Module III: Gender, rural fertility/mortality & farming systems

The agricultural gender division of labour

Linkages between fertility, mortality and farming systems and their implications

Main issues and research needs

Endnotes

Module III: Gender, rural fertility/mortality & farming systems

(Topouzis/du Guerny, SDWP, November 1995)

What have Gender and Rural Fertility/ Mortality to do with Farming Systems? ^a

(^a A farming system is a group of farm households operating in a more or less homogeneous agro-ecological setting that have similar socioeconomic characteristics and comparable resource endowments as well as similar constraints and opportunities for development.

Acknowledgements: This module is the product of a collaborative effort between the Population Programme Service (SDWP) and the Farm Management and Production Economics Service (AGSP). Special thanks are due to Karl Friedrich and Horst Wattenbach who devoted much of their time and contributed significantly to the module and to Tim Aldington, Alain Marcoux and Heiko Bammann for their comments.)

Fertility and, to some extent, mortality, levels and trends in rural populations have traditionally been analyzed in terms of socio-cultural and economic variables, including the value of children, age at marriage, malnutrition, prolonged breast feeding, education, employment, and the status of women. In the past decade, however, institutional factors and gender relations have clan hewn recognized to influence fertility and mortality patterns. This module focuses on the gender dimension of the linkages between fertility, mortality and farming systems, with emphasis on sub-Saharan Africa. In particular, the following questions are addressed: What are the implications of fertility and mortality on the dynamics of farming systems? What are the demographic implications of changes in farming systems? How can policy interventions create a desirable synergy between evolving farming systems and population variable?

About 80% of African women live and work in rural areas under conditions that support and sustain high fertility and where the expected economic contributions of children is substantial. In fact, sub-Saharan Africa exhibits the highest rates of economic activity and fertility and the highest levels of maternal and child mortality in the world. Source: Aderanti Adepoju and Christine Oppong (eds.) Gender, Work and Population in Sub-Saharan Africa, Geneva: ILO, 1994, pp. 1, 17

(¹ Fertility and Mortality in sub-Saharan Africa)

The objective of this module is to stimulate discussion among farming systems, gender and population specialists in order to: a) identify the main policy issues emerging from the interface between gender, rural fertility/mortality and farming systems; b) indicate what research is needed to address these issues; and c) define possible areas and/or points of intervention for rural development and population programmes and policies, given existing knowledge.

"What is not Counted is Usually not Noticed" J.K. Galbraith

No statistically observable relationship has been found between different type; of agricultural work and fertility to date in sub-Saharan Africa.¹ This has been attributed in part to difficulties in defining and measuring women's economic activity, as a substantial part of women's farming remains officially undocumented. Women's contributions that have most suffered from neglect and bias include 'informal sector activities,' 'home-based production,' and work on 'family farms.' Another problem is that different countries define and measure women's work in diverse ways.² Therefore, it is not surprising that demographic data linked to information on female or male economic activities are lacking or based on estimates. ^b

(^b No country in Africa has complete vital statistics and a published national census within ten years of a continuous population register, according to Oppong. The sources of demographic data available include censuses, survey and administrative records. Demographic surveys include the World Fertility Surveys (which is weak on women's economic activity rates), the Westinghouse Demographic and Health Surveys fin which only Ghana had a module on women's economic activity); surveys undertaken in the context of the African Household Survey Capability Programmer and socio-economic surveys with demographic components, such as the Living Standards Measurement Study of the World Bank.)

The agricultural gender division of labour

Although gender-disaggregated data on production is essential to making women's contribution visible and to identifying gender-specific needs and constraints, it is not the "panacea for improving the visibility of women; it only uncovers the tip of the iceberg."³ Below that lie conceptual and methodological problems that are often reflected in gender stereotypes which have a marked impact on agricultural and rural development policies and programmes. Such stereotypes are themselves a manifestation of gender bias and of the absence of gender analysis. For example, it is commonly assumed that men cultivate cash crops while women grow food crops; and that women are subsistence farmers while men are commercial farmers. Official statistics from a number of countries, however, reveal a different picture: in Ghana, for instance, one-third of farmers producing cash crops like cocoa, rice and sugar cane, and one-quarter of farmers producing tobacco, coconut and oil palm are women.⁴

Gender-disaggregated data alone do not reveal the complexity of gender dynamics, inter-relationships, interests, responsibilities and decision-making processes within the farm household. As many development projects are designed without the participation of women beneficiaries, their planned role therein is often not one they are able or willing to undertake. This raises the question of whether gender roles, needs, constraints and economic decision-making within the household have been adequately understood.⁵

Gender-disaggregated data and gender analysis of the farm household system need to be bolstered by a corresponding allocation of resources and by participatory methodologies to ensure female involvement in policies and programmes. A review of agricultural projects funded by USAID showed that projects that delivered resources to women according to their role in the farming system were more likely to succeed than projects in which women did not receive resources. Even if a mainstream project was concerned with activities that were primarily women's responsibility, women's participation was low unless delivery systems explicitly earmarked resources or services to women.⁶ Gender analysis in baseline situations did not automatically guarantee that female farmers would participate, even when there was no formal barrier to their participation. Socio-cultural norms, traditions and stereotypes, often related to female fertility and reproduction can be formidable obstacles to women's involvement in programmes and projects.

Linkages between fertility, mortality and farming systems and their implications

a) The Interface between Gender, Farming Systems and Demographic Change

Farming systems to a large extent depend on complex inter-relationships between men's and women's labour. At the center of these inter-relationships, however, lies an asymmetry o, male and female interests, duties, obligations and contributions within the Arm-household. More specifically, the extent to which activities, rights and decision-making are undertaken jointly or separately is relevant to almost all aspects or ramify life, including property ownership and management, production and management of produce. domestic tasks try etc.⁷

Five main gender patterns of farm management have been identified: separate enterprises, separate tasks, shared tasks, separate fields, and women-owned or women-managed farms. While men, women and children work to the same degree during the peak or harvest period, women and children's work is predominant in the off-season. In addition, men's labour input is most critical for a more narrow range of farming activities (especially land clearing and herding livestock), or else it is performed once yearly (construction of dwelling units, thatching huts and granaries). In contrast, women's labour input is constant throughout the year, encompassing a wide range of labour intensive, often tedious, tasks. Women's labour is also characterized by high fragmentation in terms of time.

When farm and domestic tasks are combined, women's working days are considerably longer than men's. In Asia, the Pacific and in Africa, women average 12 to 13 hours more work a week than men. In the Philippines, one study found that when housework and childcare were taken into account, the average hours worked per week by women reached 70, compared to 57 for men. In Latin America and the Caribbean, women work 5.6 more hours a week than men.⁸

TABLE 1: Women in most regions as much or more time working than men when unpaid housework is taken into account

Note: Numbers in parentheses refer to the number of studies in each region.

Source: Data are averages from a small number of studies in each region, 1975/88, compiled by the Statistical Office of the United Nations Secretariat.

Farming systems cannot be understood without taking into account the evolving and complex family roles and relationships which underpin them and the competing demands made upon women's time and energy resulting frown their productive and reproductive roles.⁹ In sub-Saharan Africa, and in other developing regions, a large proportion of economic activity takes place in familial contexts (family-based enterprises) and most productive resources are in the control of kin. This is especially the case for women who are primarily involved in family or home-based economic activities, like food production, processing and distribution. ¹⁰

Working with accurate labour/production profiles at the farm-household level and understanding intra-household dynamics is critical to identifying and addressing male and female roles and constraints as well as to assessing intra-household dynamics. This, in turn, is critical for effective policy-making and programme planning.

Key concepts underlying the collection and analysis of data for economic and demographic surveys, like 'household,' 'parenthood' and 'marriage,' have often been poorly conceived and operationalised, thereby perpetuating stereotypes. For instance, as marriage in parts of Africa is potentially polygamous, female-headed households (not resulting from male out-migration) are common (see Module I)¹¹. This may considerably complicate the definition of the farm household as an analytical unit. According to Adepoju and Oppong, "women's work has frequently been so inadequately recorded that it is virtually impossible to make satisfactory correlation’s between variations in women's economic activities and demographic evidence regarding differences and changes in birth and death rates."¹²

At the micro level, assumptions about the nature of family systems, particularly in the case of Africa, are based on the stereotype of the nuclear family unit (husband, wife and children). According to this stereotype, the nuclear or conjugal family is synonymous with the domestic group and simultaneously forms the unit for economic and demographic decision-making and related activities. For example, erroneous assumptions have been made that women's work is incompatible with high fertility and that women as wives are primarily unpaid family workers whose economic interests coincide with those of their husbands.¹³

The following working conceptual framework outlines some linkages between division of labour by gender and age, farming systems and household demographic characteristics and strategies. The framework assumes that families optimize the mix of these variables and that there is a rationale for existing patterns between them. Changes in farming systems tend to be more rapid than changes in the division of labour and the household demographic characteristics and strategies. This can cause dysfunction’s in the last two that can be systemic, triggering, in turn, changes in the farming system itself.

In short, the farming system, farm household labour and demographic profiles are linked in such a way that changes in one set of variables has direct or indirect consequences on the other two. To ensure that changes in farming systems are synergistic with changes in the labour profile and that they do not result in undesirable changes in demographic variables, it is important to look at the three sets of factors as forming an inter-dependent system. In case of dysfunction, what needs to be considered is the new optimal mix between farming system, labour profile and household demographic factors and which of these factors need to be changed through appropriate policies.

FIGURE 1: Household Labour Profile, Farming System and Demographic Profile

Development interventions have often had adverse consequences on rural women's socio-economic status and productive/reproductive roles largely because the linkages between farming systems, labour profiles, and population over time have not been systematically taken into account. Such interventions include:

- Changes in cropping patterns or crop mix, emphasizing cash crops for export at the expense of subsistence crops.

- New technologies whose introduction displaces more women than men farmers, and whose use by men only-widens the productivity gap between women and men. Some production technologies increase the labour burden on women farmers without increasing their share of or control over farm revenue.

- Changes in land access, from communal ownership in which women had secure access to land to private property in which only male heads of household can hold title.

- Agricultural extension services, credit, inputs, technical assistance, etc. which tend to be targetted to men only, thereby widening the male/female productivity gap, marginalizing women even further.

By overlooking the demographic implications of changing cropping patterns and new technologies and by not ensuring women's access to services, many development interventions have either resulted in unchanging high fertility (and mortality) levels, or may even have contributed to increased fertility, as women try to compensate for their increase in workload, displacement or marginalization by having more children.

b) Labour Bottlenecks and Fertility/Mortality

It has been shown that seasonally of labour demands on time and energy can have an adverse effect on women's ability to cope with the demands of farm work, pregnancy or lactation.¹⁴ In Eastern Uganda, women's farming work peaks during the rainy season (June-July and January).

"It is in this period that a family must acquire most of its food requirements until the next rainy season. Agricultural activities are therefore at their peak... Yet the rainy season is a period when most women are pregnant, having conceived in the slack period when there are many festivities. But they cannot sit and rest as any medical personnel would advise. They must perform all other duties as usual, that is household duties, food production and processing plus providing all these services essential for the family. The demand made on women's time and energy is too high to match their energy intake. This is because at that hectic time, particularly just before the harvest, food is always scarce in rural areas.”¹⁵

A 1994 study on the influence of reproductive status on rural central Kenyan women's time use has shown that the demands of reproduction require Embu women to decrease the amount of time spent on subsistence agriculture, tending animals, commercial activities and housework. Agricultural and economic work activities were found to be particularly curtailed in the third trimester of pregnancy and the first period of breast feeding. Over the two-year cycle of pregnancy and breast feeding, women allocated approximately 53 eleven-hour work days less to subsistence agriculture and other work activities than women who were not pregnant or breast feeding. This labour loss can place considerable stress on Embu household food production and income-generation, it is argued, particularly among poor households.¹⁶

One insight generated by this data is that agricultural development programmes need to ensure that they do not increase household labour demands to levels that cannot be met by women. The authors argue that "excessive direct demands on women's labour, or indirect demands due to changes in labour patterns of other household members may force women to work beyond what is physically good for them and their infants. This could force women to decrease the amount of time spent in agricultural programme activities to the extent that it could put those programmes at serious risk for failure."¹⁷

Periods of strenuous farm labour, such as planting and weeding, have also been associated with low birth-weights, early weaning, and thus short birth intervals and pressures toward higher levels of both infant mortality and fertility. Weeding, done in the rainy season when diarrhea illness peaks is also the period when food stocks are lowest. Pregnant women have been known to lose weight and seasonally of birth weights has been recorded.¹⁸ It has also been shown that women without a co-resident man to share tasks with, have shorter periods of breast feeding duration, indicating their increased economic constraint to work outside the home. Thus, breast feeding and weaning to some extent may affect family size by impacting on fertility and mortality.¹⁹

TABLE 2: Comparison of the Number of Hours Devoted to Activities by Non-Pregnant Non-Lactating (NPNL) Women and Pregnant/Lactating Women Over a Two-Year Period

	Hours per two-year period
Activity Category	Non-Pregnant Non-Lactating 24 months	Pregnant (9 months) & Lactating (15 months)	Difference
Child Care	403	1020	+ 617
Other work:
Animal Care	329	248	-81
Commercial	1062	814	-248
Food preparation	1049	1008	-41
Housework	1136	1048	-88
Manufacture	194	204	+ 10
Subsistence agriculture	1630	1495	-135
Wild foods	0	4	+4

Source: Derived from "The Influence of Reproductive Status on Rural Kenyan Women's Time Use," op. cit., 1994, p. 352.

c) Fertility, Agricultural Production and the Socio-Economic Value of Children

High fertility rates reflect the socio-economic value of children, particularly in rural areas²⁰ as well as high infant mortality rates. Children in developing countries contribute significantly to household labour: they assist in gathering and processing foods and in selling them in the market; they take care of animals; the girls help their mothers care for younger children and also carry water and fuelwood and help with the cooking; the boys help in irrigating, fishing and harvesting. Once they grow up, children are also expected to take care of their parents in old age.

As water and soil resources are depleted due to environmental degradation and women's agricultural responsibilities increase, the need for additional children becomes more important to compensate for the increased labour demands of the household. " (Children) provide desperately needed labour to assist in farm, home and market. They provide the links of kinship without which wives have no enduring rights in their marital homes or husband's assets, including land and thus security as well as economic status in old age. Without them conjugal links are tenuous and fragile. Without them daily laborious tasks cannot be completed. Without them a woman in virilocal marriage remains an outsider, marginal”.²¹

Table 3: Labour Profile by Age and Sex in Rajasthan, India

Time allocation (no. of hours per day) revealing segregation of activities by age and sex - Rajasthan
	age	5-9		9-14		14-19		19-34		34-44		44-70
	no. of	m	f	m	f	m	f	m	f	m	f	m	f
	sample	97	87	154	138	63	58	180	215	91	98	160	91
ploughing digging				0.05		0.57		0.75		0.87		0.79
irrigation fields				0.10		0.39		0.48		1.61		0.70
harvesting				0.05	0.08	0.14	0.25	0.10	0.04		0.25	0.06	0.21
groundnut picking			0.34	0.02	0.28	0.16	0.50	0.04	0.21	0.10	0.18	0.13	0.31
vegetable picking		0.04	0.15	0.02	0.21		0.05		0.12	0.17	0.28	0.01	0.22
cutting grass from fields		0.11	0.55	0.32	1.65	0.56	1.29	0.51	1.08	0.57	1.69	0.48	1.02
weeding fields				0.12	0.42		0.54	0.06	0.66	0.69	0.71	0.63	0.83
total agriculture		0.51	1.63	1.38	3.06	2.00	298	275	2.44	6.31	3.62	4.04	3.05
husking, winnowing			0.01	0.04	0.07		0.14	0.02	0.08	0.23	0.34	0.12	0.11
parboiling grain husking		1.12	0.87	0.61	0.81	0.54	0.20	0.35	0.12	0.56	0.05	0.30	0.18
cattle/goat grazing		0.02	0.11	0.13	0.13	0.38	0.22	0.22	0.28	0.48	0.42	0.55	0.55
making cow-dung cakes			0.16		0.33				0.18		0.17		0.10
total allied		1215	128	0.94	1.60	28	0.68	0.68	1.13	1.38	1.38	1.21	1.43
service					0.06	1.29	1.35	1.35		0.36	0.03
prod. of strawmats, ropes					0.02				0.02		0.03		0.09
selling goods (stationery, fish, grain)				0.11		0.47	0.46	0.46		0.01		0.09
manual labour							0.37	0.37	0.03	0.31		0.21
total non-agriculture				0.16	0.04	1.76	2.70	2.70	0.10	1.00	0.04	0.37	0.09
Cooking (grinding, cutting, etc)			0.26	0.04	0.74		0.03	0.03	2.34	0.10	2.79	0.08	1.60
sweeping washing clothes and utensils		0.01	0.45	0.03	0.56		0.01	0.01	1.09	0.02	1.02	0.12	0.53
fetching water			0.16	0.03	0.36				0.52	0.01	0.41	0.03	0.23
fetching fuel			0.01		0.07				0.08		0.09		0.09
total household, activities		0.01	0.89	0.10	1.75		0.05	0.05	4.05	0.14	4.33	0.25	2.47
schooling		1.71	0.50	2.61	0.41	1.72
playing children		2.55	1.43	1.14	0.24	0.39
time spent of child care		0.16	1.71	0.40	1.23	0.20	0.07	0.07	1.13	0.17	0.69	0.15	0.91
total schooling and child care		4.42	3.77	4.15	97	2.33	0.08	0.08	1.17	0.18	0.63	0.15	0.92
total include activities not listed

Source: Jain and Chand in Saradamoni (ed.), Women. Work and Society. 1985, cited in Most Farmers in India are Women, FAO, 1991, p. 6.

In Rajasthan, India, girls of 5-9 years contribute more than twice as much as boys in terms of agricultural activities, household chores and childcare and continue to contribute significantly more than men throughout their lives (see Table 3). In Africa, children under 13 years contribute around 113 of their father's labour input, less of their mothers.²² Girls of 7 years of age contribute several hours of work each day pounding millet and spinning cotton. This increases to 4 1/2 hours at 9 years (including fetching water and washing the laundry), 6 1/2 hours at 11 years (including firewood porterage) and 7 1/2 hours at 13 years. At 15 years of age, a girl spends almost 10 hours a day in productive tasks. Against this, boys of 7 years average 1 1/2 hours, reach a peak of 6 hours at 13 years, and fail to 4 1/2 at 15 years.²³

In rural areas, the "net" value of children (children's present economic value plus anticipated support in old age, minus the cost to raise them) to their parents, and especially to women, is most often positive. This means that high fertility is a rational strategy. The larger the number of children, the greater the chances of a more efficient and timely execution of farm and domestic tasks.

The persistence of large families can be, in part, explained by the division of labour, labour force needs and access to land (which is discussed at length in Module II). This is especially true of subsistence agriculture ^c, where no major changes are being introduced (e.g. in technology, commercialization, etc.) that would reduce labour-intensive production systems.²⁴ New technologies, including mechanization, fertilisers, etc. can induce changes in farming systems, but may end up increasing further the need for women's and children's labour, especially in the off-season. There is no clear evidence yet of differentials in high fertility and large family sizes between areas with rural development projects and those without.²⁵

(^c The term "subsistence", as used here, does not indicate a minimum of food and shelter necessary to support life; and "subsistence production" is not used to signify "subsistence living." Small subsistence farms range from those which produce only just enough to meet the needs of domestic family consumption to those that sell about 50 per cent of their products (La-Anyane, S., The Agricultural/lndustry of Western Africa. Accra: Ghana Universities Press, 1988, p. 20).

Children's contributions in farming systems and the farm-household economy are not usually quantitatively defined and the implications of their contributions are not built-in related programme interventions. For instance, when children are in school during peak labour season, the women are left with an increased labour burden. Poor rural households may have to take children out of school in order to have enough labour on the farm and at home.

The implications of children being taken out of school are far-reaching: it has been found that education can increase agricultural productivity by augmenting the productivity of measured inputs and enabling farmers to respond more quickly to disequilibria. The social rates of return to education have been calculated at 25-40% for Malaysia, 14-25% for Thailand and 7-11% for South Korea.²⁶ It has also been found that girls are more likely to be taken out of school than boys. Research has demonstrated that young women who have had some education are more likely to have a fewer number of children. For instance, in Namibia, women with no education have a total fertility rate (TFR) of 6.6 children, those with less than 7 years of primary education have a TFR of 6.1, those with 7-8 years of primary education have a TFR of 5.2 and those with some secondary education have a TFR of 4.1.)²⁷

At present, the division of labour is analyzed in terms men's and women's contributions. This method, however, does not take into account the entire farm-household, which also comprises children, who, as shown above, contribute substantially to the household economy and the elderly. A more accurate picture of the farm household's labour profile and a better understanding of intra-household dynamics would emerge if the contributions of boys and girls were also included on the basis of person days for each on- and off-farm activity and household task.

d) Mortality and HIV/AIDS

Maternal and infant mortality are closely linked to childbirth and to women's productive role. In recent years, economic recession and structural adjustment and stabilization programmes have contributed to a decline in the health and nutritional status of women and infants. More urgently, however, the spread of HIVIAIDS is also contributing to a dramatic rise in adult, infant and child mortality in some sub-Saharan African and Asian countries. In Uganda, life expectancy is projected to decline, in part as a result of HIV/AIDS: in 1975-1980, life expectancy stood at 47 years; by 2,0002,005 it would have risen to over 54 years, but is now projected to fall to under 44 years -- 10 years less than originally projected. In Zambia and Zimbabwe, the decline in life expectancy could even be more dramatic: while in 1975-1980 it stood at about 49 and 54 years respectively, by 2,000-2,005 it is expected to drop to 46.5 and 51 years instead of 60 and 66 years (in the absence of AIDS).²⁸

In the next three years, AIDS will cause more deaths in African children than either malaria or measles, according to UNICEF.²⁹ Similar effects are being evidenced in other countries in Africa: It has been estimated that three-quarters of the 11 million people infected with HIV in sub-Saharan Africa have yet to develop AIDS.³⁰ Parts of Asia are similarly expected to be significantly affected by the spread of the epidemic in the next decade.

Recent work by FAO in East Africa has shown that AIDS mortality is not only a medical and demographic concern but also an agricultural and more specifically, a farming systems concern, particularly in terms of the effects of AlDS-related mortality and morbidity on farming systems and rural livelihoods, the coping mechanisms adopted by rural communities and the types of intervention that can mitigate the adverse impact of the epidemic. Given the particularly important gender dimension of the disease, special attention needs to be placed on the impact of HIV/AIDS on farm women and on female-headed households. A brief summary of the findings of these studies and their implications is presented below.

Case Study on the Impact of HIV/AIDS-related Mortality on Farming Systems and Rural Families in East Africa: ^d

(^d The case studies are drawn from two FAO studies: (a) "The Effects of HIV/AIDS on Agricultural Production Systems and Rural Livelihoods in East Africa": In this study, multidisciplinary national research teams carried out HIV/AIDS impact studies in different farming systems of Uganda, Tanzania and Zambia, with emphasis on qualitative data gathering methods derived from rapid and participatory rural appraisal (RRA, PRA); and (b) “The Socio-Economic Impact of HIV/AIDS on Rural Families in Uganda. with an Emphasis on Youth": This study, conducted in six villages of Kabarole, Gulu and Tororo districts, used PRA/RRA to: assess the socio-economic impact of the disease on the rural household (household economy, health, education, etc.); identify risk factors contributing to the spread of HIV/AIDS among youth; and conduct a Knowledge-Attitude and Practice survey on HIV/AIDS among the youth. The field work was carried out as part of an ongoing agricultural extension project.)

The primary impact of HIV/AIDS-related mortality on agricultural production systems is loss of labour. Two of three farming systems examined in one FAO study on "The Effects of HIV/AIDS on Agricultural Production Systems and Rural Livelihoods in East Africa" were found to suffer from serious labour shortage, which in one of the villages was primarily caused by AlDS-related mortality. Some of the effects of labour shortage on agricultural production observed in a community that had suffered over a prolonged time from AlDS-related mortality in Uganda included:

- Reduction in the acreage of land under cultivation Delay in farming operations such as tillage, planting and weeding Reduction in the ability to control crop pests

- Decline in crop yields

- Loss of soil fertility

- Shift from labour intensive crops (e. 9. banana) to less labour intensive crops (such as cassava and sweet potato)

- Shift from cash oriented production to subsistence production

- Reduction in the range of crops per household

- Decline in livestock production Decreased food security Loss of agricultural knowledge and management skills³¹

The decline in agricultural production and the loss of household income adversely affects other household functions such as the provision of education and health, according to an FAO study in Uganda on the "Socio-Economic Impact of HIV/AIDS on Rural Families" (see Box on next page). It also profoundly changes the division of labour by gender as well as by age. AIDS mortality increases the need for child labour. Girls tend to suffer most as they are the first to be taken out of school and care for younger siblings and the sick, tend the family farm and assume household tasks.

The burden of the impact of AlDS-related mortality disproportionately affects women and girl children. The FAO study on "The Socio-Economic Impact of HIV/AIDS on Rural Families" found more households headed by AIDS widows than by AIDS widowers. Widows with dependent children tend to become entrenched in poverty. They may lose access to land, labour, inputs, agricultural extension, credit, and other support services. HIVIAIDS stigmatization compounds their situation further, eventually severing assistance from the extended family, which is often their only safety net, and the community.³²

BOX 2 illustrates how AIDS mortality affects a farm household and how it impacts on food security, on the standard of living of a family (and particularly on its nutritional and health status). The working conceptual framework presented in Figure 2 and the other tables on the division of labour by gender and age show how the loss of one or more family members can have multiple repercussions on the farming system through quantity of work, type of work and skills mix, an end to remittances and a reallocation of resources. The loss of both parents to AIDS can also lead to children- and grandparent-headed households.

The FAO studies also found that it is becoming difficult to implement agricultural programmes as a result of Al DS mortality. Agricultural extension services have been hard hit by the disease. In Uganda's Rakai district, 20-50% of all working time of extension services is lost due to HIVIAIDS. Extension staff are frequently absent from work attending burials and caring for sick relatives. At the same time, a number of extensionists are falling sick and dying. In fact, agricultural extension workers are in themselves a "high risk" group in terms of susceptibility to HIV/AIDS, given the fact that they are mobile, come to contact with a large number of people and have not been targetted by HIV/AIDS awareness campaigns. The epidemic has also made it more difficult for extension staff to meet with farmers. If a meeting coincides with a funeral, it is rescheduled. In some communities in Rakai district in Uganda where there have been as many as 10-1 5 deaths in one month, extensionists are finding it increasingly difficult to operate.³³

Helen, a widow in her late 30s, has six children. She married when she was 13. Her husband died of AIDS in 1988. She also has AIDS, is bedridden and is severely undernourished. Helen lives with her daughter, who is 19, is unmarried with a child of two years and has had five years of schooling. Her daughter is the only one taking care of her. "Without her, I could not have managed. I would not be alive today," Helen said. None of her sons work. When her husband died, they dropped out of school because she could no longer afford to pay school fees. Now they cannot find jobs. One of her sons is living nearby, but only has a small cassava shamba which is not enough to feed his own family, let alone his mother and sister. When her husband was alive, the family grew potatoes and sorghum on their 3-acre plot. Some of the sorghum was sold to buy soap, salt, meat, fish and paraffin. These days, Helen only grows cassava and millet. The husband's family has not claimed the land-yet, but she fears this may happen at any time. When she feels strong she- works as a casual labourer to earn some cash. The daily wage of casual labourers is about 1,000 US$ (about US$: 1). She would: like to grow beans-to improve the family- diet but cannot- do so at present as her yields have declined drastically. The old: cassava shamba (field) is nearly depleted and the new one is too young to harvest Helen was. not able to harvest millet: this year because she suffered from severe body: pain and was too weak to work. The family does not grow enough food to live on. Her diet consists of boiled::-cassava or millet and- a few greens without sauce : (there- is: no money to buy oil with which to prepare -the: sauce). Helen's daughter tries to prepare two meals a day, but often they only have one. Helen had not eaten fruit for-over a month. The widow has received no moral or material support -from her late husband's family or from: the village. :No one ever: comes to see her. Attitudes toward: her and her family were very negative, she said. She does not want to ask: for help: from her husband's male relatives because she fears that their wives will suspect that she is sexually involved with them. When she is not bedridden, Helen-works from 5:00 am to 9:00 pm. This long workday exhausts her but she- cannot afford to rest because she and her daughter do not have Enough to live on. She described her situation as a vicious circle: she cannot grow enough food to feed herself and her family because she is too weak and hungry; yet she would need to eat properly in order to be strong enough to work in the shamba. Source: D. Topoozis, The Socio-Economic Impact of HIV/AIDS on Rural Families, FAO, 1994 p. 15.

(² The Impact of AIDS mortality on a Rural Household)

Some AIDS widows in Uganda pointed out that as soon as their husbands died, they had lost access to agricultural extension advice which they considered invaluable (and to which few women have access to). The death of their husbands, they explained, did not only amount to the loss of their labour and income, but also to the loss of their access to credit, extension and information, access to cooperatives or other farmers' organizations, and the loss of knowledge of certain cash crops and of specific farming practices. Given the added burden of AIDS stigmatization that these widows face, losing access to extension at a time when they needed was of paramount importance to them.

It is important to underscore that even though to date HIV/AIDS appears to have hit Africa more severely than other developing regions, and certainly more is known about the nature of HIV/AIDS impact in Africa, comparable findings are likely to be relevant for a number of Asian and Caribbean countries affected by the epidemic. In fact, AlDS-related mortality is not only an African concern. To give but one example from Asia, high HIV infection rates in the Northern Thai provinces necessitate an investigation of what the impact of mortality is likely to be on rice-based farming systems. In particular, insights into the impact of HIVIAIDS on the division of labour by gender and age and the repercussions on demographic patterns could provide the foundation for policies designed to mitigate the adverse effects of the epidemic on agricultural production, farming systems and small farmers.

Main issues and research needs

i) Need for gender-dissagregated data and gender-responsive methodologies in farming systems research, including data linking agriculture and population

It has been argued that farming systems research has not successfully integrated gender issues and has not provided gender-disaggregated data.³⁴ This is only partly true, however, given the fact that no country systematically collects farming systems data as such: farming systems research usually takes piece in light of specific interventions and contexts. However, the collection of dissagregated data by age and by sex on a project or programme basis, revealing intra-household gender roles, needs and constraints, could go a long way towards establishing more accurate labour profiles and throwing some light on linkages with population issues.

The conceptualization of the farm household as a production unit can be problematic and the fact that households can comprise two or more productive subsystems has not yet been fully acknowledged by policy-makers and development planners alike. This may be the result of socio-economic factors as well as gender blindness. For instance, in much of Africa, depending on the number o, wives within a household, only one productive system is controlled by the man; the others are usually controlled by the wife or co-wives. In other instances, as a result of migration (male or female), the remittances many families rely upon represent another productive sub-system.

Furthermore, gender roles have not been sufficiently understood when it comes to technology generation and transfer and the need to link demographic data to information on female or male economic activities in the context of farming systems has yet to be addressed. The section on labour bottlenecks and fertility/mortality has shown that there is a need to explore how changes in farming techniques may affect women's farm work both in terms of increasing or decreasing their workload as well as in terms of their ability to breast-feed.

ii) Differentiate between different groups of subsistence farmers by gender and by socio-economic status, ensuring that de facto and de jure female-headed households are included among policy, programme and project target groups.

There is a general tendency to assume that subsistence farmers are a homogeneous group with uniform interests, needs and constraints. However, upon closer look, subsistence farmers comprise a wide spectrum of sub-groups according to gender, farm size, annual income, etc. Thus, one can distinguish, for instance, female-headed households, which have less access to land, productive resources and support services, landless labourers, etc.

It is also commonly assumed that rural women are a homogeneous group. Yet, research has shown that new technologies can affect different strata of rural women in very different ways. In the Philippines, for instance, the introduction of wet-seeding/direct-seeding technology, using High-Yield Varieties (HYV) in rice cultivation, which reduces the demand for pulling/seeding and transplanting (tasks normally performed by women) has had opposite effects on women subsistence farmers and on landless women. The effect of HYV on women subsistence farmers has been to lower fertility rates and maternal and infant mortality while the effect on poor landless women has been increased fertility (with preference for sons because of the greater demand for male labourers) and infant mortality.³⁵

iii) Address the Interface Between Labour Bottlenecks and Fertility/Mortality

Findings on the role of reproduction on rural Kenyan women's time use provides valuable insights into the interface between heavy labour and pregnancy on the one hand, and the agricultural calendar on the other. These need to be taken into account when designing interventions, to ensure that women's workload is not increased and that their energy intake is adequate. Little is known, however, on how the pressures on women's time at certain points of the agricultural cycle may lead to curtailment of breast-feeding and the consequent demographic effects different types of agricultural innovation may have on these.

There is a need to further explore how changes in farming techniques may affect fertility as well as women's farm work in terms of their ability to breast-feed. More specifically, the following questions need to be addressed: To what degree does this occur? What farm tasks are the most strenuous and potentially dangerous to pregnant women? How can they be compensated for by new technologies? How can such new technologies be made acceptable and attractive to farmers?

One way to begin addressing some of these questions would be to categorize farming systems not only by physical characteristic (i.e. rain-fed small farms, swidden farming in uplands, irrigated fields, etc.), but to distinguish them on the basis of intra-household dynamics. This will necessitate supporting data on the type of technologies that would accompany each type of farming system as well as a prioritization of other sub-groups of farming systems.

iv) Assess the Gender-Differential impact of the Introduction of New Technologies, in view of its Potential Effect on Fertility/Mortality, including the benefits of different types of technical change in agriculture on the two sexes

The introduction of new technologies has tended to benefit men at the expense of women, by marginalizing women, by increasing their workload or by being inappropriate for their use. By identifying and understanding the role of women in farming systems, and by consulting them prior to the introduction of new technologies, it will be possible to facilitate the generation of technologies that women farmers can use. It would also be important to empirically assess how the introduction of new technologies at certain points of the agricultural cycle affect the labour inputs of women, particularly during pregnancy and lactation, and to determine whether it is likely to contribute to high fertility and mortality.

v) Quantify, Document and Take into Account the Economic Contributions of Children by Gender

Even though it is logistically and practically difficult to measure and document the economic contribution of children, there is a need to investigate whether and to what degree new technologies can indirectly affect fertility patterns and neutralize gains in fertility reduction by increasing girls' labour inputs, leading them to drop out of school. The quantification of children's labour contribution would be most effective if undertaken on the basis of person days per activity and season. This will allow policymakers and planners to make demographic- and gender-responsive decisions about the introduction of new technologies, the introduction of new crops, etc.

Key questions that need to be addressed include the following: could changes in farming systems (cropping patterns, new technologies, etc.) better balance the gender and age division of labour and therefore offset the growing need for Chile labour and the growing responsibilities of women and girl children in agricultural production? What are the implications of children's labour contributions to farming systems and the farm-household economy? What are some of the policy adjustments that can be made to avoid children being taken out of school to make up for labour needs? What are the implications of women's increasing labour burdens on fertility, and more particularly, on decision-making concerning family size?

Finally, given the significant but invisible contribution of children and youth to agricultural and rural development, efforts could be made to:

- investigate how gender-disaggregated data on children's contributions can be incorporated in censuses, surveys, etc.;

- analyze the interface between children's and youth's farm labour and demographic factors (especially fertility and migration); and

- identify appropriate methodological tools to facilitate the development of a more holistic perspective on the socio-economic contribution of children and youth in the farm household and in the community in order to enable them to benefit from their contributions.

vi) Address the Socio-Economic and Demographic Effects of HIV/AIDS Mortality in Policies and Programmes

The systemic effects of the socio-economic impact of HIV/AIDS at the household level, including rural livelihood systems and rural men and women themselves, need to be factored in population as well as in agricultural rural development programmed, taking into account the gender dimension of the epidemic. Thus, development projects in countries most affected by the HIV/AIDS epidemic (i.e. oilseed project in Uganda, agricultural extension project in Zambia, cash crop project in Zimbabwe, etc.) need to consider the demographic and gender dimensions of the project, and in particular the potential effect of HIV/AIDS (on mortality and morbidity) on men and women farmers' livelihood systems. Responses to the epidemic also need to consider the linkages between gender, household food security and social support functions and should be people-centred.

As little is known about the impact of HIV/AIDS on agricultural production systems in Asia and related linkages with population issues, research needs to be conducted on a pilot basis. For example, in Thailand, research on the impact of AIDS-related mortality on rice-based farming systems, and in particular, on the division of labour by gender and age and the repercussions on of HIV/AIDS on demographic patterns could help guide policies designed to mitigate the adverse effects of the epidemic on agricultural production, farming systems and small farmers.

Endnotes

1. Christine Oppong, Introduction to Aderanti Adepoju and C. Oppong, Gender. Work and Population, Geneva: ILO, 1994, p. 10.

2. ibid., pp. 10-11.

3. A. Evans, "Statistics," in Ostergaard, L. (ed). Gender and Development: A Practical Guide, London: Routledge, 1992, p.12, cited in Caroline Moser, "Evaluating Gender Impacts," Conference Paper on Evaluation and Development, World Bank, December 5 and 6, 1994, p. 14.

4. K. Ewusi. "Women in Occupation in Ghana," paper presented at the Seminar on Women and Development held by the Council on Women, Legon, Ghana, 1978, cited in Constantina Safilios-Rothschild, "Agricultural Policies and Women Producers," in Gender, Work and Population in Sub-Saharan Africa, op. cit., p. 57.

5. Oppong, Introduction to Gender, Work and Population, op. cit., p. 5.

6. A.S. Carloni, "Women in Development: AlD's Experience 1973-1985," vol. 1, Synthesis Paper, Programme Evaluation Report No. 18, cited in Women, Food Systems and Agriculture, Women in Agricultural Development Paper, FAO, 1990.

7. Christine Oppong, Introduction to Gender, Work and Population in Sub-Saharan Africa, op. cit., p. 3.

8. The World's Women: Trends and Statistics 1970-1990, United Nations, New York, 1991.

9. Oppong, Introduction to Gender. Work and Population in Sub-Saharan Africa, op. cit., p. 2.

10. Ibid.

11. Oppong, Introduction to Gender. Work and Population in Sub-Saharan Africa, op. cit., p. 3.

12. Cited in Introduction, Gender. Work and Population. op. cit., p. 3.

13. Ibid. p. 3.

14. Ingrid Palmer, "Seasonal Dimension of Women's Roles," in R. Chambers, R. Longhurst and A. Pacey (ads.), Seasonal Dimensions to Rural Poverty, London: Frances Pinter Ltd,1981, cited in Oppong, Introduction, op. cit., p. 14.

15. FAO, Manual on Project Planning and Formulation of Integrated Population, Women in Rural/Agricultural Development Projects, draft mimeo, 1990, cited in Zoran Roca, "Women and Population in Agricultural and Rural Development in Sub-Saharan Africa," Women in Agricultural Development paper, No. 5, p. 19.

16. Michael Baksh, Charlotte Neumann, Michael Paolisso, Richard Trostle and A. Jansen, "The Influence of Reproductive Status on Rural Kenyan Women's Time Use," Social Science Medicine, Vol. 39, No. 3, 1994, pp. 354.

17. ibid.

18. Ingrid Palmer, Gender and Population in the Adjustment of African Economies: Planning for Change, Geneva: ILO, 1991, p. 33.

19. Christine Oppong, "Population, Development and Gender Issues in Namibia," ILO, 1994, p. 120.

20. Aderanti Adepoju, "The Demographic Profile: Sustained high Mortality and Fertility and Migration for Employment," in Gender, Work and Population in Sub-Saharan Africa, op. cit., p. 24.

21. Christine Oppong, "African Mothers, Workers and Wives: Inequality and Segregation," Working Paper No. 2, Population, Human Resources and Development Planning in Sub-Saharan Africa, Geneva, ILO, cited in Gender. Work and Population, op. cit., p. 24.

22. Palmer, Gander and Population in the Adjustment of African Economies, op. cit., p. 32.

23. Brenda McSweeney, The Negative Impact of Development on Women Reconsidered: A Case study of the Women's Education Project in Upper Volta," unpublished Ph.D. thesis, Maryland, Fletcher School of Law and Diplomacy, cited in Palmer, op. cit., p. 32.

24. Zoran Roca, "Women and Population in Agricultural and Rural Development in Sub-Saharan Africa,' FAO Women in Agricultural Development Paper, No. 5, 1991, p. 12.

25. FAO/UNFPA Case study on Population, Status of Women in Rural Development in Lesotho, Sierra Leone and Zimbabwe, 1989, in ibid., p. 19.

26. D. T. Jamison and L.J. Lau, Farmer Education and Farmer Efficiency. Baltimore: John Hopkins University Press, 1982, cited in Hans P. Binswanger and Klaus Deininger, "Towards a Political Economy of Agriculture and Agrarian Relations," Draft paper, World Bank, March 1995, p. 19.

27. Christine Oppong, "Population, Development and Gender issues in Namibia," ILO, 1994, p. 36.

28. See Selected Pages from "World Population Prospects: the 1994 Revision, forthcoming in 1995 from the Population Division, Department for Economic and Social Information and Policy Analysis, UN Secretariat, p. 137.

29. Cited in World Bank, AIDS Prevention and Mitigation in Sub-Saharan Africa: An Updated World Bank Strategy, June 1995, p. xv.

30. World Bank, 1995, ibid. p. iii.

31. Tony Barnett, "The Effects of HIV/AIDS on Farming Systems and Rural Livelihoods in Uganda, Tanzania and Zambia," FAO, 1994 and The Effects of HIV/AIDS on Farming Systems in Eastern Africa, 1995, Chapter 3.

32. Daphne Topoozis, The Socio-Economic Impact of HIV/AIDS on Rural Families with an Emphasis on Youth, FAO, 1994, pp. 17-20.

33. FAO, The Effects of HIV/AIDS on Farming Systems in Eastern Africa, AGSP, 1995, pp. 73-74.

34. Constantina Safilios-Rothschiid, "Agricultural Policies and Women Farmers," in Gender, Work and Population in Sub-Saharan Africa op. cit, p. 55.

35. Zoran Roca, "Women, Population and Environment in Agricultural and Rural Development: Policy Challenges and Responses," FAO, 1994, p. 10.