separate plots be associated to the land rights of the males - fathers, husbands or sons. The size of the separate plot(s), the quality of land, the kind of technology allocated to it can all depend on the decisions of the male heads of households. In many instances, the introduction of cash crops can result in women's food crops being allocated to the less fertile land (Bukh, 1979; McLoughlin, 1970). This, in turn has the effect of lowering yields which have to be compensated with other productive activities. Such as beer brewing or wage labour.

Thirdly, even when women have separate plots, they still have to work on the family plot(s). The amount of labour expected from women in this respect depends on the type of crops grown, the ability of the household to hire casual or full-time labour, the size of the plots, etc. Recent studies seem to imply that women can negotiate over such labour inputs (Guyer, 1980; Folbre, 1985; Jones, 1983). Other studies, however, show that if there is need for labour on the male plots, such labour takes priority (Muntemba, 1982; Okeyo, 1980). The dual role in production - as own farmers and unpaid family labour can thus have both potentialities and constraints for women's ability to provide food for their families.

ii) Joint cultivation

In areas where land is scarce or where commercialization has induced male heads of households to keep a tighter control over land and labour, joint cultivation is the more common practice. In Asia, both scarcity of land and cultural practices result in women being predominantly joint cultivators rather than producers on their own fields. The term joint cultivators, is perhaps a misnomer because it implies parity of rights. In fact, women on jointly cultivated fields are usually "unpaid labour" with few rights to the decision about the disposal of the produce from these fields. In this sense, they are at a greater disadvantage vis-à-vis women with separate plots.

Nevertheless, joint cultivation does ensure women and men's participation in at least some of the activities. Furthermore, a jointly cultivated field is more likely to have the benefit of better technology and is often larger than women's separate plots. Therefore the total product can be larger and thus more beneficial for the household's well being, depending, of course, on how this labour product is utilized.

iii) Women-headed households

The phenomenon of women as de facto and de jure heads of households is increasing as male migration to urban centres increases and widowed and divorced women lose former family support systems.

The incidence of households that are de jure headed by women is still low overall, with the lowest (22%) being in Sub-Saharan Africa. However, in all three regions, Asia, Latin America and Africa, de facto women-headed households is quite high - with as much as 63% in Botswana (Buvinic and Youssef, 1978).

Recent literature on women in production has highlighted the special problems of female-headed households. Women in such households experience great difficulties in maintaining sufficiently large plots or obtaining adequate food from the fields (Bukh, 1979). Remittances from absent males or other members of the family are often inadequate to enable the women to hire labour to make up for the loss of adult male labour. A study from Zambia shows that women from such households work harder than those households that have participating male labour. Nevertheless, the output was low because of lower productivity (Kumar, 1985). This was probably due to the fact that problems with regard to access to technology, credit and extension service are intensified when households lack male adults. Many female heads of household are thus compelled to supplement field production with strenuous activities such as casual labour or beer-brewing to provide enough food for their households.

iv) Landlessness

The causes of landlessness are varied and extensive. In Asia the main causes are land scarcity and the very unequal distribution of land. In India, the top 8 per cent of the rural population owns more than half of the arable land. In the Pakistan Punjab, almost one-quarter of the land is owned by less than 1 per cent (Murdock, 1980). The introduction of the High Yielding Varieties (HYV's) has often aggravated this situation (Griffin, 1974).

In Africa the causes are much more varied. Colonial policies of reserving large tracts of land for white farmers resulted in intolerable population concentrations and landlessness in countries like Kenya and Zimbabwe (Leys, 1980). Some aspects of such policies still prevail.

In Sub-Saharan Africa, other more culturally based issues can lead to landlessness. According to tradition, divorced and separated women cannot have access to land in their own rights (Mascarenhas and Mbilinyi, 1983; U.N., 1984). Furthermore, current trends leading to communal use rights being turned into individual male ownership rights can result in women's landlessness when male heads of households sell rather than cultivate their land (Okeyo, 1980). In other cases the loss of male labour can lead to women opting out of agriculture because of its low productivity.

Landlessness has the result of changing women's participation in direct field production from cultivating for their own households to working for others, with all the consequent vulnerabilities. Nevertheless, such work is very crucial. In Thailand, for instance, even basic food needs such as the staple, rice, have to be purchased and earnings from women's work are crucial for the survival of landless households (Tinker, 1979). In most developing countries, however, wages tend to be low so that very often such labour has to be combined with other income generating activities for mere survival, causing tremendous psychological and physiological stress.

To sum up: four different types of land use have been identified, each with its own implications concerning adequacy, security in procurement, and ability to maintain adequacy and security.

b) Control of the labour product and decision-making

One of the crucial aspects about women's desire to participate in the production process centres around their ability to control the labour product.

In the past, when production was predominantly to meet food needs, the crucial question was how much could be produced on joint or separate plots. Cultural practices acted as sanctions against laziness and neglect.

As the monetary economy penetrates the rural areas, crop and livestock production has the dual function of having to provide both food for direct consumption and cash to supply other items. The dual functions take on a greater significance as more of the households' "needs" imply the utilisation of cash. When that happens there is concomitantly a greater tendency on the part of the men to control and sell more of the produce to meet their responsibilities and needs, and on the part of the women to require a greater retention at the household level to satisfy their responsibilities and needs.

In this dichotomy the women usually lose out because of their inferior status with respect to cultural practices that associate land rights with control and disposal of the product. Cultural practices and patriarchal relations tend to sharpen as the male head of household tries to exert greater control of the labour product of the other members of the family. This aspect becomes even more dominant where the labour product from the fields or livestock is small and the control has, therefore, got to be tighter (Mascarenhas, 1983). Equity in distribution is not necessarily better among wealthier households, but at least the women's share is sufficient to meet needs and there is more opportunity to intensify production and diversify activities (Kershaw, 1976; Cheater, 1981).

Several studies show that conflict over control can seriously affect women's participation in the production process. For instance, in the Semry I Irrigation Project in Cameroon, such conflicts led to significant differences in the participation of married and independent women in rice production (Jones, 1983). The latter worked more hours in the rice fields than the former. Loss of own plots and thus a decline in the control of their labour product, frustrated women in at least two settlement schemes in East Africa and affected the efficiency of the schemes (Hanger and Moris, 1973; Brain, 1975).

c) Sexual division of labour

Labour is a primary resource in peasant production systems. Peasant studies have treated households as a common pool of labour (Chayanov, 1966; Normann et. al., 1983), but cultural practices differentiate between female and male labour at two levels:

at the level of direct productive activities
at the level of supportive activities.

Table 4.1 Household allocation of labour in agricultural tasks (by stated frequency)

Individuals doing jobs	Number of times mentioned
	Woman		Husband		Children		Relatives		Labourers
Type of job (N)	No.	per cent	No.	per cent	No.	per cent	No.	per cent	No.	per cent
Clearing land (99)	40	40.4	41	41.4	8	8.1	2	2.0	8	8.1
Cultivation* (118)	57	48.3	40	33.9	10	8.5	3	2.6	9	7.7
Planting (109)	60	55.1	30	27.5	11	10.1	2	1.8	6	5.5
Weeding (117)	59	50.4	36	30.8	10	8.5	3	2.6	9	7.7
Harvesting (109)	60	55.1	27	24.8	13	11.9	3	2.7	6	5.5
Pre-storage processing (88)	60	68.2	14	15.9	8	9.1	2	2.3	4	4.5
Storage (98)	59	60.2	21	21.4	10	10.2	3	3.1	5	5.1

Source: Mascarenhas, 1984.

* Cultivation - smoothing the fields, making ridges, etc., prior to planting.

Several processes have modified this division of labour, in most cases leading to extra labour burdens. Firstly, the migration of males to plantations or wage labour reinforced women's role in subsistence production. Secondly, with the introduction of cash crops and new technology and the resultant increased need for intensive processing, male-operated farms required a greater participation of women on their plots.

The introduction of new technology can also lead to higher labour inputs: tractor clearing of land usually means more labour in terms of weeding, harvesting etc. In Malawi women labour inputs were greater in the sprayed fields than in the unsprayed because they were more involved than men in such spraying (U.N., 1984). Better tools for weeding, transport of the harvest and the use of hired labour can ameliorate the situation but these amenities are unavailable for the vast majority of women.

In the more Muslim North African countries women are less involved in field production. They are, however, heavily involved in postharvest processing and storing, the care of livestock, sale of livestock products and some horticultural production around the house. In Egypt, for instance, women perform less than 20% of the field work, whereas they contribute more than 60% of the work with livestock (FAO, 1984d). There are also differences between countries. In Saudi Arabia women do about one-half of all farm work (see Table 4.2).

Table 4.2 Percentage of Female Labour by Agricultural Activity in the Jordan

Ploughing and land prep.	10%
Planting	30%
Weeding	60%
Harvesting	70%
Transporting crops from the field	20%
Processing crops	80%
Storing crops	60%
Marketing crops	10%
Pruning trees	5%
Animal care	70%
Dairy production	80%

Source: Hammad, Hassan I., 1980 (quoted in FAO, 1984d).

Nevertheless there are field tasks in Asia which are considered to be more appropriate for one sex than another. In nearly all rice growing areas, men traditionally prepare the land and lift the seedlings while women transplant, weed and harvest the crop. In addition, women from landless households and those with small plots are also involved in wage labour. In Java, for instance, they account for between 52% and 70% of the hired labour, particularly for weeding and transplanting (FAO, 1983).

Table 4.3 Sexual division of labour in India

	MEN		BOYS		WOMEN		GIRLS		Total
	Hours	%	Hours	%	Hours	%	Hours	%	Hours recorded
Agriculture	2.3	17	0.6	5	1.1	9	0.6	5	222.4
Animal husbandry	1.4	10	2.4	19	1.5	12	3.4	27	297.4
Food processing	0.3	2	0.2	2	1.3	10	0.7	6	113.9
Food preparation	0	0	0.2	2	2.5	20	1.0	8	178.7
Collect water and firewood	0.1	1	0.6	5	0.6	5	0.3	2	67.1
Eating and drinking	0.5	4	0.5	4	0.5	4	0.4	3	87.3
Wage labour	4.2	31	0.3	2	0.4	3	1.6	13	290.8
Child care	0.1	1	0.5	4	0.5	4	1.1	9	49.3
Leisure	3.7	28	4.2	34	3.1	24	2.8	22	577.0
Total	13.3	100	12.4	100	12.7	100	12.4	100	2103.6
Total number of days: 39

Source: Brandtzœg (1982a).

No matter what the differences in the extent and type of field labour, women from all regions in the developing world are invariably involved in performing the major part of the variety of tasks that constitute domestic labour. Increasing distances to fuelwood and water sources because of environmental deterioration can increase the burden even further.

The only recourse that women have is to relieve some of the domestic and field tasks through the assistance of children. This aspect in turn has major implications for high fertility rates which escalate child care burdens and increase the problems of sustainable development.

d) Access to economic services

Women's subordinate position with regard to land and the decision making power about its use also affects their access to other resources such as credit for purchasing tools to relieve the labour burden, and inputs to increase yields.

Some studies have hinted at the existence of indigenous savings clubs. Through this system a group of women put aside a certain sum of money each month which then goes into a common pool. Each of the women in the group takes a turn in using the common pool. There are a few examples where these have been successfully used. In Kenya the saving club method was used by a group of women to produce vegetables, pigs and chickens for sale in the urban markets. Such clubs or rotating societies are found in other parts of the world both among women and men, for example, the arisan in Indonesia, susu in West Africa, gamaya in Egypt, tanamoshi in Japan, etc. (Tinker, 1979).

More empirical data is available on the lack of agricultural information to women. Two significant studies in this context are those by Fortmann (1977) for Tanzania, and Staudt (1976) for Kenya. Fortmann used six (6) sources of agricultural information which included radio programmes, newspapers and agricultural field demonstrations, in addition to personal visits by the agricultural extension officers and found that in all cases women had less access to such methods of extension than men. This affects the quality of production and total output, which in turn affects food adequacy.

Another aspect of this determinant is related to access to the market. Women's participation in the productive sector and their ability to benefit from it depends, in many countries, on their access to marketing the surplus. In some countries cultural practices control the products marketable by men and women. As a general rule, women are less able to sell large amounts of cash crops (U.N., 1984).

Because of such constraints women are compelled to engage in non-field arduous tasks. In Sub-Saharan Africa, the main source of income is often beer-brewing which is a time and energy intensive task in terms of the labour involved in collecting fuelwood for brewing and the preparation of the brew (Mascarenhas, 1983). In Asia, the chief sources of income are often labour demanding home-based crafts or casual wage labour.

The extent of infrastructural facilities can also affect women's participation in marketing. In the Cameroon, proximity to the market was a major determinant of whether women sold crops or not. Those from villages nearby were able to sell their produce and take care of their domestic responsibilities. For women from distant villages it would have entailed an overnight stop which would conflict with their domestic responsibilities (Henn, 1983).

Finally, women's participation in marketing can also be affected by the form of transport that women are culturally allowed to utilize. The remarkable absence of women on bicycles or operating ox-carts in Africa and even in Asia, where such forms of transportation are more common, is a case in point.

4.2 Breakdown of cultural practices

There are some instances of the breakdown of cultural practices. In Africa, the introduction of cash crops has led to a limited breakdown of the rigid sexual division of labour. In Rukwa Region, Tanzania, women do all the weeding for millet, beans and groundnuts; but men participate in this activity for maize, which serves as a cash crop in that region.

In Zimbabwe, women have been able to establish themselves as successful medium size farmers, complete with Master Farmers' Certificates (Cheater, 1981). Other examples can be found from other parts of Africa (FAO, 1984d; Bader, 1975; Kershaw, 1981; Gbedemah, 1978; Mascarenhas, 1984).

In the Middle East, male migration is contributing to the breakdown of women's seclusion from direct field activities (U.N., 1984). Economic stress can also lead to a breakdown in the seclusion process. In Egypt, for instance, women from the poorer Shukyria community were very much involved in cotton picking in spite of religious taboos about women in public places (Murdock, 1979). Such stress is also the reason why generally in Asia, there is a greater sharing of tasks at least in the productive sector.

The breakdown of cultural practices can be beneficial. They can also add to women's labour burden and responsibilities. In Asia the responsibility for providing the household food has traditionally fallen to the male head of household, women have always been unpaid family "help". Economic stress has necessitated that women share and often assume this responsibility. Similar trends are becoming apparent in Africa where the need to purchase food has forced women to devise a variety of sources of income which improve the household food security at the cost of greater labour burdens.

4.3 Seasonality

A number of studies have discussed the impact of seasons on women's labour burdens and household food stocks. A very significant, perceptive and comprehensive work is the collection of articles edited by Chambers et al. (1981). It concludes that

...most of the poor people in the world live in tropical areas with marked wet and dry seasons. Especially for poorer people, women and children, the wet season before the harvest is usually the most critical time of the year. At that time adverse factors often overlap and interact... It is the hungry season and the sick season.

Table 4.4 Seasonal distribution of women's workload in Nigeria (hours/month in agriculture)

	Men	Women
May	67	46
June	147	140
July	114	145
August	141	164
September	93	114
Year's average	85	84

Source: FAO 1984d.

Even in Asia, where women are less involved than men in field activities, seasonality can affect certain classes of women. The poor and landless women are specially dependent on wage labour. Demands for women's labour in rice production, for instance, peak during transplanting/weeding and harvesting/postharvesting work. Other very labour demanding activities are cotton and tea-picking.

In all the above examples, there are additional labour inputs into domestic activities which can constitute between 4 to 7 hours a day (McSweeny, 1979; Bério, 1984a). The implications of adding these hours to agricultural labour during seasonal peak times are considerable - both in terms of expanding or intensifying agriculture; the care of children and the health of children. It also has great implications for any programmes to lower fertility since children, especially girls, are the only means that women have to lower their labour loads.

4.4 Socio-economic differentiation

A number of studies have commented on the role of socio-economic status of the households in the extent and form of women's participation in the food chain activities.

The most conventional factor of differentiation is economic status or wealth - usually of the head of household. Cheater (1981) found that women of wealthy households were often assisted by their husbands to set up independent farms. Fortmann (1977) argues that economic status overrides labour and technology constraints of female-headed households. Nestel (1985) shows that women from better off Masai families had less need to purchase supplementary food.

Socio-economic differentiation can also affect the labour burden. Women with greater access to resources either by themselves or through the higher economic status of the husband could afford to pay for fuel, wood and water collected by other women. The latter, were thus doubly burdened in this arduous task. (Folbre, 1985; Hedkvist and Mascarenhas, 1983). In Java, girls of 10-15 years averaged 94 hours a month on housework and water collection compared to 26 hours in the richest households (FAO, 1984d). In the Punjab, improved incomes at the household level could result in women's labour being substituted by hired labour (Khan and Bilgueses, 1976). In India, too, women from more prosperous farms withdraw from field tasks (Ghodake et al., 1978).

Socio-economic differentiation can also affect religious barriers to women's participation in direct field production (Salem-Murdock, 1979). In Asia, the greater sharing of field tasks between men and women is more prevalent in the households with relatively poorer incomes (FAO, 1984).

However, the result of improved economic status is not always beneficial; it depends on how the head of the household distributes the labour product. The concept of joint utility at the household level, which improves as total household resources improve, has been challenged by several authors (Folbre, 1985; Guyer, 1980). The study from India (Ghodake, 1978) shows that as women's labour is withdrawn, their decision making power as to the use of the labour product also declines. In Africa where wealth is often associated with polygamy and cattle, increased wealth at the household level may not be utilized to improve the status of existing wives and their households (Jakobsen, 1978). Women from richer households may thus have to resort to poorly paid and time-consuming activities to satisfy the household needs even in households that are overall economically well off.

Another aspect of socio-economic differentiation is marital status. Older women, for instance, have better decision making powers than younger ones in many African societies (Cheater, 1981). Some polygamous households can act as extended families with women assisting each other in productive and domestic activities. Women with older children who can help out with field and home activities have greater options about strategies to meet the household needs.

Women's age, however, does not give them greater security in case of divorce, separation or widowhood. Land rights are usually vested in male heads of households and are passed on to sons. A dissolution of marriage usually leads to loss of all rights to land, the basic resource (Andah, 1978; Mascarenhas and Mbilinyi, 1983; Okeyo, 1980; Muchema, 1977). In Sub-Saharan Africa, divorce also results in the loss of working-age children, thus losing labour as well as land. In other cases, women can be left with the children with little means to look after them.

These are some of the factors that contribute to the fact that female-headed households are among the poorest of the poor (Brown, 1981; U.N., 1984; Faruquee and Gulhati, 1983; Mencher, 1981).

Due to unfavourable inheritance laws and less capacity to withstand economic adversities, female-headed households tend to be more often landless than male-headed households, a characteristic maintained and further reinforced by agrarian reform (U.N., 1984).

4.5 Macro-level factors of change

Some reference has already been made to changes that have affected traditional practices. Three aspects, however, need to be examined in greater detail because they directly affect the dual aspects of women's participation in the food chain: the provision of food, and the constraints of resources to make this effective.

a) The spread of the monetary economy and cash cropping

The spread of the cash economy has affected rural livelihood patterns in some very profound ways resulting in both potentialities and constraints. Potentially, income can be used to purchase food and diversify diet and thus cover deficits in agricultural production. Many women now have several sources of income, the most common being small sales of surplus from the fields, cooked food, and local beer and finally agricultural wage labour.

The ability to generate income is specially important for women in some regions and in some types of households. In Asia, the generally greater scarcity of land and lower ability to utilize the high yield varieties makes supplementary cash a matter of survival. In Thailand, for instance, women could not live without money to buy food.

Even their basic diet of vegetables and salted, dried fish must be purchased in the market (Tinker, 1979).

In other cases, it can be a necessity to continue agricultural production. In Iringa, Tanzania, women from households with average economic indicators, engage in picking tea leaves in order to earn cash to hire labourers to clear their land (Mascarenhas, 1983).

In addition, income is necessary to ease some of the labour burden: cash can be used to pay for costs for milling instead of hand pounding; one can ride a bus instead of walking.

However, the penetration of the cash economy can have some very negative aspects particularly if it is associated with cash cropping and intensification. Several themes emerge in the literature on this topic.

A considerable number of studies assert that the cultivation of cash crops have been at the expense of food crops (Fleuret and Fleuret, 1980; Jakobsen, 1978; Reining, 1970). Women's rights to land in Africa become even more precarious when women's labour becomes critical to the process of capital formation through the production of cash crops. In many circumstances food fields get delegated to poorer land (Reining, 1970) or to more distant areas. In the Cameroons, land near the villages was all taken up by coffee and cocoa plantations while food fields were delegated to areas that were about six kilometers away, involving "a one and one-half-hour walk to the food fields over rough forest paths, often with slippery stream and marsh crossing" (Henn, 1976, quoted by Tinker, 1979).

The use of more land for cash rather than food crops becomes especially critical in new settlement schemes where the tendency is to give the proceeds from the sale of crops entirely to men. Land for food and for women's sources of income are either completely neglected or given a low priority (Hanger and Moris, 1973; Murdock, 1979; Brain, 1975). Depending on how the cash from the sale is used, the practice can not only affect women's contribution in the productive sector but also the amount and kind of nutrients available to the household.

Another theme, is that cash-cropping puts an extra labour burden on women. In Nigeria, for instance, a project to produce rice had the potential of doubling women's workload (Burfisher and Horenstein, 1985). In Asia, wives of small farmers who received some profit from the HYV-irrigation schemes to intensify wheat and rice production had to shoulder much higher labour burdens as "unpaid labour" (Khan and Bilgueses, 1976).

The effect on women in Asia, however, can be more varied. In some cases agricultural intensification can provide employment opportunities but here too there can be problems. The greater emphasis on cash as the form of exchange can result in women being paid in cash instead of the more preferred payment in kind. Furthermore, the expansion of HYV's has led to a decline in sharecropping, thus depriving some households of the basic means of production and thereby expanding the class of landless women and men (Griffin, 1970).

A third theme in the literature related to this aspect, is that the development of the cash economy can lead to women assuming many of the responsibilities for cash purchases that were previously considered as the man's share. Women have to assume these new responsibilities for several reasons. In the first place the combination of small plots, poor technology and declining yields, particularly in Africa, result in low outputs. With the best of intentions men have to spend a disproportionate amount of the cash received to pay off loans, and other costs incurred in the production of cash crops leaving little to share with their wives.

On the other hand, as women acquire more income, men tend to put more of their traditional responsibilities on women without compensatory sharing of work in domestic activities or even some of the agricultural tasks. The ability to acquire some cash can thus make women worse off in terms of sharing the total household labour product.

Moreover, the activities themselves can be very time consuming. In Tanzania, for instance, the most common source of income is beer brewing, but it does involve extremely heavy inputs of labour particularly in terms of collecting fuelwood and water. On an average, beer brewing required two to three times the labour and time inputs that were required for collecting fuelwood for domestic purposes. Preparation of sunflower oil, another lucrative activity, required even more (Mascarenhas, 1983). The degree of self-exploitation was therefore very high because many of the tasks were very time and energy consuming.

Finally, the need or desire to participate in such activities can be at the cost of household food security since the beer is made out of the staple foods - maize or millet. There is a false belief that it is easier to purchase rather than produce food, particularly as production techniques on the small plots are so arduous and time-consuming. The need to reduce the drudgery of work is very real. Similarly, there is a temptation to reduce the vagaries of climatic change by using cash income to make up for food shortages. In the long run there will be disastrous consequences to the food production sector if all women were to adopt this attitude. On the other hand, newer less arduous income generating activities, not involving the use of household food supplies could add to household food security.

b) Urbanization and migration

The intensification of cash generating activities among women can also be caused by the growing tendency among men to migrate to urban areas or other large sources of cash employment, resulting in the growth of the prevalence of female headed households. In Africa some sources put the rates at 45 per cent for Kenya to about 30 per cent for Malawi and Lesotho and 25 per cent for Tanzania (Fortmann, 1977; Muchema, 1977).

Evidence on the impact of such male out-migration is somewhat mixed. The more general tendency is to consider such migration as detrimental to women, in the sense that it deprives women of much needed male labour without compensatory cash to enable women to hire labour or equipment (Bukh, 1977; Palmer, 1985c; Brown, 1981). Even women on the smallest farms experienced an increase in their farm work. In some cases women were forced to lease their land to sharecroppers, thus getting less than they would if the family had cultivated the land (FAO, 1984d). Kumar (1983) asserts that women in female-headed households worked harder than in male-headed households but that their output was lower. Brown (1981) categorically states that women who were heads of households were among the poorest.

Some studies, however, argue that male absences may be beneficial to women to meet their needs including the family food needs. Staudt (1985) claims that "male absence enhances control over produce". In a study of small commercial farms, Cheater (1981) shows that women who managed farms of absentee sons or husbands had a high degree of decision making power including the use of the product. In Jordan, male migration increased women's participation in the productive sector. Fortmann (1984) uses data from Botswana to show that the ability to hire equipment or labour for clearing land was more crucial than merely the fact that adult males were absent.

Palmer (1985c) shows that, whereas in Africa earnings from male migrants do not generally benefit the women left behind, the opposite is true for Asia and Middle Eastern countries. The critical factor is the level of earnings and the duration and purpose of migration. The shorter the duration, the more likely are the men to invest in their farms to which they have to soon return. Obviously, the issue is complex and needs to be analysed in relation to other aspects of the rural households.

c) Under-development: its characteristics and effects

Micro-level communities cannot be analysed outside the context of the larger question of the characteristics of the developmental processes prevailing in the nations in which they are located. Developing countries are often categorised into different groups (see for instance World Bank Annual Reports) but some generalisations can be made.

In most developing countries the economy is characterised by its dependence on a market over which they have little control. As said earlier, prices for products and labour are very low so that survival strategies necessitate involvement in several simultaneous activities, not one of which can sometimes adequately cover the needs of the women and men in rural households. The unequal rate of exchange affects both women and men but the former are doubly affected because of their second level of unequal exchange in which tradition dictates that men have a right to use women's labour without strictly equitable recompense. Women, thus have to bear the burden of poverty over and above their traditional workload. Poverty means a low level of technology, extremely intensive methods of cultivation on small plots with declining productivity.

It also means a low level of social and economic infrastructure. In most cases such basic services such as efficient and regular transport, prepared and processed food (e.g. packaged maize flour for ugali) which make women's life somewhat easier in the urban areas are totally missing in the rural areas of the developing countries. Those who extol healthy "fresh" homemade weaning foods seldom consider the toll on women's energy and time.

Under-development is also characterised by irrational national public policies particularly with respect to food production. Several studies show how the food sector, until recently, has been neglected and demeaned, with resultant demeaning of women's activities. Murdoch (1980) has shown how the cash crop sector in India was allocated 6 times the funds for research compared to the food sector. In Tanzania, the First Five Year Plan completely ignored the food sector. Where public policies have intervened it has been to introduce new crops and techniques that have had the result of increasing food losses (hybrid maize is more susceptible to the stalk-borer), reducing variety of crops grown, or inducing the sale of traditional food crops for export or sale. For instance, in Tanzania, all sunflower seed is supposed to be sold to a government marketing organisation for export, although these seeds were traditionally used in foods to enrich the diet and are now being utilised to prepare cooking oil by several women. If women were to observe the State's regulation they would lose a valuable source of calories for their meals.

4.6 Women and technological change

The impact of technological change on women's participation in the food chain varies according to regions and the classes of women involved.

In Africa, technological change has generally meant the use of tractors or ploughs for land preparation. It is interesting that this, of all the agricultural tasks, should be of some concern in a continent where the greatest labour bottleneck is often weeding and harvesting. Unfortunately, the trend is still prevalent. In Tanzania a recent farm mechanisation study showed that of the oxen drawn equipment (insufficient as it is) 93.4 per cent consisted of ox-ploughs, 1.5 per cent of ox-carts, 3.4 per cent of harrow/weeders and 1.7 per cent of planters. The bias towards relieving men of their tasks is obvious in the predominance of ploughs for clearing land. The rationale that ploughs will lead to increased acreages has been negated in the face of evidence that shows that acreages have only increased where extra labour for weeding was available. Further mechanisation would be a boon to increasing acreages and total production and reducing labour burdens.

In Asia, on the other hand, mechanisation of agricultural tasks shrinks women's labour opportunities which are especially necessary for landless and female-headed households. Such mechanisation, albeit unintentionally, seems to affect women more than men because it is women's tasks in weeding, harvesting and postharvesting that are most affected. Yet, it is probable that for women from small farms that can generate an adequate surplus, such technologies can be beneficial in increasing productivity and reducing labour inputs.

Another aspect of the bias is the fact that technologies are more concentrated into improving production but not for activities in the other stages of the food chain - food storage, processing, preparation of food (Carr, 1981).

A third source of bias is that which makes a sharp distinction between "women's work" and "economic work". Equipment used for cash crops is not easily available for food crops or domestic activities. For instance, in Tanzania a water tanker, using a stream, was used to provide water to irrigate tobacco seedlings. A tractor-trailer was used to collect fuelwood to cure tobacco. Any suggestions that the tanker could just as easily supply water for dry season vegetable gardens or provide households with water was considered "impossible". The idea that the tractor-trailer could make two trips a week to supply the entire village with fuelwood was dismissed as "uneconomic" (Mascarenhas, 1984). Women's work in provisioning the household was obviously considered as not economic. Both in Africa and Asia, when men and women do the same field task, they use different equipment with the latter using the more traditional, less efficient equipment (Brandtzaeg, 1982a; Carr, 1981).

i) Technologies that have the potential of decreasing women's economic benefits

Other studies, however, point out that the introduction of new technologies is very complex and has both potentials and constraints, depending on the context in which the technology is introduced. In Java, the introduction of a new technology for rice hulling resulted in the loss of 125 million woman days, and a loss of $ 50 million, mostly for the women from landless and small farmer households. On the other hand, the new process lowered the price of rice - resulting in almost three times the value of the lost jobs. In India, women harvesters using small knives, used to obtain about 10-15 per cent of the harvested grain and leave about another 10 per cent in the field for the poorest families. When sickles were introduced the task was taken over by men who only obtained about 6 to 8 per cent of the harvest. Even if the ensuing harvest was bigger the lower percentage did not compensate for the loss of grain. The loss of employment opportunity for women also had the potential of reducing the total food in households. Furthermore, such loss had the potential of forcing women to look for even more difficult jobs such as road or building construction (Tinker, 1979).

The issue is obviously complex because in both cases there were gains at national level, and also among some households - the better off and even the average landowners. Thus an introduction of new technologies can have gains and losses; national interests can override local interests; the interests of some can be overlooked in the face of the advantages for others.

ii) Decline in control over the labour product

Another trend emanating from the studies dealing with the introduction of new technology has shown that when certain tasks that women performed were mechanised the control of the new technology and its product invariably accrued to men. In Nigeria, for instance, the home-based oil extraction activity was traditionally a woman's activity. When a hydraulic press was used to extract oil, the operation and all the oil went to men. Use of the press declined from 76 per cent in the first year to 24 per cent in the second (Janelid, 1975).

In Ghana, pottery was largely the work of women but with the introduction of a potter's wheel, the industry was taken over by men. Women were unable to take advantage of the new technique because of lack of credit to purchase the new wheel (Carr, 1981). Such incidences are quite frequent and can reduce the role and status of women in the food chain, and consequently their ability to provide for the necessities of their households. The assumptions of most organizations that have introduced new technologies have been that overall efficiency in output will benefit all households equally and that what benefits one member of the household is good for all. In some cases neither assumption may be true so that there is a need to see net gains or losses with respect to household food and nutrition needs.

iii) The "successes" of new technologies

All technologies have not affected women adversely. One of the best successes have been mills to grind grain. Particularly in Africa, these are usually controlled by male individuals or groups, but they are still considered as a mixed blessing. In many instances, young boys who would not participate in pounding grain in the home mortars could be relied upon to take the grain to the mill, stand in line to have it milled and return with the flour thus saving women several hours of hard work. Control of the machinery in this case is less important because there is no loss of product used for food or sale and there are sizeable gains in labour saving. Wherever donkeys or animal drawn sleds or carts are available, one often sees even men involved in such activities, as for instance in Rukwa, Tanzania (Mascarenhas, 1986). The availability of maize mills in this region has also lightened the work for preparing flour for brewing, thus giving women a double advantage.

In Kenya, and Mauritania improved bee-keeping (hives stand about 3 feet off the ground) has enabled women to participate in an activity that was mainly male dominated because it involved climbing trees (Carr, 1978). These are just two examples of how technologies can benefit women by enhancing the food and cash income at the household level.

To sum up this section, the factors which affect women's participation in the different parts of the food chain are varied. They range from cultural to economic and technological aspects, and from micro-level to macro-level determinants. Overriding these issues are modifiers such as regional differences, socio-economic differentiation and seasonality. The review has also emphasised the complexity of these determinants. Many have constraints for women, but there are also potentialities. The complexities of the effect and inter-relationship of these determinants need to be better understood. So far there has been a tendency to compartmentalise the issues and their effects, perhaps because the inter-relationships are so complex.