FAO Investment Centre
Socio-economic and Production Systems Studies

  Le Centre d'investissement de la FAO
Études des systèmes socio-économiques et productifs

  El Centro de Inversiones de la FAO
Estudios de los sistemas socioeconómicos y productivos

MALAWI

SMALLHOLDER IRRIGATION SUBSECTOR PROGRAMME

Socio-economic and Production System Study of Wetland Use

Objective of the Study

The main objective of the study carried out by the FAO Investment Centre was to increase understanding of the characteristics, needs, and priorities of communities who have access to wetlands, with a view to determining ways in which investments under the programme can help small farmer households improve their situation by making better use of the wetlands.

Study Area

For the purpose of this study, it was decided to include any wetland being utilized (or capable of being utilized) for farming purposes. The sample areas were selected on the basis of coverage of the various agro-climatic subregions, types of wetlands, and concentrations of poverty.

Areas with a high concentration of poverty were identified on the basis of data and maps prepared by the Famine Early Warning System Project (FEWS), considering the main factors of food security and degree of poverty¹. Other factors considered were frequency of female headed households and degree of stunting and wasting of children under 5 years of age. Final selection of the areas to be studied was preceded by a rapid reconnaissance visit covering about 15 sites in two ADDs.

Identification of the general area to be studied was followed by the identification of specific sites. This was done at local Rural Development Programme (RDP) with the assistance of the staff of the Ministry of Agriculture and Livestock Development (MOALD) and of the Irrigation Department (DOI).

Study Methodology

The study was conducted by a joint FAO/IC and Government of Malawi team utilizing Participatory Rural Appraisal (PRA) methods. In general, one day was spent in each village. Information in each community was gathered by way of the following:

• Transect walks, undertaken with knowledgeable members of the community, in order to analyze physical features (hydrology, soils, vegetation), socio-economic patterns (land use, land tenure, traditional utilization of wetlands) and constraints. These transects were further corroborated by village mapping, undertaken by members of the community
• Focused group interview in order to gather information on timelines of wetlands utilization, farming systems analysis, population dynamics, socio-economy of the village, village level institutions, local initiatives, village food security.
• Gathering information on farming calendars, farmers' own perceptions of the main characteristics of the different socio-economic strata in their community. constraints and priorities of the village community as a whole, and any other relevant data.

In addition, individual households, selected as far as possible from different strata , as identified by the community, were interviewed.

In total, 26 villages, covering several agro-climatic sub-regions and six major types of wetlands, located in nine RDPs in the three main regions were the object of the study (see Map of Malawi). Eighty-three households were interviewed in these villages; twenty-four (or 29%) of the households interviewed were headed by women.

THE SOCIO-ECONOMIC CONTEXT

Background to Existing Agrarian Systems

From Independence to the start of the 1990s, the agricultural sector was developed along a dual structure of estate and smallholder farming. These two sub-sectors were largely differentiated on the basis of size of holdings and on legal and institutional rules regulating land tenure, crop production, their marketing and input supply and their pricing.

The estate sub-sector is based primarily on leasehold tenure, though there are about 52,000 ha under freehold. In 1993 it was estimated that there were about 23,000 estates, occupying about 1.2 million hectares, a dramatic change from the 1970s when there were only about 229 estates occupying 255,500 hectares. This change occurred mostly in the 1980s and the early 1990s. Most of this expansion in estate land occurred at the expense of customary land taken away from the smallholders. Previous agricultural policies of forbidding smallholders from growing and marketing burley tobacco had a marked impact on the transferring of communal lands to estates.

The smallholder sub-sector is based on customary tenancy. It is subsistence-oriented, providing about 85% of domestic food production, and is the backbone of Malawi economy. Though smallholders rear all classes of livestock, smallholder farming is dominated by crops, in particular maize, which, in pure stand, occupies about 60% of the area cropped. This is mostly due to the fact that past agricultural policies placed high emphasis on increasing maize production in all agro-climatic subregions. Increased population pressure, combined with the expansion of the estate sector, has had a negative impact on holding sizes: in 1968/69, 71% of the smallholders cultivated 2 hectares or more, by 1994, only 15% were cultivating more than 1.15 ha. Because of land scarcity, more marginal lands and steep hillsides were converted into arable land, thus causing very severe environmental degradation and wetlands started to be utilized.

Traditionally, wetlands were also considered marginal lands. The increased population pressure and the ever decreasing rainfed area for cultivation has gradually, over the years, pressured farmers to turn to wetlands cultivation. The drought period, which in the early 1990s beleaguered the whole of the Southern Africa region, was yet another factor which made wetlands utilization more valuable. Lastly, as a result of removal of subsidies and of the devaluation of the Malawi Kwacha, fertilizer price trebled between 1994/95 and 1995/96 growing seasons. This was yet another factor which induced a larger number of farmers to turn to wetlands as most of their soils are, in general, more fertile. Although there are no recent statistics documenting this increase in wetland utilization, the timelines of several villages studied clearly reveal this trend.

POPULATION DYNAMICS

Demography

Malawi's total population has been estimated at about 11 million in 1994, nearly three times the 1964 level (4 million). Ninety percent of the population is rural and in 1993 there were about 1.6 million rural households. Average rural family size was about 4.9; households in the Northern Region tend to be larger (See Figure 4). Average household size of the study sample was 5.7, but family size was found to vary according to socio-economic strata.

In 1993, 32% of the rural households were headed by women; a relatively higher proportion of female headed households is found in the Southern and Central Region (see Figure 5) In the sample villages studied, about 29% of the households were headed by women. The composition of the rural population indicates a ratio of 93 males per 100 females. Almost half of the population (47%) are under 15. Dependency ratio is high and increasing, In 1977, there were 97 dependants for every 100 adults of working age, whereas in 1992 this ratio climbed to 106.

Although fertility rates have decreased in the last ten years (from 7.6 in 1982 to 6.7 in 1992), they are still quite high². This, coupled with a steady decline in mortality, has induced a high growth rate, which is estimated at 3.3% per annum³. Life expectancy at birth is estimated to have risen from 37 years in the early 1950s to 44 in 1992, still one of the lowest in the world. This is due to precarious living conditions, limited access to social services (sanitation and health care) and illiteracy. In spite of the decline in child mortality rate experienced in the last decade (from 126.1 in 1983-87 to 114.9 in 1988-92), infant mortality remains high (137.5 to 134.3 for the same period)⁴. Maternal mortality rate is estimated at 620 maternal deaths per 100,000 births. The 225,000 estimated cases of AIDS put a heavy burden on Malawi. By the year 2000, an estimated 2 million people will be infected with HIV and about 35,000 children will be orphaned⁵. The spread of the virus is increasing both child and adult mortality.

Average population density in 1987 was estimated at 87 people per km², and it has increased to about 106 persons per km² during the past 9 years. The population is unevenly distributed across the country. The Southern Region, comprising 39% of the total arable land and one third of the total land area, supported nearly 50% of the population with an average density of 125 persons per km². The Northern Region, comprising 29% of all land and 20% of the arable land, had only 11% of the total population and a density of 34 persons per km². The Central Region, which contains 41% of the arable land, had 39% of the population and a density of 87 persons per km². District densities ranged from a high of 293 persons per km² in Blantyre District to a low of 20 persons per km² in Rumphi District. Figure 6 depicts the population density in 1987; current figures are higher.

Ethnicity

In the Northern Region, the main ethnic groups are the Ngoni, Tonga, Ngonde and Tumbuka, which are predominantly patrilineal. Through the institution of brideworth (dowry), the man takes his wife to his village. Property inheritance is through patrilineage and decision making is predominantly the men's affair. Married women in patrilineal societies do not normally make decisions considered important by society; these include decisions on farming and livestock.

By contrast, in the Southern and Central Region, main ethnic groups are the Chewa, Yao, Nyanja and Ngoni⁶, which are mostly matrilineal, the only exception being the groups in Chikwawa and Nsanje districts. In matrilineal societies marriage is uxorilocal. Inheritance of property is through the female line. However, men are still the main decision makers. Moreover, the institution of Chitengwa, under which the husband can seek and be granted permission to take his wife and live in his own village, still exists. Recent studies have indicated that this practice may be on the increase as a result of factors such as rising population, and resultant land pressure. Traditionally the needs of the man's children would be met by his wife's brother but in recent times this too is changing. The man is taking greater responsibility.

Fifty-seven percent of the villages studied were predominantly patrilineal, the remaining 43% matrilineal.

Settlement Patterns

The Southern and Central regions have compact, clustered settlements in contrast to the North's dispersed ones. In the former, relatively large villages of clustered houses are located separately from people's holdings (dryland and wetland). Villages are composed mostly of people without common descent. Many are permanent settlers, immigrant by marriage, or simply land seekers, or generations following them. In the case of the Central Region, and the section of the Northern Region inhabited by the Ngoni, the villages are organized in terms of limanas, which are sub-divisions headed by a man appointed by the village headman. The heads of the limanas facilitate the administration and mobilization of the village. The number of limanas per village depends on the size of the village.

Most villages in the Northern region are relatively small with dispersed 'family' homesteads or compounds. Each family compound is composed of married brothers who may be living with their parents on land previously occupied by their ancestors. Their dryland holdings are normally located around the compound.

Holding Sizes

Smallholder cultivation in Malawi is bedeviled by severe problems of decreasing holding size, falling yields, and increasing use of marginal land. These problems are especially prevalent in the Southern and Central Regions, where a combination of high population density and allocation of land to the estate sector have resulted in increasing land pressure.

Nationwide, about 78% of the households had holdings of less than 1 ha (48.4% of less than 0.5 ha, and 29.3% between 0.5 and 0.99 ha), 12.1% between 1.0 ha and 2.0 ha and the remainder, 2 ha and above. The majority of very small holdings is concentrated in the Southern Region and on the edge of Lake Malawi.

Average holding sizes of the sample households interviewed was found to be 0.9 ha, with a standard deviation of 0.56 ha, size of holdings varies with socio-economic status. In addition to having smaller holding sizes (0.6 ha), very poor households do not cultivate all their holdings, especially the wetland ones. Over three quarters of the sample households were cultivating between 40% to 60% of their wetland holdings, the majority of the poor households were cultivating around 75% of their wetland holdings, whereas the less poor tended to cultivate all their holdings. Main constraints were: labour (ganyu work availability coincides with land preparation activities in their dimbas), capital to buy seeds and/or non availability of seeds in nearby outlets.

LAND TENURE

Land in the smallholder sector is held under customary land tenure system⁷. All land covered in this study falls under this system of tenure. in villages, land is generally subdivided into three categories:

• communal land, such as grazing land, village forest reserve (for the collection of firewood and non wood products), roads and communal meeting places;
• reserve land, i.e. land which has not yet been allocated to anybody and which is entrusted to the village headman;
• land which is subject to the exclusive use of individuals or nuclear families.

In all the villages studied in the Southern and Central Regions, village reserve land no longer exists, the same holds true for most of the forest reserve land. Population pressure and the massive influx of Mozambican refugees in the 1980s and early 1990s are the major causes of the dwindling of forests. In the North, on the other hand, several villages still hold some reserve land.

In the Northern Region, the land tenure systems are predominantly patrilineal, while in the South and Central Regions, matrilineal systems are more common. In either case, the supreme authority is the traditional chief, who, generally, transmits his power to the village headman. The customary land tenure system allows an individual only access to, and not ownership, of land. This, however, does not put security of tenure into question as it has sometimes been assumed. In fact, once land has been allocated, it cannot be easily taken away from an individual even by the village headman himself⁸. In principle, land cannot be bought or sold but it can be sub-allocated or leased to some other person, with consent from the elders⁹. In case of sub-allocation or lease to a non-resident of the village, prior consent must be given by the chief or the village headmen. Rights to the land are transferred from one generation to another in accordance with the prevailing norms and kinship systems.

In patrilineal societies, each adult male is allocated land by the village headman. Instead of allocating land to individuals, the land may be allocated to patrilineages. In this case, the head of each patrilineage assumes the responsibility of further allocating land to individual male members of his patrilineage. The head of the patrilineage also assumes the responsibility of settling disputes and of controlling land use. Appeals in disputes and in other instances, for example when the land allocated is insufficient, revert to the village headman. From the latter, appeals go to the traditional chief. The land that is allocated and is in actual use or occupation cannot be reallocated to someone else unless it is first forfeited or surrendered. In patrilineal society the transfer and sub-division of land is from a father to his male children. A married woman may, by virtue of her marriage, be allocated some land by the authorities vested with the power to do so. Usually, however, the woman obtains her rights to land through her husband. In case of death of the latter, she can exercise her rights through her male children, if she decides to remain in the village of the deceased husband. In case of divorce she will have to return to her original home/village and ask for land from her family/lineage/village headmen.

In matrilineal societies, land is transferred from a mother to her female children through the heads of the matrilineage (who, generally, are the senior maternal uncles) and sub-divided accordingly. When a man goes to stay in his wife's village, his rights to land are held through the wife. Generally, the land law, rights and interests a person may have in land in matrilineal societies are similar to those in patrilineal ones. Any person to whom the land is allocated may lease his/her land. In the case of a husband residing uxorilocally, he must first seek the consent of his wife's senior maternal uncle or his representative. A divorced or deserted woman must seek the consent of her uncle.

Increasing scarcity of virgin land for cultivation is leading to increasing fragmentation of land where population pressure is high. Also, females in matrilineal society have better access to land than those in patrilineal society.

Where land is plentiful, it is allocated freely to a village member or to a new settler upon him being accepted. However with increasing scarcity of land, it is known that land seekers sometimes have to pay some money to the village headman in order to acquire land. Such payments are usually rationalized in terms of being a "token of appreciation". Obviously, however, the extent of this practice is difficult to ascertain.

A more common and ascertainable practice is acquisition of land through renting, on a growing season basis. In this case land may be acquired by paying an agreed sum of money to the owner, or alternatively an agreed amount of produce may be used for payment after harvest.. The cost of renting land is usually not fixed community-wide and the extent of this practice varies with the degree of land pressure. For example about half of the population were renting land in some dambos in the South (like in Matimati village, Chikwawa RDP) while very few were doing so in the North. Renting of land does not guarantee security of tenure since the contract has to be renewed on a yearly basis.

In some cases, mainly in some parts of the Northern region where large pieces of land, especially dambo land, still lie uncultivated, land may be borrowed. Borrowed land can be retrieved by the owner after any growing season. A unique case of land acquisition is the renting of land across the national borders: this was found to have occurred during the drought years in villages in the South bordering Mozambique.

Disregard of prevailing tree tenure (and land tenure) patterns have contributed, in Malawi, as elsewhere in Africa, to difficulties experienced in promoting reforestation, soil conservation, and watershed management. There are two ways of gaining greater security of tenure on land under customary laws in Malawi, that is through planting trees and through cutting trees.

Most customary systems accord tenure of trees to the person who plants them. Since the trees belong to someone, the potential uses of the land on which they are located are constrained, and the tree planter most often ends up as "land-owner" as well. Knowledge of this, interestingly, may lead communities to resist individuals' efforts to plant trees. There are cases, elsewhere in Africa, where others with rights to land have uprooted newly planted seedlings to prevent loss of those rights.

Malawi's indigenous natural forests are part of the customary land and are open to the use of members of the community. Chiefs and village headmen have the responsibility to allocate land, and if there is no other land available, they grant a person access to a piece of natural forest. The land thus changes status from communal access to an individual one; however, as long as the trees are there, the community considers the trees to be fair game. Naturally the new landholder generally cuts the trees as quickly as possible. Thus, `the tragedy of the commons" is not averted by individualization.

Land tenure in wetlands in general, and in dambos in particular, was different depending on the areas visited. Main patterns were as follows:

• in instances where land was scarce, all land had been allocated either through the lineage or the household: this was the case in most of the villages in the Southern and Central Regions;
• when new wetland was available, either because of the cyclical drying of what used to be marshland (such as the Elephant Marsh in Chikwawa, Southern Region) or as a result of larger portions of wetland being available due to the higher rainfall experienced this year, the new wetland had been allocated to both residents of the village, to whom traditionally the wetlands belonged, and to residents of other nearby villages who had no or little access to wetlands;
• access to wetland not previously cultivated - due to labour and/or capital constraints experienced by local residents - was given, either on a permanent or temporary basis, to outsiders who often came from different districts. This was found to be the case in several villages in the Northern Region, especially those near large estates in Nkhata Bay.

HOUSEHOLD FOOD SECURITY

Preliminary results of a Food Security and Nutrition Survey carried out in 1994 indicated that in Malawi, 7 out of 10 households run out of food in the pre-harvest season¹⁰ . Figure 8 shows the distribution of smallholder households, by district, who ran out of food during the pre-harvest season in 1994. Clearly, food insecurity is pervasive in Malawi. In 1994 there were ten Districts where over 40% of the smallholders ran out of food in the pre-harvest season. It is likely that nation-wide, the situation has worsened on account of the trebling of fertilizer prices.

The Table below indicates the food sufficiency status of the households interviewed. Clearly the overwhelming majority of the households are food insecure. In fact, only 11% of all the households interviewed were food secure all the time and an additional 7% are food secure in good years. Moreover, 14% of the households in good years and 44% in bad years are self-sufficient for only six months or less.

Household Food Sufficiency

DEPENDENCY ON WETLANDS

A typology of the villages studied was established on the basis of their dependency on wetlands. The ensuing distribution was quite even, with a slightly higher proportion of villages with an average dependency on wetlands. The Table below summarizes the main features of each type.

Dependency on Wetlands

The above typology demonstrates that the higher the degree of food insecurity and poverty, the higher the degree of dependency on wetlands. In addition, high dependency on wetlands was often found to be associated with a high degree of motivation to improve the status of their wetlands.

HEALTH AND NUTRITION

Child malnutrition is widespread in Malawi: stunting, an indicator of long-term malnutrition, affects nearly one in every two children under five; moreover, one in four children is underweight. Poor diets, short birth intervals and inappropriate feeding habits are the major determinants of malnutrition¹¹.

The most common health problems in Malawi are malaria and pneumonia, which concern everybody, and malnutrition, which affects mostly children and pregnant women. Other health problems include respiratory diseases, abdominal/gastrointestinal diseases, eyes and skin conditions. Amongst the parasitic diseases which have been steadily increasing are malaria and schistosomiasis. Although it is among the top twenty diseases, the latter is grossly underestimated, due to lack of appropriate laboratory equipment.

Schistosomiasis is highly endemic, particularly among children and women. Both groups are particularly at risk because they come in contact with contaminated water. According to some surveys¹², confirmed by the study, between 40% to 50% of the population have the infection and about 80% are at risk. Reportedly, the type found in the wetlands, Schistosome haematobium, affects up to 80% of the population in some areas¹³.

INFRASTRUCTURE

Rural Water Supply

According to the 1994 Food Security and Nutrition Survey cited above, nearly 60% of rural households in Malawi draw drinking water from unprotected sources (open wells, lakes/dams, rivers, streams)¹⁴. Results from the sample villages studied are not very dissimilar as 61% of them were drawing their water from unprotected sources. In addition, the quality of the water was judged very poor or poor by 35% and 33% of the cases respectively. Only 32% of the villages studied found the water quality good; several villages in the South were served by piped water which was quantitatively inadequate for most of the year and dried out completely in the latter part of the dry season. This meant that they then had to draw water from their dambo wells.

Accessibility

Roughly 60% of the villages visited were in close proximity to a feeder road (less than 5 km), the rest were much further away (between 8 to 15 km). The status of the feeder roads was, however, quite poor: only one third of them were in fair condition, 40% were in poor condition, the rest in very poor conditions. Most bridges were in a dangerous condition and nearly unusable, and many of them are regularly washed away during the rainy season, rendering the villages inaccessible for several months. Nearly all village access roads were in a deplorable state. This makes it difficult for farmers to access inputs and markets, and to reach other facilities such as health clinics.

Health and School Facilities

The survey cited above found that 54% of Malawi smallholders had difficult access to health care facilities and 28% to an under five clinic¹⁵. Results from the sample villages studied indicate that the average distance to a health facility was 6 km and that 73% of the villages studied had difficult access to them.

The survey also indicated that 10% of the people had difficult access to a primary school, whereas, according to the findings of this study, average distance to a primary school was 1,5 km. On the other hand, average distance to a secondary school was found to be 15 km.

Market/Trading Centre

Marketing of farm produce, especially vegetables, requires proximity to nearby markets. Yet, average distance to market or trading centre for the sample villages was found to be around 6 km; more than half of the villages studied had difficult access to markets, e.g. a market or a trading centre was more than 5 km away. This, combined with the poor road conditions, is a severe constraint to marketing.

NON-FARM EMPLOYMENT

Opportunities for off-farm employment are limited. Less poor households are often involved in trade, or have a shop. Those with skills work as artisans, for example as carpenters or tinsmiths. Other households are able to earn money from activities such as charcoal making, cutting thatching grass, or fishing. Women earn cash from brewing beer. However for poorer households, the main and often the only source of non-farm income is ganyu work, paid for either in cash or in food. Families did not report much difficulty in finding work, though they sometimes travel to neighbouring areas to get it. Near Mozambique and Zambia many people cross the border to work on the farms there. The peak times for ganyu employment are for planting and weeding of dryland crops and for land preparation for wetland crops. Women are employed for threshing and winnowing, particularly on rice and sunflower.

Although most payment of ganyu is in kind, it is worth about K 7 (US$ 0.46) per day for men and K 4-5 per day for women, for about six hours of work from sunrise to noon, but it is possible to achieve higher daily earnings by working longer hours at times of peak demand. In the Nkhata Bay area, very few of the indigenous people do ganyu work; it is mainly done by migrants from the plateau.

WOMEN'S ROLE

Malawian women play a key role as producers. According to the 1987 census, 70% of all farm work in the smallholder sub-sector was carried out by women. Usually women share with men most farming related tasks. Clearing of land for wetland cultivation involving digging out the roots of reeds or tall grasses, or turning the sods on land with shorter grass cover is usually done by men, but the heavy work of digging over the land for the dry season crops is often shared. In addition , women generally carry out harvesting of most crops, threshing and winnowing, transport, processing and most of the marketing. In the very poor and poor households, where men are heavily involved in ganyu work, most of the work of looking after the homestead and farming falls on women, although women do ganyu work as well. This, combined with arduous household tasks, child bearing and raising, collection of water and fuelwood and engagement in other productive activities, exerts a severe strain on their health, time and energy.

In addition, more women than men contribute to the incomes of the poorest households. Female household heads are the only income earners; in joint families, women's income contributes substantially to the total household income (nearly half). The socialization of women through the family, local traditional institutions and rites contribute women's home-bound role, division of labour and responsibilities, respect and obedience for male authorities, village elders and leaders. Cultural factors influence intra-household dynamics, decision-making in the household, the allocation of resources and the distribution of food.

Yet, investment in women's development has received a very small share of resources and attention in terms of training and skills development, of maternal and child health care and of education in general. The urgency of enhancing their economic productivity is a strategic choice to be made in order to improve theirs and their families living standards. It is a documented fact that increases in women's income impact positively on child and family nutritional status.

SOCIO-ECONOMIC DIFFERENTIATION OF HOUSEHOLDS

Following the categorization of the farmers themselves, the study team defined three main categories of smallholders, based on household resources (land and assets), access to employment and involvement in off-farm activities (which are very limited in rural Malawi). The three categories are very poor, poor and less poor. These categories have to be looked upon in the local context since, in general, with very few exceptions, poverty is widespread. The following table highlights the main characteristics of each category and their distribution in the sample of households interviewed.

The Table below indicates that the less poor are adopters of technology and initiatives. They use their dryland and wetland gardens and, to a lesser extent, their livestock as productive economic assets. The poor and about half of the very poor are mostly subsistence farmers; they try to imitate the initiatives of the less poor but their attempts are often frustrated by lack of resources. The recent price hike of inputs, combined with the stagnating (maize) or decreasing price of their main cash crops (sunflower and paddy), have adversely affected them the most, thus considerably lowering their living standards. They aspire to improve their status and get promoted to the upper category and they are willing to learn.

At least half of the very poor are less likely to respond to initiatives as they have labour constraints, since they have to seek ganyu work in order to survive, and are often bedeviled by chronic sickness. These households are locked in a vicious circle of poverty and have very limited capacity to invest resources into activities that have potential to let them escape it.

Household Typology and Characteristics

Household Composition and Labour Force

Although the average number of household members in the areas studied was 5.7, the corresponding averages were 4.7, 5.9 and 7.13 for the very poor, poor and less poor respectively. Similarly, labour availability increased with the strata, from 2.3, 2.5 and 2.9 respectively (see Text Figure 5.1 below).

The less poor households, in spite of their comparatively larger family size, have only slightly more family labour at their disposal; This is mostly due to the fact that they send all their children to school. However, they compensate for seasonal labour shortages by hiring ganyu and/or permanent labour¹⁶. Most of the members of the other two strata tend to send only young children to school and only for a couple of years (drop-out rates and school absenteeism in these categories may reach between 70% to 100%). Generally children start to work in the field at very early age (at about 8), doing light tasks; by the time they are 11-12 years they assist in all tasks, including land preparation and watering dimbas.

Food Stocks

At the time of the study (June/July), harvesting of staples had taken place in March/ April in the South. It had just been completed and/or was undertaken in the Central and Northern regions; in the cassava-based system. Cassava is harvested as needed. The team estimated, with the people being interviewed, the amount of staple food they had in stock.¹⁷. The result of this exercise, by categories of households, is shown in Text Figure 5.3.

In other words, the average staple food in stock per adult equivalent was about 84 kg for the very poor category, 184 kg for the poor category and 237 kg for the less poor¹⁸. This is a further indication of the poor food security situation, especially in view of the good rains of the past growing season.

Food Self Sufficiency and Number of Meals per Day

As mentioned before, average number of months of self-sufficiency for the whole sample was just over 9 months in good years and slightly less than 7 months in bad years. Comparable figures for the three strata appear in Text Figure 5.4.

Average number of meals for the very poor households were two after harvest, dropping to one in the pre-harvest period; these were complemented by wild fruits and some bananas. Poor households had an average of two meals per day in the post-harvest period and to one in the pre-harvest season.

Assets

Livestock

Reportedly, livestock numbers in the areas studied have decreased substantially as a result of high incidence of pests and diseases, population pressure, lack of adequate veterinary services and thefts (for cattle). Percentage of households having livestock and the average number for those owning them is indicated in Text Figure 5.5. Nearly three quarters of poor household have poultry and about one third have goats or pigs; about one in ten has cattle, as compared to less than half of the very poor having chicken and about one in ten owning goats or pigs. In spite of the fact that over three quarters of the less poor own goats , still only about half of them own cattle.

Trees

Trees are highly valued by rural households. Fruit trees play an important role in their diet, and provide some income, while other trees¹⁹ are an invaluable source of timber in a country where degradation of forests is so advanced. Text Figure 5.6 indicates tree ownership by type and numbers by the different strata. Over three quarters of both the poor and less poor category own fruit trees, though the average number of trees owned by the former is three quarters less. In the case of other trees, the situation is quite different. Only 3,5 out of 10 poor households own commercially valuable trees, as compared to 7.5 out of ten less poor households. Only few very poor households, less than one out of ten, own valuable trees.

Farm Implements and Other Assets

Other assets were few in number and concerned only a small proportion of households. Very poor households had none except one or two hoes and, rarely, a watering can. Poor households also had only hoes as farm implements but about half of them had watering cans, nearly 25% of them owned a bicycle or a radio. Just over a half of the less poor ones had a bicycle, 40% a radio; all had about two watering cans, 45% had a sprayer, 20% had an oxcart and 35% had a plough and a ridger (these latter implements were found only in the Northern region).

Input Use and Crop Diversification

Hybrid maize for dryland fields was used by one in five very poor households, one in three poor households and by all less poor ones. These latter were also using hybrid maize in their wetland fields while the proportion dropped to 20% for the very poor and to 50% for the poor households.

According to the people interviewed, fertilizer use had dropped considerably during this past growing season as a result of the price hike: amongst the very poor and poor households fertilizer application, reportedly, were halved or more than halved; in addition, the quantities applied were far less than those used previously. Only about 20% of the very poor households and 43% of the poor households had applied any fertilizer on their dryland crops in the 1995/96 growing season; on the other hand, nearly all the less poor ones had applied fertilizer on the dryland crops (mostly on maize and tobacco). Apart from small amounts of Actellic used for storage, the only pesticide applied on dryland crops was by a handful of farmers who were growing tomatoes during the rainy season.

With regard to wetland crops, fertilizer was applied in very small quantities. About 18% of the very poor and 30% of the poor farmers were applying small amounts of fertilizer, especially on the nurseries; three quarters of the less poor were applying fertilizer in both nurseries and on the fields. Although pests and diseases appeared to be a big constraint to most farmers, few were applying pesticides, being 15%, 17% and 33% respectively in the three different socio-economic groups.

Another survival strategy of most farmers is to try to diversify the types of crops they grow, provided they have the resources to do so. Average number of dryland crops for the sample households was 3 for the very poor households, 4 for the poor households and 5 for the less poor. Comparable figures for the wetland were 4, 5 and 6 respectively.

Household Income and Expenditures Levels and Patterns

Average income levels for the whole sample was about K 3,350, or a per capita income of K 670²⁰.Average income for the very poor was K 904 (or 226 per capita) , K 2,640 (or 528 per capita) for the poor and K 8,680 (or 1450 per capita) for the less poor (See Text Figure 5.7). However, since most of the ganyu work is paid in kind, the figure for the very poor and poor households may have not captured this income in kind.

The figures seem to point out to very large disparities; the findings of this study are confirmed by past nation-wide surveys, which indicate a highly skewed income distribution.

Sources of Income

Not surprisingly, given the nature of the sample, the main source of income for the households interviewed was the crops grown in the wetlands, which accounted for nearly 34% , followed by the sale of dryland crops (27%), the majority of which (16%) from tobacco. A variety of sources make up the rest.

Ganyu, sales of wetland crops and petty jobs²¹ were the main income source of the very poor households. Sales of wetland crops, on the other hand, were the main source of income for the poor households; petty jobs, sales of dryland crops and ganyu were other significant income sources for this group. On the other hand, for the less poor, sales of dryland crops, especially tobacco, constituted their main source of income, followed by the sales of wetland crops.

The expenditure level for the whole sample was found to be about K 3,100, (890, 2,590 and 7,800 for each strata respectively). Grain purchase accounts for the single major expenditure item for the very poor (40%). If all the food through ganyu work was monetized, grain purchases would make up more than half of their budget; health expenditure for this group is proportionally larger than for the other two groups. Expenditure patterns for the less poor differ considerably: grain purchases accounted for only 5% whereas hiring of ganyu and input purchases accounted for 30% of their budget.

INCIDENCE OF POVERTY BY TYPE OF WETLANDS

Incidence of poverty was cross tabulated by type of wetlands studied; the results are summarized in the Table below.

Although the sampled Footslopes and Streambank types were too few to allow for generalization, it would appear that the highest incidence of poverty seems to occur in valley floor dambos and in floodplain, followed by upland dambos, as dependency on wetlands was found to be strongly related to a high incidence of poverty. If poverty alleviation is the main concern of the proposed programme, priority interventions would need to take place in these types of wetlands.

INCIDENCE OF POVERTY BY DISTRICTS

District level incidence of poverty is depicted in Text Figure 5.9 below. This confirms the findings of other studies and surveys, mainly that although poverty is widespread, it tends to concentrate in the Lake Chilwa - Phalombe, Lower Shire Valley, and Northern Lakeshore regions where all the sample households were in the very poor or poor category.

GENDER AND POVERTY

There is a gender dimension to Malawian rural poverty. Female headed households were found to be invariably worse off than the population at large: one in two female headed household falls into the very poor category and nearly all the others in the poor (see Table below).

Status of Female Headed Households

Female heads of households are either widowed, divorced, unmarried or the wives of men working elsewhere. Those in the first three categories, who are the majority, are the poorest since they do not receive any cash remittances.

The situation of women household heads in patrilineal societies is very precarious as widowed or divorced women returning to their home village may become semi-landless as most of the available land would have been taken up by their brothers. The dependency ratio in female headed households is very high. Consequently, lack of labour was usually their main constraint. Most of the women heads of households encountered work twice as hard as the married ones, since they must look for ganyu employment, for which they are remunerated far less than a man for similar tasks, in addition to cultivating both their dryland and wetland fields and to caring for their children. In spite of their eagerness and of their efforts, they lack the financial resources to purchase hybrid seeds or other inputs, which has an adverse effect on productivity and on their food security. This in turn leads to malnutrition of both mothers and children. They are the most vulnerable amongst the groups at risk. The first priority of the majority of them was to obtain interest free credit for inputs in order to escape the poverty trap.

CAUSES OF POVERTY

Factors that the villagers perceive as leading to poverty are: drought; holding sizes too small; declining soil fertility; poor crop production; changing of river course and abnormal floods (in the specific case of floodplain); and sickness. The recent increase in fertilizer prices and the low and fluctuating producers prices were worsening their situation and contributing even further to their poverty. Death or ill-health of the male head of the household, or of the wife, with the consequent loss of labour unit, leads to the decline in the status of the household and ultimately to poverty. The same applies to divorce, which impacts negatively both men and women. Women in patrilineal societies and men in matrilineal societies are adversely affected by divorce because, in addition to losing the labour necessary for the normal functioning of the production unit, they lose their users' rights to the land, acquired through their spouses. They must return to their original household and try to acquire land there, but, in areas where land pressure is high, or where the original household has very limited land resources, they may end up semi-landless.

CONSTRAINTS AND PRIORITIES AS PERCEIVED BY FARMERS

Constraints

Constraints to increased production and incomes from wetlands can be divided into three categories: household level constraints, wetlands-specific constraints and institutional constraints.

Household Level

Constraints at household level can be summarized in order of importance for all strata, as:

• lack of capital to buy inputs (the high cost of inputs, combined with the high cost of credit, puts the use of fertilizer and pest/disease control beyond the reach of most farmers)
• lack of markets for perishable produce
• lack of clean drinking water (this has resulted in endemic water-borne disease, sickness and loss of productive capacity in family labour)
• lack of ready supply of good quality vegetable seeds
• pests and diseases of wetland crops (because of inability to buy chemicals)
• drying out of the wetlands in the latter part of the dry season
• low and declining productivity on the dryland.

Other household level constraints included a lack of accessible health facilities, flooding and waterlogging, transport, and the labour involved in watering vegetables. For very poor and poor households lack of food and general poverty were additional constraints.

It is interesting to note that watering of vegetables was not perceived as a very important constraint in the farming systems overall: it was mentioned by only 7% of the less poor households (against 2% of the very poor and 3% of the poor households). Nevertheless, farmers do consider watering of vegetables (carrying water more than lifting it) a strenuous task but, compared with all the other problems they are faced with, it is not perceived as a major one.

This point is illustrated clearly by the Text Figure 6.2. The team asked villagers to rank, on a scale from one to ten, all the tasks related to farming, the higher the score the more arduous the task. Watering dimbas received a score of seven, less strenuous than land preparation and weeding of the dryland fields, but more difficult than all other dimba related tasks. This could obviously have important implications for the appropriateness or otherwise of manually operated pumps.

Wetlands Specific

Generally, there was convergence between village level constraints and those of individual households, except that village level interviews revealed two additional ones, namely lack of fuelwood and environmental degradation, mentioned especially in the footslope dambos, in the upland dambos and, to a lesser extent, in the floodplains.

The main wetlands-specific constraints can be summarized as follows:

• in upland dambos lack of markets for fresh produce, lack of capital, size of holding, pests/diseases and input supply
• in streambanks lack of capital, marketing, input supply, pests/diseases, drying out of the dimba and drinking water
• in valley bottoms input supply, marketing, lack of capital, pests/diseases, waterlogging and drinking water
• in footslopes marketing, lack of capital, holding size and environmental degradation
• in floodplains poor water control, flooding, lack of capital, drying out of dimbas, waterlogging (of dryland) pests, input supply, transport and drinking water.

As can be seen, the main constraints across all wetland types are lack of capital, marketing, input supply, pests and diseases, drying out of dimbas and lack of safe drinking water. However, each type of wetlands had its own specific problems: for example in floodplains the widespread and frequent occurrence of flooding is the most important constraint. The list of constraints is manifold and there are concerns that go beyond improvements of wetlands cultivation.

Priorities

Household

Household priorities common to all strata were higher producer prices, (e.g. better marketing of vegetables but also an increase in the Government guaranteed price of produce such as maize, paddy, sunflower), credit, improvement of drainage and flood control measures, availability of vegetable seeds, early maturing varieties, especially for paddy and maize, and safe drinking water. Not surprisingly the introduction of water lifting devices was a priority for 20% of the less poor farmers (against 2% of the very poor and 6% of the poor) since they operate their dimbas on a commercial scale and cultivate larger areas.

Village Level

Priorities at village level, by type of wetlands, corroborate the suggestion above that there are concerns that go beyond improvements of wetlands cultivation. In fact, village level priorities fall into four main categories:

• those directly related to improving the status of cultivation of the wetlands, e.g. flood and water control measures for floodplain, river diversion and/or training and water control in valley bottoms, improved water delivery mechanisms in upland dambos and streambanks, etc.;
• those associated to improvement of backward and forward linkages or services: input supply, marketing, early maturing varieties, transport, credit and erosion control measures;
• more general ones aimed at improving living standards such as safe drinking water, health facilities, woodlots
• those directly related to policy decisions, e.g. an increase in the floor price of field crops such as cereals.

1 Exemplified by the percentage of the Extension Planning Areas (EPA) falling into the lowest 40th percentile of income. The cut off point was EPAs with 40% of households and below falling into that category.

2 World Bank (1996), Malawi- Human Resources and Poverty.

3 The HIV epidemic, however, is expected to slow population growth.

4 Infant mortality rate = number of children dying before their first year for 1,000 live births. Child mortality rate = number of children dying between 12 and 59 months per 1,000 live births.

5 Information based on World Bank Draft (1995), Southern Africa - Development and Aids Challenge.

6 While the Ngoni in the North are patrilineal, those in the South are matrilineal.

7 The only exception is found in Lilongwe ADD where in the 1970s about 52,000 hectares, held under customary tenure, were converted in freehold titles.

8 In the Southern region, however, where land has become very scarce, the study team came across several cases where surplus land held by households had been amicably reallocated , i.e. reallocated with the consent of the person who had the land, in order to avoid landlessness.

9 Lack of objection is often implied as consent.

10 K.M. Kwofie, State of Food Insecurity in Malawi, March 1966, Volume 1, Part 2, pp. 4-7.

11 Traditional beliefs concerning child care and nutrition persist nearly everywhere.

12 Sasaki, Y. et al (1994), Bilharzia survey in and around Lilongwe in 1994, (Interim Report), Malawi, DHS, (1994), 1992 Demographic and Health Survey, Summary Report, and Malawi, National Schistosomiasis Control Programme, Ten Year Plan, 1991-2000 (proposal)

13 Schistosomiasis transmission occurs through snail intermediate host. In general transmission tends to be influenced by the presence of water in rivers, natural man-made lakes and in dambos.

14 Ibid. pages 22-23.

15 Difficult access was defined as follows: for water source and primary school walking distance (one way), 3 km or more; for all other services 5 km or more. Ibid., p24.

16 This latter often comes from Mozambique.

17 For cassava and sweet potatoes, crop estimates were made by the household members being interviewed. Cassava was converted into dry weight, ( 30% of the wet weight).

18 The equivalency scale applied here is the WHO scale developed for sub-Saharan Africa, which is based on average calorie need for each age and gender groups. The scale took the following values: AE = 1 if adult; AE =.95 for child 10-12; AE =.78 for child 7-9; AE =.62 for child 4-6; AE =.36 for child 1-2; AE =0 for less than 1 year old children.

19 Eucalyptus, neem and gmelina are the most popular species found; in one or two villages, as a result of recent efforts to promote agroforestry, some farmers were starting to plant Acacias.

20 Figures calculated on an adult equivalent basis.

21 Such as grass cutting and sale, sale of firewood and charcoal.