The contemporary crisis in Africa is having contradictory political effects. On the one hand, the increasing subordination of the continent’s political life to the calculus of factional interests ensures that even cabinet reshuffles become violent affairs, tragi-comic ‘revolutions’. On the other hand, the deepening socio-economic crisis of the African countryside has left ever-widening circles of society without organized political participation or voice. Everywhere is heard the chorus of dissatisfaction and the demand for ‘democracy’. While a number of progressive writers on Africa have recently warned against the attitude that dismisses democracy as a mere ‘bourgeois’ question, peripheral to the struggle over the relations of production,footnote1 it is also time to warn of the opposite extreme—the tendency to eulogize democracy only in the forms conceptualized by the middle classes. By putting a halo around a single demand—that for a multi-party system and fair elections—as the sum and substance of the democratic process, it is possible to make the ‘democratic opening’ so narrow that it would only grant meaningful freedom to rival bourgeois factions, leaving most of the popular classes beyond the pale. To reconstruct this debate from the standpoint of the producing classes, it is necessary to recognize democracy as more than an artifact of formal institutions. For if democracy is to be an activity of meaning to all classes in society, and in particular to the peasantry and working classes, then its form and scope must indeed be meaningfully related to their living conditions. In other words, a discussion of the political form of the state can not be divorced from an analysis of existing production relations or systems of exploitation. In the contemporary African context this means linking the question of democracy, first and foremost, to an analysis of the agrarian question—because Africa remains, preeminently, a ‘continent of peasantries’. While this essay will consider different perspectives on the relationship between the peasantry and democracy in Uganda, I hope its general significance to the broader African scene will be apparent.

Since the 1960s a substantial literature has examined labour reserve economies in Southern Africa.footnote2 Seeking to explain the crisis of peasant small commodity production, these analyses agree that this crisis can not be understood without focusing on the role of peasant households as labour reserves for large-scale capitalist enterprises. If this approach has had the merit of emphasizing the relations of production, the same unfortunately cannot be said of most of the literature on peasant small commodity production in Africa outside the labour reserve economies. Writings on this subject have been characterized by two predominant tendencies. The main tendency has involved a shift of focus away from relations of production to relations of exchange. Lines of debate have been drawn within these shared parameters: is the crisis primarily the result of ‘external’ exchange relations (adverse terms of international trade leading to unequal exchange and flow of value to imperialist centres) or of ‘internal’ exchange relations (unfavourable terms of trade between agriculture and industry reinforced by adverse and overvalued exchange rates)? On one side stand the proponents of dependency theory, on the other the technocrats of the World Bank and the imf.footnote3

In contrast, the second tendency of analysis—best represented by the writings of Goran Hydenfootnote4—does indeed begin with an investigation of relations of peasant production, but is heavily coloured by results of previous discussions of the agrarian question in Asia and Latin America, where agrarian relations are synonymous with landlord-tenant relations. This is why, as he confronts the Africa of small commodity producers and finds that except in small patches (the largest being Ethiopia) the peasant in Africa does not confront an immediate social overlord, and that land here is not a scarce commodity, Hyden is led to conclude that the African peasant is free, unexploited, and ‘uncaptured’. In his futile search for a Latino-Asian-type reality in Africa, Hyden even fails to identify the relations that do obtain, let alone analyse them.

To consider the agrarian question to be synonymous with the land question is to generalize from one concrete historical situation to others. It can only characterize new historical experiences through formal analogies with the ‘original’ experience—that here the peasant confronts no social overlord, and thus there is no land question—as Hyden understands the Africa of small commodity producers by formal analogy with the Latino-Asian experience. To go beyond the limitation of these two tendencies—one that is confined to the analysis of exchange relations, and another that tries to investigate production relations through formal analogies with an earlier historical experience—it is necessary to pose a single question: Why do peasants enter into unequal, exploitative relations? The answer leads one to recognize the variety of production conditions faced by peasant households.

The unequal relations that peasants enter into have a dual character about them. One set of relations are the product of social relations obtaining in the villages. They are entered into because of the force of objective circumstances. In the sense that there is an absence of any direct compulsion, and only in this sense, these relations can be said to be entered into ‘voluntarily’. Their starting point is the unequal position of peasant households and their result is the deepening of the differentiation internal to the peasantry. What are these objective compulsions that peasants face? Let us consider the Ugandan case.

In the Ugandan situation, broadly speaking, one can identify two different contexts: where landlordism combines with an acute land shortage, and where the peasant confronts neither an immediate overlord nor an acute shortage of land. The main location of landlord-tenant relations in contemporary Uganda of the first type is Buganda. The history of these relations predates the 1900 Agreement that restructured pre-colonial landlordism by granting miles of land to a class of British-connected landlords. A tenants’ movement in the 1920s compelled the colonial state to introduce rent control legislation, the 1928 Busulu and Envujjo Law, setting a limit on both the ground rent (obussulu) and the commodity rent (envujjo) a landlord could demand from his tenant. Since the law only covered existing tenancies, the passage of time loosened its effect as population increased and new lands were brought under cultivation, which landlords could sell or utilize as they wished. It was, however, the 1975 Land Reform Decree of the Amin regime which gave a new lease of life to landlordism in Buganda. By abolishing the 1928 Busulu and Envujjo law, and by turning the landlords into lessees of the state and the tenants into sub-lessees, it removed the protection of the law from the tenant, and left rent in each case to be worked out by the push and pull of class relations on the ground. The effect of this history can be seen on the distribution of land in one village in Buganda in 1983. The data have been culled from a larger study on the agrarian question in contemporary Uganda.footnote5

It is clear that the rural poor in Kitende enter into exploitative relations because of one predominant fact: they are land-poor. Production conditions, however, in the bulk of the Ugandan countryside are quite different. Here, there has been no history of landlordism. In the colonial period, land was relatively plentiful, the population relatively sparse, and access to it governed by customary tenure. And yet, in spite of the absence of a land shortage, a differentiation can be observed amongst the peasantry, leading the rural poor to enter into exploitative relations ‘voluntarily’. To understand the reasons behind this, it is necessary to go beyond the formal ownership of the land to the capacity of each stratum of the peasantry to bring the land under cultivation. The table below, taken from the same study, summarizes production conditions obtaining in a village in the northern part of Uganda.