| Social exclusion and Africa south of the Sahara: A review of the literature |
| Chapter 3: EXCLUSION FROM AGRICULTURAL LIVELIHOOD |
  
Access to agricultural land is a necessary condition for making an agricultural livelihood, but it is not a sufficient condition. Livelihoods based on land depend on the mobilization of other factors of production - the mobilization of labour and application of a range of productive inputs, such as fertilizers, pesticides, animal traction, and improved implements. Secure livelihoods also depend on the maintenance of yields and the avoidance of agricultural practices which lead to land degradation.
In indigenous land tenure systems in which a household is allowed to put under cultivation as much land as it can use, the ability to mobilize labour resources, through control of the household, is a central determinant of wealth and poverty when land is abundant (see Binswanger and MacIntyre, 1987; Meillassoux 1981; and Smith 1991 for a summary of the literature). As available land becomes scarce, the application of inputs which enhance land productivity become more important than expansion of the area of production in determining total production, and access to those inputs become more important in determining agricultural livelihoods. "Modern" inputs can offer a way of cultivating land intensively without degradation (though this is not always so).
With the commercialization of production, livelihoods also depend upon the ability to cultivate high-value crops and access to output markets.
These considerations mean that exclusion from agricultural livelihoods depends on the interaction between: firstly, restrictions on access to land resources and patterns of land poverty (discussed in section II); secondly, exclusion from access to productive inputs, high-value crops, and output markets; and thirdly, processes of land degradation.
These interactions are complex. They are taking place within an international frame, in the sense that the prices of many of the agricultural commodities produced in Africa are determined in international markets, and agricultural policies are the subject of negotiations surrounding the implementation of structural adjustment programmes.
Recognizing these complexities, it may be stated that whilst the pattern of exclusion from land resources depends critically (though not only) on social identity, the pattern of exclusion from productive inputs, undergraded fertile land, and remunerative output markets depends critically on poverty.
The rest of this section briefly summarizes some case studies which indicate how poverty is implicated in the process of exclusion from agricultural livelihoods. A key point which emerges from this material is that exclusion from agricultural livelihoods is not now solely dependent on what is happening in the agricultural sector, but is closely linked to lack of access to remunerative off-farm employment opportunities.
 3.1 Access to "Modern" Inputs
A principal feature of African agriculture is that agricultural commercialization has proceeded rapidly throughout the continent during this century, but in a spatially uneven way. Commercialization has also been unbalanced, in the sense that many producers are engaged in commercial sales, with or without production for self-provisioning, but they do so with a low level of capitalization in terms of purchase of inputs. Reyna and Downs (1988), who emphasize the importance of this "unbalanced agricultural commercialization", indicate that the utilization of agricultural inputs is much lower than in other Third World areas, and they dramatize the situation by stating that "although most farmers now produce for the market, they do so with essentially Neolithic tools" (p.12).
A recurrent theme of development policy in Africa in the postcolonial period has been to "modernize" agriculture. This has many strands (see Bernstein 1992), but an important component has been the promotion of greater marketed output, increased production and improved productivity through the application of modern inputs (fertilizers, new varieties, pesticides, herbicides). Bernstein suggests that there are three major approaches to "modernizing" agriculture when production is organized on the basis of peasant farms: (i) "the smashing of the peasantry through direct dispossession typically achieved by violence"; (ii) "'bypassing' peasant farming" through large scale production (capitalist or state farm) on unoccupied cultivable or grazing land;and (iii) "to 'lock in' peasants (or at least those commanding adequate resources) ... through higher - and controlled- levels of input and credit use, and controlling (increased) output through the organization of marketing and processing, thus achieving greater commoditisation, specialization, and standardization" (p.8-9). In Africa, the first approach was applied in white settler and plantation regions in the colonial period, and there are localized examples of dispossession and usurpation of land since Independence. There also examples of the second approach (such as the expansion of rice farming in Northern Ghana). However, it is the third approach which has provided the main way of "modernizing" agriculture in Africa. It has been highly selective in that it has focussed:
i) on high potential areas with higher and more reliable rainfall versus lower potential and marginal areas
ii) on farmers with more resources versus those with fewer resources
iii) on men versus women farmers, which tend to be associated with (ii) (Bernstein 1992:9).
Bates (1981), in his influential study of the politics of agrarian policy in Tropical Africa, also concludes, from a completely different theoretical perspective to Bernstein, that policy has been inherently exclusionary. He suggests that a major thrust of agricultural policy in the 1970s involved output pricing policies which depressed returns for all producers alongside input pricing and supply policies which selectively subsidized the production costs of a few farmers. Reviewing evidence on access to "modern" inputs, he concludes that "It is commonly and almost universally found that the poorer, small-scale, villagelevel farmers do not secure farm inputs that have been publicly provisioned and publicly subsidized as part of programs of agricultural development" (p.55).
Agricultural inputs can become a stake in clientalist politics (see section 5.1, below), and access to those in state office can become important in gaining access to inputs. Crehan (1991), in a down-to-earth description of the situation in North-West Zambia notes that "access to almost any market beyond the immediate village community depends on various extra-economic relations ... on the managing of political relationships as on the manipulation of market forces" (p.194), and that "the poorest group of villagers are those who tend to have least direct contact with the state institutions and officials, and to be least involved in making specific claims on such bodies" (p.206). Eyoh (1992a and 1992b) provides a close study of the politics of access in the Lafia Agricultural Development Project in Nigeria. He argues that local power structures are of central importance in determining the allocation of resources, and that favoured groups include both the leading strata of the peasantry (so-called "progressive farmers") and also members of particular cultural communities.
A major theme of structural adjustment programmes in Africa has been to dismantle the agricultural policy structure, which was seen in terms very much similar to the analysis of Bates (1981), by raising producer prices, and by reducing state controlled output marketing, subsidized inputs and state-controlled distribution of inputs. For households with very little land, improved producer prices have not brought major benefits. Removing subsidies on inputs and promoting their distribution through the free market has meant that access to inputs is now regulated by price and purchasing power rather than political connections and bureaucratic allocation. The poor are still excluded.
3.2 Land Degradation and Exclusion
Land degradation is proceeding rapidly in many parts of Africa and as it occurs the link between access to agricultural land and access to agricultural livelihood is being severed. Land degradation is becoming (or is likely to become) a major mechanism through which people are being pushed off the land. In a land tenure system in which a major mode of access to land is through membership of a local community, localized land degradation can have effects which are similar to land dispossession. Persons who live in areas where yields have fallen find it difficult to find land in other areas under indigenous tenure systems if restrictions on access to land for "outsiders" are tightened.
Suliman (1992) argues that an important excluded group in Africa are "internally displaced people". These are persons who are forced to leave their home lands, but who do not cross international frontiers and cannot be classified as "refugees" and thus qualify for international assistance. Using figures published in 1992, he estimates that there are 16 million internally displaced people in Africa, including 4.5 million in Sudan, 4.2 million in South Africa, 2 million in Mozambique, 2 million in Somalia, 825,000 in Angola, 500,000 in Uganda and at least 500,000 in Liberia. The ratio of internally displaced people to refugees is 4:1; most are women and children, and they move to urban centres. Suliman argues that persons are displaced mainly as a result of environmental degradation and armed conflicts, which are sometimes related to environmental resource competition. He notes that in the drought of the early 1980s, more than 10 million people had to abandon their homelands in search of food and water by early 1984, that one million people in Burkina Faso (a sixth of the country's population) migrated to urban centres, and that the number of food aid-dependent people in the Sahel region was 5-10% of the total population.
One important mechanism through which poverty is linked to land degradation is what has been called the "simple reproduction squeeze" (Bernstein 1977). A simple reproduction squeeze occurs when a peasant household which is land poor and producing for the market is faced with deteriorating terms of trade (between necessary purchased commodities and agricultural inputs on the one hand and agricultural outputs on the other hand) or declining yields, and can only reproduce itself (even with hunger) if it intensifies labour inputs and land exploitation. In these circumstances a vicious circle can set in as actions to reproduce the household in the short-term lead to land degradation and a tightening squeeze in the medium-term. To the extent that households are irreversibly integrated into markets (and particularly the extent to which their subsistence is money-mediated), they cannot escape this squeeze without other sources of income. Withdrawal from markets to escape a "simple reproduction squeeze" can also involve decisions with long-term consequences - such as withdrawing children from schooling.
Land degradation can also be associated with spatial marginalization, if people are pushed on to poorer and more ecologically fragile land through dispossession for agribusiness (for analysis of an example of pasture degradation in the Sahelian zone, see Blaikie 1985, ch.7). A study which is important in understanding the relationships between such exclusion from land resources, land degradation and the search for alternative sources of livelihood through labour markets is provided by Arrighi (1970).
Arrighi's classic study analyzes African responses to the expropriation of three-quarters of their land by European settlers in colonial Rhodesia. His analysis is of importance as it looks at a process of exclusion in terms of the practices of the excluders and the excluded, and shows how an initial act of exclusion from a critical resource, land, led, over a long period of time, to the irreversible closure of livelihood options of the excluded population, until finally a larger and larger proportion, although they retained access to some land, could only seek work on labour markets to survive. The analysis is complex, but central structural elements are: (i) the increasingly necessary, rather than discretionary, character of involvement in the money economy, which was initially provoked by the introduction of taxes and rents, and which subsequently depended on rising expectations regarding "necessary subsistence"; and (ii) the comparative "effort-price" of money income obtained through the sale of agricultural produce and the sale of labour-time.
Arrighi observes a key shift in the behaviour of labour markets in the 1920's in that "While before 1922 African participation in the labour market did not increase in periods of falling wages, after that year it always increased irrespective of whether real wages were falling, rising or remaining constant" (Arrighi and Saul 1973:191). This shift is associated with the rising effortprice of African participation in agricultural produce markets, which was the result of falling yields in the "Native Reserves", the term used to designate legaly the poorer quality land distant from the railways which had been left for the African population. Falling yields were the result of higher population densities, associated with natural increase and the shift of the African population into the "Native Reserves", and also a pattern of agricultural innovation and investment on the Reserves (in cattle and ploughs) which had a land-consuming bias. A slump in prices of cattle and maize in 1921-23 accelerated movement into the Reserves, and thereafter land shortages appeared and land degradation (soil erosion, drying-up of springs, exhaustion of pasture) set in. It may have been possible to arrest this pattern through shifting the pattern of investment of surplus towards high-value tobacco crops. But this was not undertaken, partly because of the smallness of the surplus and partly because in the late 1930s there was a major re-orientation in the pattern of use of surplus towards expenditure on education.
The details of this analysis have been challenged (see, for example, Low 1986), and literature since the mid-1970s has emphasized the gender dimensions of patterns of involvement in labour markets (see below). However, Arrighi's approach offers one promising mode of analysis of exclusion.
3.3 The Importance of Access to Remunerative Off-farm Income
An recurrent theme emerging in studies of patterns of poverty in rural areas is that the poorest groups are those without access to remunerative off-farm income. A framework within which differentiation is analyzed by various authors is to identify processes of upward mobility which operate over short or long periods of time and to explain the poverty of particular groups at a given observed point in time in terms of their exclusion from these processes owing to various constraints. In this frame of reference, access to remunerative off-farm income has been identified as a key factor in upward mobility in rural areas.
This off-farm income may be earned through activities in rural areas, notably when a farmer also engages in trading activities. However, often off-farm income is derived from urban activities, including both earnings from formal and informal sector employment and from state offices. This pattern of investment activity, which bridges rural and urban locations, formal and informal activity, and the public and private sector, was originally identified in Kenya by M.P.Cowen (see Kitching 1980), and called "straddling". According to Iliffe (1983), straddling is much more common in East Africa than West Africa, but Bayart (1993) notes examples from Cote d'Ivoire, Liberia, Sierra Leone, Nigeria, Zaire and Mali, and further examples are reported in Downs and Reyna (1988).
In the rest of this sub-section, three case studies are presented. In all three, access to remunerative off-farm income is an important condition for upward mobility, and in the last "straddling" is evident.
Hill (1972) is a classic analysis of economic inequality and poverty in a Hausa village in Northern Nigeria. A major aim of her study is to explain why, even with abundant land and reasonable soil fertility, a certain proportion of farmers within the village were experiencing some degree of destitution. In broad terms, Hill summarizes the explanation of the situation by the fact that some farmers were "too poor to farm", and in a "concluding speculation", which she suggests may be applicable to other areas of Hausaland, or even "other West African savannah peoples", she states that "some degree of destitution is likely to occur in most communities where permanent cultivation [of manured farmland] is the preferred agronomic system, and where poorer farmers have few opportunities of significantly supplementing their income by growing special crops (other than grains or groundnuts), or by pursuing remunerative non-farm occupations" (p.191).
In the village which Hill did her study, cultivation was based on manured farmland, which was bought and sold, in the area around the village, and a broader zone of bushland, in which members of the village could establish usufructory rights. The main crops grown were grains (millet and guinea corn) and groundnuts. The main source of labour for farming was household labour organized through the Hausa institution of paternal gandu, "a voluntary, mutually advantageous, agreement between father and married son in which the son works in a subordinate capacity on his father's farms in return for a variety of benefits including a share of the food supplies" (p.38). But labour hiring was also occurring in the village. Women were in seclusion, but engaged in trade, particularly in cooked meals from their houses.
Within the village, there were significant inequalities. In concrete terms, the number of working-men in each of the richest farming units was twice as great as the number in the poorest units; the richest 15% of the household heads owned 43% of the manured farmland, whilst 12% of the heads had less than one acre; the ten richest farmers accounted for one-half of the hired labour; nearly all the poorest farmers sold manure.
Hill argues that there is a general tendency for wealth to increase with age, and she identifies "short-term spirals of upward mobility" in which particular households get richer by consolidating and extending the household, purchasing land, hiring labour to cultivate a larger acreage, and purchasing manure to increase yields. Accumulating profits from remunerative non-farming activities was an essential part of this spiral of upward mobility. The richest farmers had remunerative secondary occupations as traders, craftsmen and Koranic students and teachers. Particularly important for them was local graintrading. Most rich men aspired to store grain, buying it from the poorer farmers when prices were lowest and re-selling it later when prices were much higher.
Poorer households, in contrast, tend to get stuck in a situation of "individual impasse". A key aspect of this was that most of the poorer farmers did not produce enough grain to last throughout the year and they subsisted on bought grain for most of the year. Moreover the poorest farmers did not have the capital or special skills needed to engage in remunerative off-farm activities and most were forced to engage in collecting and manufacturing freely available goods (especially firewood), or to undertake miscellaneous poorly remunerated tasks. In the individual impasse of poor farmers, poverty itself results in an interlocking array of exclusions from opportunity. As Hill puts it:
Poor men applied less manure to their farms and obtained lower yields per unit of effort; poor men were those who sold their grain immediately after harvest when prices were lowest; poor men had unremunerative types of non-farming occupations; poor men often had 'no time to farm' - their granaries being empty soon after harvest, they were obliged to pick up a living from day-to-day by working in odd jobs for others or by collecting 'free goods' such as grass or firewood; poor men (being farm sellers) often owned insufficient manured farmland to set their sons to work: poor men could seldom borrow money, being considered bad risks (p.164-5).
In short, in this village, it was possible to be "too poor to farm". But Hill emphasizes " there is "an absence of class", as the spirals of upward mobility are only short-term, and when the gandu-head dies, property is divided and the farming business falls into disarray.
Kitching (1980) is a long-term study of class formation in Kenya from 1900-80. As part of this study, he examines patterns of adoption of new high-value crops (coffee, tea, pyrethrum and hybrid maize) during the period 1952-70. The introduction of these crops, along with land consolidation and enclosure, is sometimes described as an "agrarian revolution", but Kitching shows that the new crops had, by 1970, been adopted by approximately one-third of smallholders. The adoption of the new crops increased the incomes from agricultural activity, and using survey evidence, he suggests that by 1970 households could be divided into four categories according to their farm incomes: (i) large holdings with one or more of the new small enterprises (high incomes); (ii) small smallholdings with one or more of the new farm enterprises, and (iii) large smallholdings without any of the new farm enterprises (both intermediate income); and (iv) small smallholdings without any of the new farm enterprises (low incomes) (p.330). He seeks to explain the pattern of differentiation and argues that off-farm income is important to both the richest and the poorest households (for different reasons) and that there is a "close relationship between above-average farm incomes from larger holdings and access to off-farm incomes" (p.361).
Focussing on the poorer households, "those left out in the cold by the 'agrarian revolution"', he writes:
Thus those left out were those with the ecological opportunity but without the land, capital or labour power to take that opportunity .... those without the ecological opportunities and without enough land to make up for yield deficiencies by the size of cultivated area .... and those with the land and the opportunity, but essentially without the capital.
Common to all of them was the marked narrowing of alternatives due to the lack of access (either in 1952-70 or before) to sufficiently large sources of off-farm income. With such sources land could be bought or, if the latter was poor or semi-arid, better land could be bought elsewhere. It was thus the 'elbow room' given by access to above average off-farm income (particularly wages and salaries) for which it was so prized. Failing access to it, the only alternative was the diversion of labour power off the holding into lowpaid off-farm employment to take the subsistence strain of the holding and perhaps raise household subsistence levels a little. But of course there was always the hope that the latter form of adaptation to rural poverty (what might be called a migration out of poverty into poverty) could be turned unto the former by upward mobility. The aim of every migrant was to turn 'average' or 'below average' off-farm earnings into 'above average' earnings, and thus start on the path of accumulation which he had seen others tread. (p.373- 4)
An important element in Kitching's account is that off-farm income was crucial for both richer and poorer households (either for accumulation or to supplement subsistence). What mattered in determining the pattern of differentiation was both the level of remuneration (which depended on what the off-farm employment was), and also "the time at which it was acquired, the period for which it was available and the price of land over that period as a proportion of income". He describes a situation of closing opportunity in the sense that with increasing land shortage and escalating land prices, "the 'land purchase' power of the nominally higher salaries in the early 1970s was probably less than that of smaller money incomes obtained in the 1940s and 1950s" (p.371). By the 1970s, too, "the type of labour market to which a household sent its labour power was a good indication of its relative prosperity", in that 'la poor household would be involved -almost entirely in the urban and/or rural informal labour markets, while a more prosperous household would be far more likely to have its labour power concentrated in the higherincome occupations of the urban formal sector and to be hiring in the labour power of the poorer households" (p.403). Significantly, Kitching argues that the interrelationship between agricultural livelihoods and off-farm labour markets changed over time - "Whereas up to 1952 trends in the wage labour market were primarily explicable by agricultural change on the smallholding, after 1952 to a very large degree trends in agricultural production became the dependent variable controlled by the off-farm labour market" (p.374).
Kitching's study is a classic example of "straddling", and it points to the importance of exclusion from labour markets as a critical factor in determining exclusion from remunerative agricultural livelihood. A further study by Cousins, Weiner, and Amin (1992), which summarizes the literature on current patterns of social differentiation in the communal lands (former "Native Reserves") of Zimbabwe, also shows the importance of off-farm employment, and complements Arrighi's historical analysis (above). The study shows that differentiation through straddling is also important under indigenous land tenure systems, and it throws further light on the "land-grabbing" described in the last section.
Cousins et al. identify four distinct categories of household in the communal areas: (i) "households which regularly produce a surplus, invest in the agricultural means of production, and hire significant amounts of wage labour" ("rural petite bourgeoisie"); (ii) "households able to reproduce themselves from rural production alone" - including agriculture, craftwork, beer-brewing, construction, etc. ("petty commodity producers"); (iii) "households able to reproduce themselves through a combination of rural production and wage labour" ("worker-peasants"); and (iv) "households unable to reproduce themselves, from whatever source, without external assistance from either other households (usually within the some kinship network) or from the state (usually in the form of drought relief)" ("a lumpen semi-peasantry") (pp-11-13). The critical processes enhancing social differentiation are a series of interlocking patterns of unequal access - unequal access to land, livestock, technology and agricultural capital; inequalities in wage labour incomes; and unequal political clout at the local level, which is important as access to resources is "skewed by political capture" (p.16). These processes intersect with differentiations by agro-ecological zone, gender, and generation.
Important observations of Cousins et al., which probably have wider applicability, include the assertion that "The distribution of land is usually less skewed than other assets such as livestock and technology" (p.13); and "Inequalities in wage labour incomes contribute to inequalities ownership of means of agricultural and livestock production, in both the short term (the purchase of inputs for annual cropping activities and he hiring of local wage labour), and the longer term (investments in tools, equipment, livestock)" (p.14). Cousins et al. argue that there are a number of constraints on differentiation, including the indigenous tenure system, kinship relations in rural areas, and the general difficulty of accumulation in what are marginal areas.
The historical circumstances of the 1980s are also foreclosing options. Because of a lack of formal sector employment opportunities (net new jobs are said to absorb only 10% of the million school leavers since Independence), "the strategy of capitalizing an agricultural base through savings from wage employment is becoming less viable over time" (p.16). Moreover, whilst there is no landless class, the numbers of households who may be described as part of the "lumpen semipeasantry" are increasing. As Cousins et al. put it:
The combination of unemployment, environmental degradation and drought is responsible for a 'reproduction squeeze' for many of rural households; in some regions relative over- population is stretching the resource base to the limit increasing the number of households falling into the ranks of the lumpen semipeasantry. A significant proportion of these are female-headed and relatively young households, and in general this trend is more marked in the semi-arid regions of the country (p.19).
3.4 Female-Headed Households
To conclude this section,it is worth underlining the position of female-headed households with agricultural livelihoods, who are often poor because of limited ability to mobilize labour for farming, restrictions on access to credit and inputs, and exclusion from off-farm income. These are an important "excluded group" in southern Africa, where a common pattern is for the men to migrate to work in the urban areas and the women to work land to which they have access through continued use under indigenous tenure systems (see Bush, Cliffe and Jansen 1986). The internal dynamics of these "doubly-divided' households is complex, and conflicts, and abandonment, may particularly reflect the impossible position both women and men find themselves given their allotted gender roles in conditions of economic stress (see Whitehead 1990 for a brief,but subtle account). But whatever the dynamics,
There is a strong correlation between the destitute, 'poor peasant' families and the women- headed households who number roughly 30% of rural households in most of southern Africa. many of these are without regular remittances of earnings from elsewhere and their own production is massively constrained as they have less chance of access to land, and to oxen for ploughing, and are also short of labour at crucial seasons, with the result that, in Botswana, 80% of all those working for rations' in drought relief projects were women. (Bush, Cliffe, and Jansen 1986: 298-99).
A good case-study, which documents the process of exclusion of women from credit and inputs in an agricultural development scheme in Malawi, is Chipande (1986). That study clearly shows how the inability of female-headed households to mobilize labour led to a poor credit rating for those households and, as a result, severe restrictions on access to inputs.
|
