Reforming Peasant Production in Africa: Power and Technological Change in Two Nigerian Villages Dickson L. Eyoh ABSTRACT zyxw zyx Differential access to state-allocated incentives, based on socio-economic inequalities in rural society, is commonly assumed to be a key determinant of change in rural Africa. This article argues that, given the spatial diversity of Africa’s rural political economies, analysis of the politics of rural change needs t o be premised on a n appreciation of the multiplicity of social relations through which rural power structures are configured. Data from a field study of a World Bank assisted agricultural development project in Lafia, Nigeria, are used to illustrate the manner in which spatial and inter-community variations in responses to commercialization, cultural divisions and the reorganization of political relations during the colonial era combine to sustain regional power structures which are defined by such differences. A comparative analysis of two village communities at opposite ends of the regional spectrum of commercialization is employed t o demonstrate how such power structures provide a framework within which the political conditions of access operate to the advantage of both dominant socio-economic strata and members of particular cultural communities. 1 . INTRODUCTION A widely shared consensus now exists that state domination of economies has been the root cause of poor agricultural performance in sub-Saharan Africa. This has encouraged the view that the promotion of macro-economic frameworks within which resource allocation and producer decision-making are governed by market signals is essential t o improving productivity in peasant-dominated agricultural systems (see Crook, 1988). Development and Change (SAGE, London, Newbury Park and New Delhi), Vol. 23 (1992) No. 2, 37-66. zy zy zy D.L. Eyoh The advocacy of a greater reliance o n markets articulates a demand for a restructuring of domestic power relations so as t o expand the range of producer incentives. For neo-Marxists, who presume peasants t o be more technically competent than state bureaucrats and their external allies in relating their economic goals t o capabilities, the enhancement of local decision-making autonomy is critical for liberating their productive potential. If combined with a nurturing of autonomous local organization, market decentralization will facilitate peasants’ ability t o defend their interests against predatory ruling classes a n d states (Brett, 1986; Dutkiewicz and Williams, 1987; Williams, 1985a). From the standpoint of the rational choice perspective of Bates (1981), the best prospects lie in expanding the ranks of large commercial farmers who a r e better placed to promote rural producers’ interests within policy-making circles. As the World Bank has come t o assert with increasing boldness that Africa’s economic malaise is primarily the result of ‘poor governance’, the political thrust of its pro-market agenda namely, that a greater reliance on market forces will ensure the rise of farm gate prices by mitigating the power of ‘rent-seeking’ urban elites and what it claims are comparatively overpriced urban workforces - has become less obscure (see World Bank, 1989: 22-9). Whether African states are capable of sustaining such dramatic changes in the political environments of economic policy-making remains a n open question. However, even if their role were restricted to creating conditions which are conducive t o the growth of private markets, the dissemination of agricultural research and the development of infrastructure, it is certain they will continue t o influence the conditions of agriculture development. In other words, political forces will always impinge on the structures through which incentives are allocated, and rural power relations will remain important in determining the conditions of differential access t o resources, which students of African agriculture have long recognized as a crucial ingredient for agrarian differentiation (Berry, 1984; Williams, 1985b). While directing attention to the significance of the political context in shaping agricultural policy choices and performances, many of the contributions to the debate on agricultural reform agendas for sub-Saharan Africa have been handicapped by a proclivity to what may loosely be termed an ‘urban bias’ perspective o f staterural relations. This is the view that the interests of rural producers zy Reforming Peasant Production in Africa zyxzy 39 are overwhelmed by the political power of coalitions of diverse urban groups which represent a greater threat to ruling elite incumbency. The experience of state-organized collective victimization of producers arising therefrom establishes a complementarity of interests which overrides differences in interests between groups in farming communities. Indeed, the endorsement of modified versions of ‘trickle down’ which emerges from many micro-societal level studies of the politics of agricultural change is related to an adherence to this perspective. A common conclusion of such studies is that rural patronage networks led by richer farmers are the best political means open to peasants to extract resources from states. Because of shared membership in communities, richer farmers boost the moral authority which is necessary to mobilize peasants on the basis of common grievances and concerns. Although they tend to capture a disproportionate amount of resources acquired through collective political action, such inequities are the inspiration to leadership without which surbordinate strata will remain all the more vulnerable to state-sanctioned exploitation or neglect (see Clough and Williams, 1987: 187-92; Holmquist, 1984). The distribution of political power between peasantries and urban-based classes in Africa elicits empathy for the ‘urban bias’ perspective and support for ‘trickle down’ as a redistributive mechanism. Nonetheless, because it is inclined to reduce rural power structures to expressions of primarily socio-economic differences, macro-societal as well as micro-societal level analyses which are informed by this perspective are not very attentive to how other politically relevant social divisions bear on the abilities of varied rural groups to harness, or to be excluded from, incentives. To raise the issue of allocational inequalities is not simply to express a normative commitment to social justice. Even if it is true that richer farmers are most capable of defending peasant interests within state arenas, it is by no means obvious that economic benefits derived through their leadership will be shared with subordinate peasant strata. Secondly, as has been amply documented, richer farmers are inclined to shift their economic interests to other areas, for example, transportation, commerce and money-lending, which offer higher returns on investments. Thus allocational processes that are biased in their favour are not guaranteed to assist the goal of enhancing productivity in peasant agriculture. Lastly, and of immediate relevance here, is the fact that commercialization and zyxw 40 zyxwvuts z D . L . Eyoh zyxw state penetration of rural society has often been accompanied by other change-inducing phenomena, such as migrations, the growth of new religions and other modes of cultural expression. The interaction of such processes has engendered regionally distinct agrarian political economies whose power structures incorporate economic and ethnic divisions (Barker, 1989: 70-100; Cliffe, 1977; Samoff, 1980). They sustain political contexts in which the conditions of access to resources operate, without reference to individual capabilities as producers, to the advantage of some classes, groups and communities over others. Seen from the standpoint of even this limited range of considerations, issues related to allocational inequities warrant serious consideration in the debate on the reform of peasant production. A deeper appreciation of the probable consequences of allocational processes, which are subject to the play of political forces, on productivity reform programmes therefore necessitates a more sophisticated view of the political contexts and dynamics than is offered by the ‘urban bias’ perspective. It requires attention to the multiplicity of social relations through which rural power configurations are constituted and expressed in particular local/regional settings. Such a focus is all the more important because of the enormous spatial variations of agrarian production systems, rural social hierarchies and political-economic networks through which peasants are linked to wider structures of class and power. This article provides an example of how regional structures of power shape the nature of political responses in two village communities to administratively imposed attempts to increase the productivity of peasant farming systems in Nigeria. It does so through a case study of a World Bank designed and partially financed agricultural development project; the Lafia Agricultural Development Project (LADP) in Plateau state.’ The article is in four parts. This introduction is followed by a summary of the project’s design, objectives and major outcomes. The third section outlines the defining characteristics of regional production systems and sociopolitical structures so as to explain the major determinants of those outcomes. The fourth illustrates how the constitutive elements of IocaVregional power structures bear on the abilities of different strata and community groups to gain access to externally emanating resources in two village communities. These four parts are followed by a conclusion. zyxw zyx zy zyxw Reforming Peasant Production in Africa 41 2. THE LAFIA PROJECT: DESIGN, OBJECTIVES AND OUTCOMES Agricultural development projects (ADPs) emerged in the mid- 1970s as the main institutional device employed by the Nigerian state to reform its peasant-dominated food production systems. Starting with three pilot projects in Kaduna, Bauchi and Sokoto states in 1976, the ADP programme was extended to the Middle-Belt states of Benue, Niger and Plateau in 1977. The programme was declared the cornerstone of a national ‘green revolution’ strategy in 1980 and a commitment was made to establish projects in all states. The core component of ADPs is their crop development programmes, which are geared to increasing the yield levels of peasant food production systems through the dissemination of ‘green revolution’-type technologies: crop packages comprised of improved seed varieties, fertilizer, pesticides and tractor services. A number of infrastructural supports, most importantly rural road networks, farm service centres for input delivery, intensified extension services and soil conservation measures, are developed to facilitate this objective (Beckman, 1987; Bonat and Abdullahi, 1989; Clough and Williams, 1987; Wallace, 1980). ADPs are able to instigate changes in peasant production systems by acting as vehicles for the sudden infusion of large quantities of material and human resources into the selected localities. Their inherent capacity to lubricate changes in rural social structures stems from a derivative potential, which is based on their in-built ‘discriminatory logic of resource allocation’ (Beckman, 1987: 133), for altering patterns of resource distribution and opportunities for accumulation. The design of crop development programmes rests on a critical division of farming communities between ‘progressive’ and ‘ordinary’ farmers. Because they are more commercially oriented and thus presumed to possess a greater capacity for innovation with new farming technologies, the projects specify that available resources should be preferentially allocated to ‘progressive’ farmers. This logic of allocation ensures that the dynamics of local power relations play a crucial role in regulating access to resources. The Lafia project encompassed an area of approximately 9400km2 which make up the Lafia and Awe local government areas in the south-eastern corner of Plateau state (see Figure I).* Contained within its estimated population of 250,000 are slightly over 50,000 households or farm families.’ The population is a zyx zyxwvu P N z z zyxwvuts zyxwvutsrqpo zyx P rc Figure 1. La fa Agricultural Development Project: 1977-83 zy zyxw zyxwv Reforming Peasant Production in Africa 43 socio-cultural mosaic of over thirty-five ethnic communities, which has been formed through a process of immigration by various groups over the last five centuries. The major groups, which together account for approximately 70 per cent of regional households, but none of which is numerically preponderant, are the Eggon (25 per cent), Alago (22 per cent), Tiv (19 per cent) and Migili (12 per cent).4 Over 90 per cent of the economically active population pursue farming as the only or the primary occupation, and depend on manual methods of cultivation with the aid of rudimentary technologies (hoes and machetes). Edible staples, of which yam and cassava are the main root crops and sorghum, millet and maize the main cereals, predominate in the region’s economy. The LADP’s crop development programme identified ‘ordinary’ smallholders as its main target group and estimated that around one-third of regional households (randomly selected) would benefit directly from the programme. The direct beneficiaries were to serve as the medium for spreading information about crop packages and improved husbandry across farming communities. The programme also had a second component which was presented as a subsidiary element. This was to create conditions to encourage ‘technological development and make agriculture attractive to private capital . . . by attending t o the special needs of large farmers and progressive smallholder groups who displayed sufficient business acumen to advance beyond the level of basic crop husbandry and yield improvements’ expected of ordinary smallholders (World Bank, 1977: Annex 7, p. 10). The bias towards uneven patterns of resource allocation inherent in both components of the programme was advanced through a number of institutional mechanisms. First there was the bifurcation of extension activities between a General Extension Divisiori geared to serving ‘ordinary’ farmers and a Farm Management Service attending to the needs of ‘progressive’ and ‘large-scale’ farmers. In spite of a recognized paucity of largescale farming in the region, 2.786 million naira (exclusive of contingencies) was budgeted for the Farm Management Service slightly in excess of the 2.7 million naiia allocated to the General Extension Division. Because of the supposedly high level of fertilizer subsidies (then over 80 per cent), ordinary farmers were excluded from credit programmes.’ Second, both extension units were dedicated t o approaches which favoured a concentration of available resources on a restricted number of farmers. The most prominent examples were the ‘group zy 44 zyxwvut zy D.L. Eyoh farming’ and ‘demonstration farmer’ programmes. In the case of the former, which was meant to encourage mechanized cultivation by smallholders, select groups of progressive farmers were offered credit packages of inputs and cash (to hire supplementary labour) to farm adjacent plots. In the case of the latter, in exchange for their willingness t o be called on at short notice for demonstration purposes, selected farmers received free packages of inputs to farm l ha of their land under close supervision by extension staff. Between 1981 and 1983, for example, between 700 and 800 farmers participated in the group farming programme. During these three years, participants in the programme received approximately 7 per cent, 10 per cent and 6 per cent of available fertilizer, respectively, as components of credit packages. Moreover, during the recurrent periods in which fertilizer was in short supply and the project was forced into rationing available stocks, participants in these and other programmes for progressive farmers were accorded priority and granted larger allocations. Third, the project created organizational structures to mediate relations between management and farming communities. The most important of these were farmers’ advisory committees constituted at the village, farm service centre (FSC) and district levels. Formation of village advisory committees (VACs) was commonly initiated by extension agents, who would inform the village leadership of the desirability of such a committee, and request that a meeting of villagers be convened to select members and an executive. The FSC committees were comprised in turn of executives of VACs within their geographical area of service, and district committees of chairmen of FSC committees in each of the six project administrative districts. These committees were closely involved in the selection of participants for programmes through which resources were preferentially allocated. The relationship between the composition of VACs and the beneficiaries of the group farming and demonstration farmer programmes is central to the following analysis of the political dynamics of unequal access in the two village communities. But first, let us examine the two main outcomes of the project and their primary determinants. The LADP had no noticeable impact on peasant production practices. According to its official assessors, gains in regional crop output over the implementation period were not due to increases in farm productivity that could be attributed to the use of crop packages, but to expansion of cropped area caused by immigration zy zy zyx zyxwv Reforming Peasant Production in Africa 45 and natural population increases (APMEPU; 1982a). Peasants across the board rejected the crop packages, adopting only individual components which could be conveniently incorporated into existing production systems. Of these fertilizer, which accounted for over 80 per cent of the cumulative value of inputs, was the most widely adopted; this was followed by improved seed varieties. The majority of farmers showed scant interest in insecticides and pesticides; over 90 per cent of this class of inputs was sold through credit packages for progressive farmers. Evaluation of the two programmes for progressive farmers which formed the institutional basis of uneven access to resources, confirmed a similar absence of impact on the production practices of their beneficiaries and on existing patterns of income distribution. There was no local history of ‘group farming’. ‘Ephemeral in time, space and numbers’ (APMEPU, 1982a: Vol. 1, p. 22), ‘groups’ amounted to little more than a temporary banding together of farmers for the expediency of securing credit packages. Upon receipt of credit packages, most groups disintegrated and their members either cultivated sections of the group land individually or not at all. The credit programme was to be operated through a revolving loan scheme. After the 1982 farming season, when austerity measures combined with a loan default rate of over 60 per cent to compel a withdrawal of the cash component, interest in the programme evaporated. The ‘demonstration farmer’ programme also failed as a vehicle for disseminating new husbandry practices. Apart from the inappropriateness of the changes in production practices which it fostered (more on this below), its failure was partly due to what project managers readily admitted was the poor quality of the extension staff - especially juniors - who were in immediate contact with the farmers. Farmers’ evaluation of the usefulness of advice received from them often bordered on contempt (see Eyoh, 1989: Ch. 8). Most proved incapable of comprehending basic instructions on the very inputs and husbandry practices which they were supposed to teach to others after completion of their basic training (see APMEPU, 1982a: Vol. 2, Annex 4, pp. 22-7). This outcome was accompanied by geographical and intercommunity variations in the level of input adoption. In other words, the distribution of inputs tended to be concentrated on a limited number of farm service centres. For example, between 1981 and 1983, seven out of twenty-eight farm service centres (two in Lafia zyxwvu 46 zyxwvut zy zyxwv zyxwvu D.L . Eyoh town and the other five in settlements located along major tarred roads and within a 40 km radius of Lafia town) accounted for between 58 and 66 per cent of fertilizer sales. A similar pattern obtained with regard to the distribution of improved seeds (Eyoh, 1989: Appendix 3). As illustrated through data in Table 1, derived from a survey of 250 households (randomly selected but weighted to reflect the numerical position of the four major ethnic groups in the region), the level of fertilizer adoption varied between ethnic communities and strata. Table 1. Fertilizer Application by Ethnic Group (per cent) Ethnic group Tiv Eggon Alago Migili Others Proportion of project cultivated area Proportion of cultivated area receiving fertilizer 29 I2 22 8 30 2 14 30 (24') 45 36 Average for all ethnic groups 21 (18') Note: indicates exclusion of one household with strong effects on overall totals. Source: APMEPU (1982a: Vol. 2, Annex 4, p. 4). 3. EXPLAINING OUTCOMES: PETTY COMMODITY PRODUCTION, DIFFERENTIATION AND POWER IN HISTORICAL PERSPECTIVE A distinguishing characteristic of ADPs is that they operate without a commitment on the part of the state to impose new (capitalist) relations of production on existing relations of production. Because of this, their managements lack the power to assert any measure of direct control over the conditions of access to the main means of production (land, labour and credit) and thus to impinge on the boundaries of peasant decision-making in production. Consequently, the pattern of responses to the technologies and production forms they promote is primarily shaped by the character of production relations and constraints in particular localities. These same factors also define the capabilities of different strata to undertake zy zyx zyxwvu Reforming Peosant Production in Africa 47 changes in production, and the framework of probable changes in rural social structures which could be instigated by uneven patterns of resource distribution.6 Taking this into account, it is clear that while the impact of administrative and organizational deficiencies both within and outside the project cannot be discounted, the explanation for its twin failures must be sought in the interplay of the project’s developmental priorities with three key features of the region’s political economy: (1) the social relations and constraints of household production; (2) geographic and, coincidentally, inter-ethnic variation in the level of commitment to petty commodity production and in the degree of socio-economic stratification; and (3) the overlapping of (2) with an ethnically based political hierarchy which determines the manner in which local groups are linked to national structures of class and power. Agricultural production in contemporary Lafia remains the province of household-based units which combine production for domestic needs and for market, and cultivate an average of 2.5-3 ha annually.’ Regional farming systems are overwhelmingly characterized by inter-cropping in mixtures ranging from two to ten crops spread across several plots. Although the transfer of land rights by short-term leases and direct sales is not uncommon, access to farmland is primarily determined by customary usufruct. The bulk of agricultural labour derives from households. Exchange labour arrangements, followed distantly by casual wage labour, are the main sources of supplementary labour. Notwithstanding the low position of commodity relations in the means of production, a significant degree of inequality in the distribution of farm size and area cultivated by households does exist. However, in the absence of differences in the technological conditions of production, variations in household farm size and therefore the gross value of household production, are primarily contingent on the size and structure (including gender-based divisions) of domestic labour pools (Table 2).* The differences in the proportion of marketed output relative to the gross output of major crops are marginal between households with different farm sizes (Table 3). This indicates that the parameters within which the commercialization of production has proceeded since colonial times has precluded the emergence of class-based divisions in Lafia, and that households retain considerable autonomy over the organization of 48 zyxwvu D.L. Eyoh zyxwvuts zyxwvutsr zyx Table 2. Domestic Labour and Farm Size Average farm size (ha) Average size of family workforce Average ha per worker I .6-2.5 2.6-3.5 3.6-5.5 > 5.6 I .05 2.21 3.11 4.35 7.27 2.15 I .98 2.53 2.94 4.14 0.49 1.12 1.23 I .48 I .76 Overall average 3.46 2.65 1.31 Farm size (ha) < 1.5 Source: APMEPU (1984: 16). Table 3. Farm size (ha) < 1.5 1.6-2.5 2.6-3.5 3.6-5.5 > 5.6 Value of Crop Production and Sales by Farm Size Total value of output (naira) Value of crop sales (naira) Sales as percentage of total 477 68 I 888 956 1288 I93 282 385 42 I 65 I 40 41 43 44 50 Source: APMEPU (1982b: 7-9). agricultural labour processes. The paucity of alternative sources of labour remains the major constraint on the ability of households across all strata to intensify or expand production. Regional systems of production are well adapted to this constraint and to servicing the twin objectives which govern peasant decision-making in production. The LADP’s emphasis on farm enterprise specialization was not only at variance with these priorities but, given the major constraint, it amounted to an attempt to convince farmers ‘to adopt a system which was less productive than their “unimproved” traditional systems’ (APMEPU, 1982a: Vol2, Annex 4, p. 9).’ Furthermore, in the absence of a concerted attempt on the part of the state to redefine the conditions of access to the means of production, the ability of beneficiaries of preferentially allocated resources to alter their production practices to an extent which could have impinged noticeably on the production conditions of others remained structurally limited. Reforming Peasant Production in Africa zyx zy 49 The variation in responses to the project was a reflection of differences between the various communities in degrees of commitment to commercialized production and specialization. The location of farm service centres in large settlements, and the afore-mentioned concentration of input sales in a limited number of these, ensured that access to inputs was much easier in communities with a greater degree of involvement in commercialized production, where the predisposition for experimentation with production-enhancing technologies was bound to be higher. As data in Table 4 indicate, the clearest differences in levels of commercialization run along ethnic group lines, reflecting intercommunity variation in the intensity of commitment to commercial production and in the degree of specialization associated with it. This aspect of differentiation is linked to the settlement patterns of the different ethnic groups, which can be traced to such factors as the timing of their migration to the region and the colonial era transportation networks linking producers with external markets. Since pre-colonial times, settlement in Lafia has followed two broad patterns: (1) nuclear settlements (villages and towns) which, although now multiethnic, originated as foci of earlier arriving groups (Migili, Alago and Kamberi especially) with whom they remain associated; and (2) more ethnically homogeneous and geographically dispersed hamlets inhabited mostly by later arriving groups, especially Tiv and Eggon, whose migration into the region accelerated during the colonial era." In general, the groups which arrived earliest gained most advantage from the transportation network and access to external markets, and thus were exposed to a more intense process of market incorporation." The overlapping of economic and communal divisions arising from this settlement pattern was given political expression by the colonial state which imposed an ethnically based political hierarchy on the region. The area was declared an emirate in 1903, and Lafia town's Kamberi leader was proclaimed the emir and designated a 'second-class' chief. The traditional rulers of the new emirate's constituent districts of Keana, Assiakio and Doma were in turn proclaimed district heads and designated 'third-class' chiefs. Because of the presence of small groups of Islamic converts among the Alago of Assiakio and Doma, the emir was empowered to appoint the heads of both districts. The apportioning of the region into village administrative areas in the 1920s, and the practice of appointing village heads or chiefs from groups which were considered zyxw 50 zyxwvu zy zy D.L . Eyoh indigenous to an area because of early settlement extended this hierarchy to lower levels. The hierarchy continues t o inform the inter-ethnic distribution of power within institutions of political administration and to influence the manner in which regional groups are linked to national structures of class and power. The coterie of individuals at the top of the regional social hierarchy - in effect the regional representatives of the national ruling class, whose members derive their economic and political status from commerce, public sector contracting and high positions in public bureaucracy - is comprised primarily of Kamberis and, secondarily, Alagos and Migilis.12 In the case of traditional authority structures, the emirship of Lafia, for example, whose occupant is both the Lafia district head and nominal leader of Muslims across the region, is the most powerful ‘traditional’ political position in the region. The Kamberis in Lafia town still supply emirs, and the district’s traditional council which is the last point of arbitration for conflicts not resolvable at the village level, and for decision-making over matters of district-wide concern, is dominated by Kamberis. Although no longer filled by appointees of the emir, the chieftaincies-cum-district headships of Assiakio, Doma and Keana are controlled by Alagos, who also dominate the traditional councils of those districts. The continued reliance on village areas as local units of political administration and the selection of salaried village heads from groups defined as ‘indigenes’ of concerned localities, sustains the replication of this pattern of political power at lower reaches. The imbalance in political power is equally manifest in patterns of representation within ‘modern’ state institutions.’’ Although the particulars of local power configurations vary, the overall effect of the overlapping of economic and ethnic divisions and their expression in institutions of political administration is to form a framework within which the political conditions of differential access to external resources operate in a manner which favours both dominant village strata and members of some communities over o t h e r ~ . Being ’~ the most likely centres of commercial activity and the location of branches of public bureaucracy, large villages and towns are the preferred residence not only of older immigrant groups, but also of the group of farmer-traders and petty bureaucrats w h o are key components of leading social strata in rural Nigeria. Therefore, as arenas of competition for access to public resources shift from hamlets to villages and towns, they become zyxw zyxwvutsr zyxwvutsrqp zy Table 4. Ethnic group Tiv Alago Eggon Migili Average for a!l ethnic groups Value of Crop Production and Sales", by Four Major Ethnic Groups zyxz zyxwvutsrqpon All crops Total Sold Total Sold 310 459 240 794 5 22 684 208 987 29 365 120 6% 1066 892 460 1150 89 I 385 Yams 545 Nores: Value in naira at prevailing local prices. Indicates insignificant value. Source: APMEPU (1982b: 17-18). 21 1 Sorghum Total Sold 214 97 66 66 I33 103 22 16 45 55 Millet Maize Total Sold Total Sold 71 15 61 II 22 8 $ 8 29 25 39 21 30 23 .; 3s * 30 30 g z L 13 3 a' 52 z zyxwvutsr D. L . Eyoh more complex. Social stratification within farming populations becomes more pronounced. Also, the range of social and cultural relations (for example, ties of kinship, religion and shared residence) which are important in cultivating the political networks through which access is mediated, become more varied. Members of more recent immigrant groups who tend to reside in remote hamlets, and who are generally subordinated within the political hierarchy, are at an enormous disadvantage when it comes to manipulating dominant political networks or creating effective alternatives. It would be misleading to infer from the above that political relations in Lafia are primarily founded in communal divisions or that such divisions are the basis of entrenched political cleavages at local and regional levels. It would also be misleading to intimate that LADP extension staff were prone to demonstrating a bias in resource allocation in favour of members of particular ethnic communities. What is most relevant to understanding the political dynamics of unequal access in Lafia is not the cultural biases of public officials, but the constitution of local power structures through which their relations with local communities are mediated. The manner in which these structures are shaped by the interpenetration of economic and communal divisions, combined with the progressive farmer emphasis of ADPs, sufficed to encourage the pattern of resource allocation. zyx 4. VILLAGE STRUCTURES OF POWER AND POLITICS OF ACCESS The relationship between village advisory committees and the composition of beneficiaries of the group farming and demonstration farmer programmes in the two communities of Ruttu and Deddere show clearly how such power arrangements, in conjunction with state-sanctioned frameworks of participation, determine the outcomes of allocational processes. The village communities dealt with here differ in ethnic composition, and are representative of opposite ends of the regional spectrum of commercialization and economic stratification. They were selected to evaluate ( 1 ) the extent to which responses to new production-augmenting technologies were influenced by past histories of commercialization and (2) the impact of socio-economic stratification on the ability of various groups to gain access to preferentially allocated resources. The analysis of the Reforming Peasant Production in Africa zyxw zyxw 53 determinants of responses to the LADP confirmed the influence of these factors on levels of adoption and interest in inputs. However, as borne out by the experience of both villages, membership in the leading socio-economic strata was only one of many conditions relevant to the capacity of village groups to gain access to preferentially allocated resources. In other words, the configurations of local power structures have resulted in access being determined by both socio-economic and communal divisions. ’’ Ruttu Village Community The village community of Ruttu contains eleven hamlets.I6 The settlement from which the community derives its name lies approximately 50 km north-west of Lafia town, and 20 km west of Doma town, the district’s administrative centre as well as the location of the nearest farm service centre (see Figure 1). In the early 1980s the community contained around 700 households or, based on an estimated mean of eight members per household, a total population of between 5000 and 6000. Tivs accounted for slightly over 70 per cent of its households and Alagos and Eggons for an estimated 11 and 8 per cent, respectively. Hausas and a scattering of other minority groups made up the balance (LADP, 1980: 14). The populations of the three largest hamlets - two of which are situated along, and the other within 1 km of, the road linking the main hamlet with Doma town - are ethnically mixed. Although some Tivs live in these three hamlets, the majority of Tiv households are in isolated hamlets at distances of 3-15 km off the Ruttu-Doma road, to which they are connected by bush tracks. The village economy is entirely based on food production. Due to its remote location and, even by regional standards, low population density, the estimated average household farm size of 5 ha is appreciably above the regional average of 3 ha (LADP, 1982b). From all indications, including the scarcity of such visible symbols of consumption as bicycles, motor cycles and houses with corrugated iron roofs, which could be used to infer disparities in monetized wealth, social stratification in Ruttu is far from developed. Given the paucity of non-farm income opportunities, economic inequalities are primarily due to disparities in the size of domestic labour pools and the related ability of households to mobilize additional labour resources. 54 zyxwvutsr zy zy D . L . Eyoh Household farm size, together with the local reputation for proficiency in farming, were supposed to have been the main criteria for selection of the so-called ‘progressive farmers’ by the LADP staff. No complete inventory was undertaken for any village during or after the formal implementation period to determine the distribution of household farm size, the staff responsible mainly relying on information received from village headmen, heads of wards and farmers. Nonetheless, those designated progressive farmers all belonged to the leading strata in their villages. In Ruttu, the ethnic composition of this stratum was in inverse proportion to that of the community as a whole. Interviews with the group of primary informants confirmed the centrality of domestic labour force size to inter-household inequalities. The thirteen project-designated progressive farmers within this group (six Alagos, three Hausas, two Eggons and two Tivs) were all heads of compounds with two or more households, with an average of eleven members and two or more young adult males. The use of casual waged labour, often engaged for only one or two days at a time, appeared to be common, if not as widespread as in other areas, although only three ‘progressive farmers’ admitted to having engaged ‘young men’ to help prepare fields for planting since the project commenced. In contrast to Deddere, I encountered no farmers in Ruttu who worked for wages on a regular basis. All the primary informants drew on labour resources of households within their compounds; nine depended regularly on exchange labour arrangements involving age-based groups formed with members of compounds within and beyond their own hamlets, and three (including two of the three who used waged labour) had mobilized work parties with more than ten persons during the previous three seasons.” While most informants admitted that farm land could be secured through cash payments and short-term leases, wide-ranging enquiries revealed no instance of such transactions in the years since the project had begun. All informants emphasized that the ready availability of land through customary usufruct accounted for the rarity of such transactions. The organization of the regional food trade had some effect on stratification, but the numbers of those involved in Ruttu were considerably less than in other villages (see note 11). The three major farmer-traders (two Hausas and one Alago) in the community were project-designated progressive farmers and functioned as intermediaries of traders in Lafia town. Two of the farmer-traders were zyx zy zyxw zyxw Reforming Peasant Production in Africa 55 among those who admitted use of waged labour and work parties. As in other multiethnic settlements, members of Ruttu’s constituent communities enjoy a large degree of autonomy in the organization of their political affairs. This includes the choice of heads of hamlets (outer unguwas) or wards (inner unguwas) within large settlements whose residents are drawn predominantly from the same ethnic group. The village traditional council which is composed of community representatives and elders of Alago lineage is the main public forum for deliberating village-wide affairs. But despite the seeming egalitarian nature of representation, the distribution of power between communities is characterized by an imbalance which broadly parallels the ethnic colouring of the dominant stratum. By virtue of their claim as original inhabitants of the area, Alagos remain in control of the salaried position of village head and of local representation in the Alago-dominated Doma district traditional council. The composition of Ruttu’s Village Advisory Committee and participants in the group farming and demonstration farmer programmes bore the imprints of these relationships. Ruttu’s Village Advisory Committee was constituted in 1979 through a process which began with an information session for farmers with project staff. At a subsequent gathering, which was called t o organize the committee and select its executive, the local extension agent suggested that the committee be formed by representatives from all hamlets. The result was a fairly large forum of over twenty-five, but one which allowed for the representation of constituent village groups. But the eight-member executive of the Village Advisory Committee (five of whom were from Ruttu, two from Mudu and one from Aragye) failed to replicate the level of communal representation of the VAC. It was composed of five Alagos, two Hausas and one Mopun. Most informants agreed that the choice of executives and, in particular, the designation of chief of farmers (Sarkin Norna)18as chair and the inclusion of the village head, was popularly agreed upon by farmers present at the meeting when the committee was formed. Most executive committee members interviewed readily conceded that they had solicited the support of several farmers prior to the meeting, which was attended mostly by farmers from the three largest settlements. Their interest was based on assessments of the potential benefits of membership gained in part from official descriptions of the functions of the commit tee. 56 D. L. Eyoh Between 1980 and 1984, five project-designated progressive farmers participated in the demonstration farm programme. Three of the demonstration farms were in Ruttu and the other two in Mudu, which is situated about 2 km from Ruttu. In 1981, two farmers groups, each with ten members devoted to the cultivation of paddy rice on ten acres, were formed in Ruttu. A third group with twelve members was organized in 1982 to cultivate a 16-ha mixture of maize and cassava in Mudu. Two of the demonstration farmers in Ruttu were Alago and the third, Hausa. In Mudu, one was Alago and the other Hausa. Two of Ruttu’s demonstration farmers (an Alago and the Hausa farmer-trader) were also the respective leaders of the Ruttu groups as was the case in Mudu. The first Ruttu group had seven Alagos, two Hausas and one Eggon. With three of its members also belonging to the first group, the second Ruttu group had six Alagos, three Hausas, two Eggons and one Tiv. The Mudu group had three Hausas, four Alagos, one Eggon and two Mopuns. Contrary to the formal criteria stipulated by the project for determining eligibility for credit programmes, the Village Advisory Committee was not collectively consulted in the process of selecting participants for both programmes. Primary responsibility for the choice of beneficiaries of both was confined to members of the executive, acting in concert with the local extension agent. The leader of the first Ruttu group was chairman of the Village Advisory Committee, the leader of the second a member of the executive, and the leader of the Mudu group the deputy chairman. According to all three, and other informants, they had been told by the local extension agent - as members of the Village Advisory Committee’s executive - of the project’s intention to have groups organized in the area, and had ‘volunteered’ to be group leaders. Upon securing their mandates, group leaders approached or were contacted by interested farmers. The final list of participants was determined in consultation with the extension agent. All group leaders claimed that they chose ‘hard-working progressive farmers’ who could be trusted as they had been informed that they would be held ultimately responsible for ensuring prompt repayment of loans made to groups. The end results of this narrowing of the arena of participation was that the beneficiaries of both programmes reflected the village power structure. zyxwvu Reforming Peasant Production in Africa zyx zyx zy 57 Deddere Village Community Deddere village community is an agglomeration of sixteen settlements, subdivided into six administrative wards.’’ The largest nuclear settlement, from which the community’s name is derived and which was the location of a farm service centre, lies about 20 km south-west of Lafia town (see Figure 1). In the early 1980s the community contained about 1400 farm families, or, based on an estimated mean of six members per farm family, between 8000 and 9OOO people. The population was characterized by a high degree of ethnic heterogeneity. Migilis, the original settlers of the area, and Eggons each represented approximately 24 per cent of households. Gwandaras and Tivs accounted for roughly 13 and 15 per cent respectively, and Hausas, Alagos and Kamberis for a combined 14 per cent of households. The balance was made up of members of numerically insignificant groups (LADP, 1980). The Migili, Alago and Hausa populations were concentrated within demarcated areas in the two largest settlements, although some members of virtually all ethnic groups also lived there. Tivs and Eggons lived mainly in hamlets situated at some distance from the main road traversing the community. Deddere’s average household farm size of 3.6 ha was slightly above the regional mean. Its position on the all-season Lafia-Obi road, which has been in existence since early colonial times, ensured favourable access to external markets. Because of this, and its close proximity to Lafia town, Deddere was among the leading communities in terms of the commercialization of food production. From the obvious evidence of consumer goods such as previously mentioned, and indeed the presence of at least three small buses used for commercial transportation in the main settlement and in Kadunatana, the community as a whole was visibly far more affluent than Ruttu. Its population was also more internally stratified along economic lines. Interviews with the group of primary informants, among whom were sixteen project-designated farmers, confirmed a higher intensity of commercialization and a greater impact on stratification. Random audits in several hamlets supported the official estimates of an average of six members per household: all progressive farmers among the primary informants were heads of households with an average of nine members. While exchange labour arrangements were the main source of supplementary agricultural labour, there 58 zyxwvutsrq zy D.L. Eyoh was also a much more widespread reliance on casual waged labour than had been observed in Ruttu. The demand for such labour was concentrated in the two largest settlements, which attracted young men from Tiv and Eggon hamlets seeking farm work on an intermittent basis. The higher incidence of casual waged labour is related to the greater degree of stratification in Deddere’s farming population. Nine progressive farmers among the group of primary informants admitted regular use of hired labour for a variety of farm tasks over a number of years, and six admitted hiring casual labour occasionally due to temporary depletions in their domestic labour pools. While most of those who worked for wages claimed that they did so intermittently and without detriment to their farm activities, there were cases in which participation in wage labour seemed to be a necessary condition for household subsistence. In contrast to Ruttu where no such case was observed, three young married men encountered in the main village had been working for wages over a number of years. All claimed that they were obliged to do so because they were without adequate farmland and could not as yet afford to expand their farm holdings. The transfer of land rights through short-term lease arrangements (aro)and direct sales were readily acknowledged to be long-standing local practices. A number of related factors account for the higher incidence of ‘commercial’ land transfers in Deddere. The community is situated within the most densely settled area in the region and the pressures on farmland are therefore greater. As most of the land in the immediate vicinity of the large settlements is likely to be permanently claimed and under continuous cultivation, for most farmers acquisition of new holdings through customary tenurial relations often entails greater commuting distances between residences and fields. Households with more financial resources are at an obvious advantage in securing additional land in the vicinity of large settlements. The influence of these factors was confirmed in interviews with farmers, among them a young Migili former employee of Lafia local government council and secretary of Deddere’s VAC, and a prosperous Hausa farmer-trader, who bought and leased adjacent fields in the vicinity of the village so as to take advantage of tractor services offered by the project. The varying intensity of involvement in commercialized production and in the regional food trade between ethnic communities was similarly more pronounced in Deddere. For example, while all zyxwv Reforming Peasant Production in Africa zy zyx zy 59 communities were represented among food traders in the village’s main market, Migili and Alago females were the major sellers of yam (the region’s most important market crop). Migili women in particular travelled distances in excess of 10 km to sell yams at markets along the interstate highway crossing the region, where they could get higher prices. The sale of food on discrete plots before harvest is relatively widespread in communities near Lafia town, but was not found at all in Ruttu. This is another indication of the variations in commitment to commercialized production between areas and ethnic communities. In Deddere and other places, yam was the main commodity involved in transactions of this nature. All traders encountered w h o bought as well as farmers who sold food by this rnethod claimed that those primarily involved in this practice were Migilis and Alagos. None of the Tiv farmers interviewed admitted engaging in transactions of this type, which they commonly characterized as signifying ‘a hunger for money at the expense of feeding ones’s family’. Finally, participation in the regional food commerce had a more profound impact on stratification. Of the four leading farmer-traders in the community (two Hausas, a Migili and a Kamberi), three were independent traders and one a client of a Lafia town trader. The three independent traders were the owners of the commercial vehicles already mentioned. The organizational basis of political authority in Deddere is broadly similar to that in Ruttu: leadership is formally vested in unguwa heads, and Migili lineage heads and leaders of major communities sit on the village traditional council. By virtue of their claim as original settlers of the area, Migilis continue to occupy the inherited and salaried position of village head and to provide local representation in the Lafia district traditional council. While control of these crucial positions might suggest that Migilis are the best placed group in terms of political administration and in intercommunity political relations, the situation is more complex. A part of Lafia district, the community remains politically wedded to the Kamberi-dominated power hierarchy which is centred on Lafia town. As demonstrated by attendance at the mosque in the main settlement on Fridays, Islam was the major religion, with practitioners from long-established Islamic groups such as the Hausa and Kamberi, as well as relatively recent converts from the Migili and Alago Communities. Thus, despite the institutional arrangements which appear to favour Migilis, the village social hierarchy and the political networks of access to public resources are more strongly 60 zyxwvuz D. L . Eyoh defined by an interpenetration of economic divisions with religious and other communal affinities. Although the underlying dynamics varied, Deddere’s Village Advisory Committee also emerged from a process which began with the convening of an information session for all interested farmers. This was followed by a second general meeting which was attended by extension agents assigned to the FSC and the LADP’s district development officer for Lafia, during which the VAC and its executive were selected. Positions on the fifteen-member committee with a six to seven member executive, which the district development officer instructed were to be filled by progressive farmers, were decided through an open contest involving rival candidates. Reflecting the patterns of population distribution and economic and political stratification, the resulting committee was characterized by a restrictiveness both in terms of the number of hamlets from which members originated, and its ethnic mix. Ten of its members were from Deddere, four from Kadunatana and one from Gidan Raisana. Seven members of the committee were Migili, three Hausa, two Alago, two Kamberi and one Eggon. The executive, with four members from Deddere and two from Kadunatana, was comprised of three Migilis, two Hausas and one Alago. Once again, the composition of beneficiaries of both group farming and demonstration farmer programmes also bore the imprints of the local power structures evident in the constitution of Deddere’s VAC. Between 1980 and 1983, five farmers took part in the demonstration farm programme, while four ten-member farming groups three devoted to cultivation of 10 ha of maize/sorghum and one to maize/cassava - were also organized. Three of these groups were located in Deddere and the other in Kadunatana. Of the five demonstration farmers, two were in Deddere, t w o in Kadunatana and the fifth in Gidan Raisana. Just as in Ruttu, the process of selection of participants in both programmes was confined to interactions between members of the VAC executive, local extension agents and the district development officer. All demonstration farmers and group-farm leaders were on the VAC executive. With a Hausa farmer-trader as leader, the first of the Deddere-based groups had two Hausas, six Migilis and two Alagos; the second group which was led by the Surkin Noma of Deddere, a Migili, had seven Migilis and three Alagos; and the third, also headed by a Migili, had six Migilis, two Hausas and two Kamberis. The sole Kadunatana group, whose leader was an Alago, had five Alagos, zyxwv zy zyxw zy zyx zyxw Reforming Peasant Production in Africa 61 three Migilis, one Eggon and one Kwandara. With the exception of the Gidan Raisana case, all Deddere area demonstration farmers were also members of farming groups. CONCLUSION This article has sought to illustrate the manner in which regional and local power structures which are constituted by multiple dimensions, impinge on the conditions of access to resources distributed through state channels and, by so doing, enhance economic opportunities available for some groups at the expense of others. In the case of the Lafia project, the allocation biases could not be justified by claims that those farmers deemed ‘progressive’ had a superior capacity or potential for innovation with new technologies of production. In fact, even if the reverse were true, the constitution of local power structures through which access to resources was mediated ensured that many who could be or were so classified stood to be excluded from access because of their community backgrounds. Those engaged in debates about the parameters of agricultural reform in sub-Saharan Africa tend to agree that the urban-centred universe of state power has worked to the detriment of national economies by narrowing the range of incentives available to agricultural producers. The nurturing of broad-based coalitions of agricultural producers is thus essential for creating and sustaining political contexts which are conducive to increasing incentives for producers. This perspective permeates micro-societal level studies of the politics of agricultural change, which conclude by endorsing what has been termed a modified version of ‘trickle down’. The need to enhance the power of agricultural producers is not in dispute here. The plea of this article is that such advocacy be tempered by a greater sensitivity to the fact that different classes of agricultural producers have convergent as well as divergent interests, and that their capacities to secure access to incentives are not uniform. An appreciation of how economic and cultural divisions are expressed in local power structures, and how these inform the networks through which rural groups are linked to external structures of class and power, is therefore crucial to understanding the importance of political context in determining which groups gain or are excluded from incentives. This is all the more important when D.L. Eyoh the organizational mechanisms through which incentives are distributed are subject to the play of political forces and power relations in rural society. I n Africa, differential access to productive resources channelled through state institutions has been a key dynamic in rural economic differentiation. The prospect of introducing processes of industrialization which are capable of absorbing large streams of marginalized rural groups is not, for the foreseeable future, a viable option for most African states. They can ill afford processes of agricultural change which intensify rural inequalities. This, as well as the issues raised in the introduction, should prompt social scientists to pay closer attention to how the complexities of rural social divisions and power relations promote allocational inequities which impede the ability of various rural groups to realize their productive potential. z zy zyxwvut zyxwvut zy zyxw NOTES I . See Eyoh (1989) for full details of the case study which was based on information derived from records held at the Nigerian National Archives (Kaduna branch) and data from field work in the region between May 1985 and July 1986. The main sources of information during field work were: ( I ) official documents that were available at the project's head office in Lafia town, at the Jos headquarters of the Plateau State Agricultural Development Project which superseded the LADP, and at the Federal Department of Rural Development's Kaduna-based Agricultural Project Monitoring, Evaluation and Planning Unit; (2) a socio-economic survey involving 240 randomly selected farmers; (3) observation of project activities across the region; and (4) extensive interviews with project personnel, farmers and local political leaders. 2 . Local political-administrative boundaries were redrawn on a number of occasions between 1977 and 1984. The two local government areas contain six administrative district$ (Assiakio, Awe, Doma, Keana, Lafia and Obi) which, prior to January 1984, were all local government areas. 'The project retained the districts as its administrative units, and much of the data pertaining to its activities are district based. 3 . A typical residential compound in the region contains more than one household, and households co-operate on a wide variety of economic tasks. Except where otherwise noted, a household or farm family (both concepts are used interchangeably) here represents persons who work regularly together. share a common residence and have their main meal from the same pot. 4. Most regional groups are referred to by more than one name; the more popular names are used here to minimize confusion. 5 . The project relied on fami size to distinguish between classes of farmers. Farm holdings of under 5 ha were classified as small-scale or ordinary; between 5 and 10 ha were judged mediurn-scale or progressive; and those with more than 10 ha were Reforming Peasant Production in Africa z zyxw 63 zyxwvu zyxw classified as large-scale. The estimated number of large-scale farmers in the project ranged from ten to forty in different survey years. Most of these were recognized to be absentee claimants of large tracts of uncultivated land whose main economic interests lay outside agriculture - in commerce and public sector contracting. In the final analysis, extension personnel found it difficult to distinguish between smalland medium-scale farmers. 6. The theoretical significance of these observations is elaborated in Eyoh ( 1992). 7. The character of socio-economic and political changes under colonialism and social relations governing agricultural production in contemporary Lafia are described in detail in Eyoh (1989: Chs 3 and 4). 8. Data in Tables 2, 3 and 4 are derived from a region-wide random sample of 218 households. 9. The superior efficiency of mixed cropping in serving the needs of the household and in forming the economic basis for rejection of sole-crop packages was strikingly obvious. For example, traditional yield levels for sorghum/millet mixtures were found to be 25 and 35 per cent higher than estimated yields for sole-cropped sorghum. Even setting aside the labour-saving advantages. mixed cropping was more profitable in terms of market value; traditional per hectare yields of grains grown in various mixtures were estimated to be at least 50 per cent higher than sole-cropped grains and worth 140-260 naira more at prevailing prices. The market value of y a d m a i z e and yam/maize/sorghum mixtures was 120 and 275 naira above solecropped yams (for more detailed evidence see APMEPU, 1982a; Vol. I , p. 22; Eyoh, 1989: Ch. 7). 10. The distribution of ethnic communities by settlement type persists to the present. In Doma and Keana districts, which are predominantly Alago and Tiv, the Alago population is concentrated in the towns of Alago and Tiv and in a number of large, roadside villages within close proximity. Outside these towns and villages, the population becomes predominantly Tiv. In Assiakio district. Migilis and Alagos (who are in the numerical minority) are concentrated in Assiakio town and in villages which are strategically located along the interstate highway link between Lafia town and Kaduna, whereas Eggons (who are numerically preponderant) reside mainly in remoter hamlets. In Lafia district, Kamberis are concentrated in Lafia town and a few villages within a 15-km radius of it, and the preponderant Eggons in hamlets further off the main interstate highway. I I . The relationship between settlement patterns, market access and agricultural cash incomes was recognized by colonial officials and accounted for Tivs being subject to a much lower per capita tax rate during most of the colonial period. The social basis of organization of the regional food trade has also contributed to this pattern. While food marketing is undertaken by a variety of agents, the greatest returns accrue to those engaged in bulk purchasing, especially for export outside the region. This dimension of the food trade involves intricate patron-client based trading networks which are centred in Lafia town and are dominated by Kamberis and Hausas. and secondarily. lgbos who are part of southern trading networks. Locally resident Hausas and individuals from other ‘Islamized’ communities such as Migilis and Alagos have predominated historically as intermediaries in trading networks and thus within the ranks of farm-traders who are a key component of dominant village strata (Eyoh, 1989: Ch. 4). 12. In a blatant act of political patronage intended, in the words of a senior project 64 zy z zyxwvuts zyxwvut zyxwvut . D. L Eyoh zyx zyxwvut manager, ‘to secure the goodwill of men of calibre’, all credit disbursed during the project’s first year went t o fifty individuals reputed to be the most influential in the region. A scrutiny of the list of beneficiaries of credit packages, which ranged in value from 400 to over SO00 naira, revealed that at least forty-three were Kamberi, Hausa, Alago and Migili. Only one could be readily identified as Tiv. 13. Popularly elected local government councils were disbanded following the coup &Plat which toppled Nigeria’s so-called second republic in 1984. After the last national elections in 1983, for example, ten of the twelve members elected to the Lafia local government council were Kamberi and/or Muslims. In the case of the Doma local government council, eight of the elected councillors were Alago and only two belonged to the numerically preponderant Tiv community in the district. 14. I t needs t o be emphasized that the advantages of access to public resources do not apply only t o incumbents of these positions but are also communal. The political effectiveness of chiefs in multiethnic communities requires the maintenance of cordial relations with all constituent groups. As was confirmed in several locations, decisions on common concerns, such as organizing labour for communal roads or a school, are arrived at through consultations with representatives of all groups. However, as decisions over incumbency t o these positions are in the hands of so-called ‘indigenes’ of a given area, in cases where external political demands and resources are involved, the common practice is for village heads to first consult with lineage heads from their primary group and other prominent village personalities. The effect is that members of the village heads’ primary group are generally at an advantage in terms of information about the availability of resources and, as shown below, in terms of the social and political networks which assure access to these resources. 15. See Eyoh (1989: Ch. 8) for a more elaborate analysis of what follows. The bulk of information used t o reconstruct project-related political activity was derived from primary informant groups of twenty farmers in Ruttu and twenty-five in Deddere, who consented t o be interviewed o n two or more occasions. 16. Ruttu Village Community Population Wards Ruttu Tarko Tiv Gajeri Sabo Dogon Kurmi Mudu Gidan Allu Unguwa Giwa Aragye Kogon Lemu Unguwa Tiv Total Hamlets Compounds Households I 93 25 3 84 23 164 I I I I I so 5 7 I I5 73 23 6 2 I I12 33 26 171 42 60 I1 419 709 1 I I5 Source: LADP (1981) 17. This last source of non-household compound labour is of special significance, as the ability to mobilize large work parties depends o n the organizer’s ability to Reforming Peasant Production in Africa zy zyx zy 65 provide large quantities of food and beer. Most work parties are organized at the beginning of the planting season when supplies of food and grain needed for brewing are low; thus households which can afford the necessary largess are among those most likely to produce substantially above their domestic requirements and therefore be in a position to market a large amount (in volume, if not in relative share) of their total output. 18. The Sarkin Noma is a symbolic position which commands a measure of respect in the community and is usually occupied by an elder with a local reputation as a diligent, hard-working farmer. The Sarkin Noma’s main duties include organization of agriculturally related cultural festivals and. although he has no actual authority to adjudicate, he is the first mediator of conflicts over farmland. Sarkin Nomas were prominent in LADP-sponsored farmers’ advisory committees at different levels. This illustrates the ability of village title-holders to convert symbolic authority to political advantage in struggles for external resources. 19. Deddere Village Community Population Wards Deddere Gidan Raisana Osung Ashupe Kadunatana Mutum Daya Total Source: LADP (1982a). zyx zy zyxwv zyxwvu Hamlets Compounds Households I 7 4 2 1 1 592 26 I5 63 I I4 45 898 46 35 I46 243 71 16 855 1439 REFERENCES APMEPU (1982a) ‘Lafia Project Completion Report’, 3 Volumes. Kaduna, Nigeria: Federal Department of Rural Development. APMEPU (1982b) ‘Crop Production and Sales in Lafia ADP: Patterns of Variation by Farm Size Category and Ethnic Group’. Kaduna, Nigeria: Federal Department of Rural Development. APMEPU (1984) ‘Lafia: Agronomic Survey for 1983/84’. Kaduna. Nigeria: Federal Department of Rural Development. Barker, J. (1989) Rural Communities Under Stress: Peasant Farmers and the State in Africa. Cambridge: Cambridge University Press. Bates, R. (1981) States and Markets in Tropical Africa: The Political Basis of Agricultural Policy. Berkeley, CA: University of California Press. Beckman, B. (1987) ‘Public Investment and Agrarian Change in Northern Nigeria’, in M. Watts (ed.) State. Oil and Agriculture in Nigeria, pp. 110-37. Berkeley, CA: Institute of International Studies, University of California. Berry, S. (1984) ‘The Food Crisis and Agrarian Change in Africa’, African Studies Review 29(2): 59-1 12. z zyx zyxwv D.L . Eyoh Bonat, Z. and Abdullahi, Y. (1989) ‘The World Bank, the IMF and Nigeria’s Agricultural and Rural Economy’, in B. Onimode (ed.) The I M F , the World Bank, and fhe African Debf, Vol. 1, pp. 153-76. London: Zed Press for the Institute of African Alternatives. Brett, T. (1986) ‘State Power and Economic Inefficiency: Explaining Political Failure in Africa’, IDS Bulletin 17(1): 22-9. Cliffe, L. (1977) ‘Rural Class Formation in East Africa’, Journal of Peasanf Studies 4(2): 195-224. Clough, P. and Williams, G . (1987) ‘Decoding Berg: The World Bank in Rural Northern Nigeria’, in M. Watts (ed.) Stare, Oil and Agriculture in Nigeria, pp. 168201. Berkeley, CA: Institute of International Studies, University of California. Crook, R. (1988) ‘Farmers and the State’, in D. Rimmer (ed.) Rural Transformafion in Africa, pp. 117-39. London: Belhaven Press. Dutkiewicz, P. and Williams, G. (1987) ‘All The King’s Horses and All the King’s Men Couldn’t put Humpty-Dumpty Together Again’, IDS Bulletin 18(3): 39-44. Eyoh, D. (1989) ‘State, Capital and Agrarian Transformation in Nigeria: Politics and Production in Lafia, Plateau State’, unpublished P h D Thesis, York University, Canada. Eyoh, D. (1992) ‘Structures of Intermediation and Change in African Agriculture’, African Sfudies Review 35( I ) (forthcoming). Holmquist, F. (1984) ‘Self-Help: The State and Peasant Leverage in Kenya’, Africa 54(3): 72-91, LADP (Lafia Agricultural Development Project) (1980) ‘Mid-Term Evaluation Report’, Lafia Town (mimeo). LADP (1981) ‘Water Resources Survey for Doma District’, Lafia Town (mimeo). LADP (1982a) ‘Water Resources Survey for Lafia District’, Lafia Town (mimeo). LADP (1982b) ‘FRADYS Survey for 1982’, Lafia Town (mimeo). Samoff, J . (1980) ‘Underdevelopment and its Grassroots in Africa’, Canadian Journal of African Studies 14(3): 5-36. Wallace, T. (1980) ‘Agricultural Projects and Land in Northern Nigeria’, Review of African Poliiical Economy, 17: 59-70. Williams, G. (l985a) ‘Marketing With and Without Marketing Boards’, Review of African Poliiical Economy 34: 4- 15. Williams, G. (l985b) ‘Taking the Part of Peasants: Rural Development in Nigeria and Tanzania’, in P. Gutkind and I. Wallerstein (eds) The Political Economy of Contemporary Africa, pp. 144-80. Newbury Park, C A and London: Sage. World Bank (1977) Survey Appraisal Report for Lafia Agricultural Developmenf frojecf. Washington, DC: The World Bank. World Bank (1989) Sub-Saharan Africa: From Crisis lo Sustainable Developmenr. Washington, DC: The World Bank. zyx Dickson L. Eyoh is Assistant Professor and Director of the Undergraduate Programme, Division of Social Science, and Co-ordinator of African Studies at York University, 4700 Keele Street, North York, Ontario M3J 1P3, Canada.