Emancipatory Politics: A Critique
edited by Stephan Feuchtwang and Alpa Shah

Previous Chapter / TOC / Next Chapter

Download as PDF

Part 3: Armed Movements and State Formation in Nepal, Nicaragua and Mozambique

The Politics of Production, Frelimo and Socialist Agrarian Strategy in Mozambique

Bridget O’Laughlin


Socialist revolution is not just a single moment when power is defied and defeated; it must be a continuing process of class struggle to change both the forces and relations of production. In Mozambique, Frelimo rejected any strategic redistribution of land to the peasantry from settler farms, estates and plantations and steered both investment and consumer goods towards industry, urban areas and state-farms. Its management of the farms – patterns of cropping, the organisation of work, labour recruitment and stratified wage-forms – and its gendered assumptions about the subsistence orientation of the peasantry mirrored those of the class relations typical of the colonial period. The farms were a failure in economic terms and undercut the political legitimacy of Frelimo in the countryside. They also contributed to private accumulation of capital in parallel commercial networks and to patterns of debt that opened Mozambique to the liberalisation policies of the IFIs.

Introduction: Revolution and Millenarianism

The social revolution of the nineteenth century cannot take its poetry from the past but only from the future. It cannot begin with itself before it has stripped away all superstition about the past. The former revolutions required recollections of past world history in order to smother their own content. The revolution of the nineteenth century must let the dead bury their dead in order to arrive at its own content. There the phrase went beyond the content – here the content goes beyond the phrase (Marx 1852).

This much cited passage from Marx’s analysis of the fall of the French Republic in 1851 (The 18th Brumaire of Louis Bonaparte) captures the millenarian expectations arising from Marx’s theory of socialist revolution. All should change; all is possible; the past is past. There is much here to encourage revolutionaries to think that socialist revolution is that moment when they have seized state power and can set out to build a new and better world.

There was a millenarian cast to everyday life in Mozambique in the late 1970s. The Mozambique Liberation Front (FRELIMO, from Frente de Libertação de Moçambique) formed the first independent government in 1975 after a prolonged armed struggle and declared itself a vanguard Marxist-Leninist party in 1977. Many felt that they were living in revolutionary times, a moment when people could do, say and imagine things that had been impossible only a few years before. Dockers changed out of their work clothes to attend primary school classes in the middle of the day. Young students went to the countryside for literacy or vaccination campaigns. Manual workers, clerks and managers, doctors, nurses and patients called each other comrade. Neighbours got together to clean the rubbish from the streets. Artists painted images of workers and peasants on the walls of the city. Production councils kept rural cotton gins and urban factories running even though the owners and managers had left. Women, rural and urban, stood up in meetings and criticised those in charge (mainly men) for not listening to them. Ideas, big and small, were borrowed from different socialist traditions: betterment campaigns (like killing flies) from Mao, dynamising groups from Cuban neighbourhood committees, choreographed May Days from North Koreans, women driving buses and tractors from the Soviets. The president, Samora Machel, would announce a new political slogan (palavra de ordem) in a broadcast speech, and all over the country people would offer different interpretations of what it meant, and act upon them in sometimes contradictory ways.

Yet a teleological vision that reduces socialist revolution to such transcendent moments would belie the importance that Marx himself assigned to the primacy of production and thus to socialism as a process of transforming both the forces and relations of production. This can only be a process, not a single moment, and it is a process that is contested and contingent, its course and outcome unknown. Because forces and relations of production are a contradictory and mutually constitutive unity, neither the seizure of state-power nor the physical flight and elimination of the bourgeoisie, nor state-ownership of enterprises silence class struggle. Relations of class remain embedded in the ways we work, in how distribution and consumption are organised and in how we understand and organise the politics of emancipatory change.

As Feuchtwang and Shah point out in their introduction to this book, socialist strategies are defined by the emancipatory ends envisioned by revolutionaries. In that sense, Marx wrote that social revolutions must take their poetry not from the past but from the future. But socialist strategies also depend on conceptions of beginning points; they must trace the ways we move from where we are to where we want to go. For socialist revolutionary movements, these beginning points have been historically quite variable, though each was embedded in a world dominated by the dynamics of capitalism as a mode of production. The historical forms of the relation between capital and labour have been diverse and shifting in ways that mattered deeply in the everyday experience of people. They have thus also mattered deeply to the revolutionary movements that navigate in their midst, forging alliances, defining the opposition, broadening the basis of support for revolutionary change. And sometimes revolutionary movements have erred in their analysis of their beginning points, in ways that profoundly and tragically shaped their strategies of socialist transformation.

This chapter deals with one such movement, the Frelimo Party of the 1970s, its analysis of Mozambican class structure, and the consequences of this analysis for how it conceived a socialist strategy for the countryside,[1] a strategy that concentrated resources, not just investment but also consumer goods, in state-farms. The chapter reconsiders the question that Frelimo could never satisfactorily answer – why were these state-farms not able to recruit the workers they needed when they were needed? It uses this very specific focus to interrogate both how Frelimo’s emancipatory strategy envisioned its specific beginning points and the underlying assumptions it shared with many other socialist revolutionary movements.

There are few arguments to be made for dissecting the experience of Mozambican state-farms on economic grounds: they accumulated huge debts, both to the state financial institutions that extended almost endless credit and to their workers, who in the last few years seldom received their wages. But state-farms have figured in the strategies of agrarian transformation for many socialist movements. Their version of worker ownership should thus be interrogated as part of the critique that revolutionary movements must make of past strategies. If socialist revolution is a contingent process, not something to be simply read off the right recipe, what can we now say about the class politics of agricultural enterprises in the post-independence period? For Mozambique this means reviewing rural political economy at the end of the colonial period, discussing Frelimo’s socialist vision of the countryside, how it framed conceptions of what was to be done, and considering how class relations and struggles shaped the outcomes of its strategy.

The Theorisation of Strategies of Socialist Transition

Much of the classic theorisation of socialist strategy posits a linear series of steps that takes as a universal beginning point 19th century experiences of rapid industrialisation, wide proletarianisation and the emergence of well-organised urban working-class movements. Post-socialist critics are right to point out that this is also the basis of modernisation theories of development and that a kind of teleological vision of historical inevitability often infuses both traditions. There are, however, major differences between them. Marx insisted that the rapid pace of economic expansion and commodification that has characterised the capitalist world is not the product of inexorable adaptive efficiency or universal human progress. What drives change is capital’s life or death competition for profit and labour’s unending struggle against exploitation. Underlying the seemingly balanced terms of market expansion are violent historical processes of dispossession and proletarianisation.

Marx’s theory describes the general tendencies of capitalism, but these are dynamic structural possibilities, contingent moments, not predictable outcomes. The exercise of collective agency, embedded in particular histories, but more than the summation of individual conceptions and desires, shapes these outcomes. Revolutionary moments will always be constructed though understanding particular cultural and historical experiences and lived through local struggles, but they also transform these experiences in collective action.

There is nothing in Marxist theory that precludes recognising that capital may rely on non-free or non-commodified labour. The first was frequent in the late 19th and the first half of the 20th century when colonial power underlay forced cropping and labour regimes that were deliberately embraced to drive down the cost of labour and to discourage labour militancy. The second continues to be embedded in the notion of a sphere of ‘domestic labour,’ tasks that are necessary but labelled as unproductive because what is not bought and sold cannot directly produce profit. Some Marxist analysis has universalised the conceptual distinction between productive and unproductive labour whereas it actually applies only from point of view of capital. This naturalises a historically constructed and hierarchical division of labour in which women have done much of this ‘unproductive’ work.

Mechanical developmentalism has characterised some Marxist approaches to socialist strategy, illustrated in an extreme form by manuals on socialising production with formulaic recipes such as ‘the law of mechanisation.’ It has also established a causal relationship between property and efficiency reminiscent of neo-liberal argumentation, though in an inverted form: suppression of private property and the market allows productivity to be increased through planning of material balances that replaces market signals with norms of efficiency and the setting of quantitative production targets. Protest against low wages, the intensification of work and monopoly pricing, defined as anti-capitalist struggle in the pre-revolutionary period, was redefined as a manifestation of counter-revolutionary attitudes when property belonged to the socialist state.

The central role assigned to state-farms in classic socialist strategies of agrarian transition reflected a focus on the immediate suppression of commodity production, the leading role of the proletariat and the suspension of class struggle. The lacunae in this strategy made it particularly inappropriate in Mozambique, but arguably it would also confront difficulties in the contemporary world of precarious employment and uneven commodification of work.

There has, however, always been a Marxist alternative to linear developmentalism. Just as the development of capitalism is contingent and historically variable so is socialist revolution a contingent process rather than a moment in which all is irrevocably transformed. Class struggle continues and workers’ collective agency is as important as the efficiency of planning in the socialisation of production. This is the ‘immanent critique’ within Marxism itself discussed by Burawoy (2009: 47). It is present in classic debates about theorisation of the experiences of socialist transition, which suggest that the politics of socialist production are much more complex than the suppression of capitalist ownership and the development of the productive forces. These reflections include extended work by scholars such as Bettelheim (1976), but also the writings of critical ‘participant scholars’ like Bukharin and Chayanov, and texts of Lenin himself on ‘socialist emulation’ and cooperativisation. This body of work forces one to look at the class struggles embedded in the discourses and practices of management, in how labour processes are organised and different groups of producers, both within and between work-places, relate to each other.

Class and the Politics of Rural Production at the End of the Colonial Period

In the political narrative of colonial history written by Frelimo during the years of armed struggle and the first years of independence (Mondlane 1995), it was customary to speak of 500 years of Portuguese colonialism. Though the dynamics of coastal trade, and particularly the slave trade, had enormous impact on the history of eastern and southern Africa during those 500 years, the Portuguese political presence beyond coastal trade settlements was minimal until after the 1878 Congress of Berlin and even then only selectively evident. Direct colonial rule and settlement was most firmly established in southern Mozambique and financed by the export of labour to the mines of South Africa. In central Mozambique, administration and plantation economies were at first organised by concessionary companies. Colonial revenue in northern Mozambique drew principally on taxes on companies trading in commodities produced by peasants, particularly the cheap cotton that provisioned the growth of the Portuguese textile industry from the 1930s through the 1950s.

Colonial Mozambique was divided between two interdependent social spaces. First were the ‘native reserve’ areas where peasants maintained their rights to land, organised smallholder production (sometimes subject to forced cropping) and lived (and struggled) within selectively adapted pre-colonial political institutions. Second were the urban, industrial and plantation areas where they went as temporary alien workers, subject to forced labour regimes and controls on their movement, and excluded from the rights of citizenship extended to settlers. In this first period of effective colonial rule, race – or nation – was almost perfectly congruent with class, though differences in the forms of exploitation established clear regional differences in the dynamics of households.

Gradual economic and political changes in the post-World War II period led, however, to much greater complexity of rural class structure and to a blurring of sharp regional differences. The reasons for these changes were partly due to the restructuring of capital in Mozambique, Portugal and the southern African region, which itself reflected growing and changing forms of opposition to the terms of colonial rule in Africa generally and in Mozambique and the southern African region in particular. Tanzania, Malawi, Zambia, Botswana, Swaziland, and Lesotho all gained political independence in the 1960s and national liberation movements of different historical depth were active in Mozambique, Angola, Namibia, Zimbabwe and South Africa. Attempts to stabilise the basis of white rule included a series of economic and political reforms: incentives for white settlement, including in areas once classified as native reserves, modernisation of productive capacity, including reinforcing industrialisation, freer regimes of labour recruitment, and schemes to develop a class of small specialised black commercial farmers.

In Mozambican rural areas these reforms included the formal abolition of forced labour and forced cropping in 1961 and 1962, administrative experiments with marketing cooperatives for small groups of black farmers in the 1950s and 1960s (Adam 1987), state funding for smallholder Portuguese settler schemes in new areas, with limited participation of black peasant families in the 1960s and 1970s (Hermele 1986), and experimentation with extension of crop credit to black farmers by the cotton and cereal authorities in the 1970s (CEA research Baixo Limpopo, Angonia, Manica). By the 1950s, the Zambezia plantations relied almost entirely on contract labour rather than forced labour (Vail and White 1980). State investment in expansion of the rural transport system facilitated labour movement and flexible recruitment.

Shifts in the nature of urban employment and use of labour controls also had an impact in rural areas. Economic growth in both Rhodesia under UDI and South Africa increased traffic and thus jobs in the ports of Beira and Maputo. With the expansion of fixed employment in urban centres there was a corresponding expansion of informal service jobs. In the 1950s women began to move from the countryside into domestic labour in the cities (Penvenne 1995: 152), despite the fact that this required the authorisation of their chief as well as the male head of household. The loosening of influx controls and diversification of urban employment drew young people from the countryside. Shortage of casual wage-labour in rural areas thus co-existed with urban unemployment, particularly in Maputo. Attempts to recruit agricultural workers from among the urban unemployed were not successful, however (Schaedel 1984).

The construction of the Cahora Bassa dam on the Zambezi and the expansion of coal production in Moatize increased off-farm non-agricultural employment in central Mozambique. Off-farm employment also increased in northern Mozambique with the development of the Nacala port and construction of the rail-line from Malawi. Ironically, the formation of FRELIMO and, in 1964, the launching of armed struggle in the north of the country, also increased off-farm wage-employment in the Pemba port and in construction with the expansion of the internal road-system and the building of military installations. Increased urbanisation and off-farm employment boosted the demand for staple food crops, particularly maize, rice and cassava, the latter almost exclusively a commercial crop of black farmers. At the same time, small and medium Portuguese settler farms continued to expand in areas of prime commercial potential, sometimes with government support for credit and even the provision of labour. In cotton growing areas, some new settler farmers were allowed to subvert the monopsony of the cotton companies by purchasing cotton from surrounding peasant producers and selling it with their own production, paying the company only a ginning fee (CEA research Nampula 1979). Other settlers provided inputs and marketing crops such as potatoes from peasant farmers (CEA research Angonia 1982).

The reshaping of rural life at the end of the colonial period thus reflected competition between different forms of capital and the formal political struggles of liberation movements in Mozambique and more broadly in Africa. At the same time, it reveals the struggles of working people within production and exchange. We know quite a lot about forms of individual resistance to forced labour or cropping – absenteeism or desertion on the plantations, for example, or sabotaging one’s own cotton crop (Isaacman et al. 1980). Forms of resistance propelled proletarianisation and differentiation of the peasantry – embedding livelihoods more tightly in the world of commodities. Escaping the monopsony power of a concessionary cotton company in Nampula often meant cultivating an alternative cash-crop such as sunflowers, or planting more cashew-trees. Resistance to the terms of forced labour often meant signing up for a contract on the mines or plantations. Collective resistance was more limited, given high levels of repression and the exclusion of most black workers from trade unions.[2]

In the last decade of colonial rule, therefore, Mozambican class structure was much more complexly related to the dynamics of the market than the narrative of forced cropping and forced labour torn from rural native reserves might lead one to think. There were greater numbers of permanent and contract workers on farms, mines, and construction, and a larger and more stable black urban population. Incentives were given in terms of credit, differential pricing and labour recruitment to settler farmers that no black commercial farmer received, but there were enclaves of specialised black farmers scattered across all areas of Mozambique, many of them related to chiefs but also using the regular wage income of a household member as a basis for investment and working capital.

Yet movement between rural and urban areas was still controlled by administrative documents – the identity card, tax-receipt and pass. Local administrators and chiefs continued to be involved in forcible recruitment of workers both for settler farmers’ seasonal labour demands and local infrastructural maintenance. Colonial agricultural statistics continued to distinguish on the basis of colour: yellows, whites, Indians, blacks, mixed. Only black workers were distinguished by gender, allowing one to trace the substantial importance of women wage-workers for agriculture in southern Mozambique. The casual labour of boys under 18 years was important in agricultural enterprises across all provinces. Extension workers pursued and beat farmers, particularly women, who were not giving priority to timely weeding of their cotton fields. If we asked in our rural fieldwork when forced labour and cropping ended in the region, the answer was almost invariably ‘with Frelimo.’

The administrative institutions of a bifurcated colonial state created a dualistic political order, as Mamdani (1996) and others have emphasised. But they also promoted an evolutionary ideology and vision that explained the need for continuing oppressive state intervention in labour and agricultural markets as being rooted in the supposed subsistence orientation of the peasantry. For colonial planners, the poverty of the people of rural areas, including their deprivation of schools and health facilities, reflected their primitive backwardness, a condition to be slowly addressed, according to the development plans of the 1950s and 1960s (Planos de Fomento) through economic modernisation. This dualist vision did not fit well with the complex interdependence of rural and urban issues in Mozambique by the 1970s, but it continued to inform the vision that many Mozambican intellectuals had of their own society (O’Laughlin 1996).

Frelimo Strategy for the Socialisation of the Countryside

When in 1974 there was a left-wing coup in Portugal, FRELIMO was internally organised as a coherent party rather than as a front, crucial in its unified approach to the war. It was ready to push for independence, its rhetoric was firmly anti-imperialist, and settlers expected to lose property and privilege. FRELIMO had experience in administrating liberated zones in northern Mozambique, but it had no real experience in collectivising production from which to work (Adam 2006). Land was nationalised at independence.[3] This was closely followed by legal practices (including notaries), medicine, education and rental property, though banks and productive enterprises were not nationalised. As settlers fled, however, the state gradually took on administrative control of a wider range of different kinds of property.

Plantation capital and some settler farmers hung on after Independence, but most settlers, particularly the smaller farmers in the irrigation schemes and new settlements, rapidly abandoned their holdings.[4] Administrative commissions integrating farm workers and Frelimo commanders were formed to keep some of the farms going. Still much of the infrastructures, stock, crops and equipment were lost, some sold off or sabotaged by the departing settlers, sometimes appropriated by local people of influence. Administrative commissions could bring the crops in, but they had neither the capital nor experience to organise a new agricultural campaign.

Although the formation of the state-farms was initially a defensive response to the problems of managing these abandoned farms, settlers deserting their farms and jobs were responding to the angry tone in Frelimo’s criticisms of colonial exploitation and to the popular support, sometimes menacing, they generated. Samora Machel’s speeches and Frelimo documents made it clear that the old bourgeoisie were to go and would not be replaced. A Ministry of Agriculture document characterises the first phase of its activity as one where they knew what they did not want the abandoned farms to become: ‘We wanted to avoid and combat the formation of a national bourgeoisie substituting the former land-owners, and, on the other side we did not want self-management appearing.’[5]

Thus the abandoned farms were first managed directly by the ministry as a series of holdings. Provincial planning offices were then formed to coordinate management of the farms and link with centralised planning of crops, inputs, marketing and investment. The largest units and clusters of settler-farms were consolidated into state-farms. Cooperatives were formed on some of the smaller or isolated settler farms, sometimes by a group of workers who stayed on, sometimes by a group of local residents, and very occasionally by a combination of the two. Some private farmers, often of Indian, Chinese, Portuguese or mixed origin, continued farming in peri-urban areas and in other regions with good market access or infrastructures for commercial agriculture. A few small-scale black Mozambican private farmers managed to appropriate abandoned marginal settler lands which were neither integrated into state-farms nor farmed cooperatively.

Frelimo’s strategic directives on agricultural enterprises were not adopted until the Third Congress in 1977 when FRELIMO the front became Frelimo, the Marxist-Leninist party. Frelimo’s strategy of rural socialisation was formalised in the theses for the Third Party Congress in 1977 and widely debated in the period of preparation for the Congress (Frelimo 1977). The basic lines of policy were, however, clear before that: all land belonged to the state but could be held by individuals or enterprises as long as they used it well. There was to be no general nationalisation of settler-owned farms, but where farms had been abandoned or were being badly run, they were to be taken over by the state. Abandoned or seized settler and plantation holdings would not be handed over to individual farmers but kept together to be cultivated either as cooperatives or state-farms. Collectivisation of the countryside was to be based on residential regrouping of the population in communal villages that would have as their productive base either cooperatives or a state-farm. Family farms were to be gradually reduced to one hectare of rain-fed land or half a hectare of irrigated land.

State-farms were to be the focus of investment because their large area, rational organisation of resources and access to mechanised means of production would allow them both to feed the nation and to be centres of agricultural science and teaching. The priority given to food production was intended to address the problem of feeding the cities and industrial centres in a strategy of rapid industrial accumulation. Agriculture would be the base and industry was defined as the dynamising element, i.e. surpluses for investment in industry would come by improving productivity in agriculture and establishing mechanisms for mobilising the resulting agricultural surplus.

Frelimo also used evolutionist Marxist-Leninist language to describe as ‘semi-feudal’ many of the institutions of the rural world, distinguishing those that it thought were bad – particularly traditional chiefs, beliefs in sorcery and witchcraft and bridewealth payments from those that were good – such as music, dance, herbalists and midwives. Such a view of rural people did not really depend on the texts of Marxism-Leninism, however, since it so permeated colonial thought. In the absence of a clear class enemy, underdevelopment became the principal enemy of the revolution; the word itself conjured up the hoe and hut of rural life.

Commercial profit was a potential source of accumulation, but Frelimo sought immediately to restrict the autonomy of the market as much as possible. Pan-territorial prices were declared, effective only if the state subsidised transport costs. Initially the network of rural shops, crucial for distribution of consumer goods and inputs and in marketing of peasant surpluses, was taken over by state-run ‘people’s shops’ (lojas do povo) and seasonal government crop-buying brigades, pre-empting private trade, though by 1979-80, the latter had been reinstituted, giving selected rural traders quotas of consumer goods at official prices. Rural services, many of which had been provided by private farmers and traders, were also taken over by the state. Tractors left behind were either absorbed by state-farms or centralised in state-operated machine-posts that were to privilege provisioning of farm machinery and fuel to cooperatives.

The size of peasant farms was normatively restricted though in most areas this was difficult for Frelimo to enforce. In the Zambezi and Limpopo valleys the floods of 1976 and 1977 facilitated both the obligatory movement of people into nucleated settlements on higher ground, the communal villages, and limiting the size of peasant holdings of irrigable valley land (though not to the declared ½ ha.). Similarly the Rhodesian attacks in Manica facilitated regrouping of people into protected communal villages and effectively restricted easy access to land. Across the country, with the consolidation of scattered settler farms into large-scale state-farms, former settlement scheme farmers and small specialised commercial farmers lost their land.

Restrictions placed on market agents and the extension of prices controls and subsidies (neither alien to the previous colonial regime) meant that the state took over coordination between different groups of producers. Planning began as an effort to coordinate the use of resources controlled by the state, but in the period between 1977 and 1981 the institutional and legal framework was established to subject markets to the planning process (Wuyts 1989). A national planning commission was set up and all economic ministries subordinated to it. Advisors from socialist countries introduced material balances planning to replace monetary accounting. State enterprises were legally required to produce according to the plan, given annual production targets, to exchange with other state enterprises also subject to the plan and had to transfer financial surpluses to the state-budget.

In 1979, it was announced that the country would be preparing a ten-year plan, the PPI. Production councils were formalised in all state enterprises, but their principal tasks were not defined in the directives of the Third Congress as giving workers input into the planning process or assuring that the plan protected workers’ interests. Class struggle was redefined as maintaining worker discipline, striving for higher productivity and transforming production (Wuyts 1989). Nonetheless, in practice the councils and party cells often disputed management’s interpretation of how targets should be reached and explanations for why they were not.

Frelimo was particularly concerned about the potential resurgence of petty bourgeois political power in collective organisations that extended beyond workplaces. In the theory of Marxist-Leninist parties, mass organisations (women, youth, teachers, workers…) play an important political role in building alliances between working classes. Frelimo instituted many of these, but despite debate over the question within the party, did not permit the organisation of a national peasants’ organisation. The National Union of Peasants (UNAC) was not organised until 1987 and then as an autonomous organisation not linked to Frelimo. Even the central trade union organisation (OTM), which was subordinated to Frelimo until 1990, was not formed until 1983. The orthodox wing within the party thought the production councils were not sufficiently ‘mature’ and feared that the clout of a national union led by a labour aristocracy might lead to an escalation of wage demands.

With control over the state apparatus, the flight of most settlers and the scope of private property limited, Frelimo treated both national liberation and the dissolution of capitalist relations of production as accomplished.[6] They spoke of racial inequalities in the past but not in the present; the worker-peasant alliance (aliança operário-camponesa) was declared the order of the day; almost everyone was considered to be either a peasant or a worker.[7] Behaviour, such as ambition or corruption, were defined as capitalist, not ownership of property. Gradually the definition of the political struggle shifted; by 1980 the class enemy had become very abstract – under-development with an analysis of causality vague enough to bridge capitalism, colonialism and semi-feudalism. In the large amphitheatre in the Medical Faculty in Maputo where political lectures and debates were often held, the national director of planning instructed everyone to pull in their belts, for the next decade was to be decade of the victory over under-development.

Retrospectively this linear view of socialist revolution – a march through history set at the moment when revolutionaries seize state power and impeded mainly by the actions of external and internal enemies – may now seem naïve. Yet in the late 1970s it did not appear to be so, either to revolutionaries or to western intelligence agencies, both of whom in the wake of the United States defeat in Vietnam saw generalised socialist revolution in southern Africa as a real possibility. In this context a materialist analysis of failure, such as that of the state-farms was extraordinarily difficult.

The State Farms as a Locus of Contradictory Relations of Production[8]

Marx’s analysis of capitalism showed that commodities are not just things circulating in the space of markets but fundamentally the expression of underlying social relationships. Challenging the logic of markets thus means much more than outlawing or regulating them; it means transforming relations of production: not only the relations of exploitation between capital and labour, but also the frequently gendered relation between commodified and non-commodified production. Frelimo’s strategy for displacing the logic of the market in rural areas was to concentrate resources in the hands of the state and then to plan how they would be used to fuel investment in state-farms and cooperatives.

It was initially expected that planning agrarian production would be relatively easy since most of the settler agrarian bourgeoisie had fled and the rural population was thought to be made up of a semi-feudal subsistence-oriented peasantry mainly living outside the market and employing rudimentary techniques of production. Instead very few peasants joined the cooperatives and when they did, their productivity, at least as measured by what was sold, appeared to be much less than what they achieved on their family plots.[9] As for the state-farms, planning failed miserably at finding a socialist answer to what has been arguably the central problem of capitalist agriculture: how to have enough workers when you need them and not have them when you do not.[10] It also failed to transform capitalist relations of production in the rural economy and consequently fuelled parallel processes of accumulation and class dynamics that undercut its emancipatory objectives. The reasons for this failure have to do with the inaccuracy of Frelimo’s understanding of its strategic beginning point, class structure at the end of the colonial period, but they also reflect more general theoretical problems entailed in linear millenarian views of revolutionary transition.

Though each state-farm had its own particular history and distinctive management and cropping patterns, there was a common contradiction between the planning of labour use and real labour used and paid on most farms. Figure 1 below illustrates this with data from 1982 for one block of CAIA, an agro-industrial complex specialising in the production of maize and potatoes in Tete province.


figure 1

The number of workers required in the peak labour month, November, the first weeding period for maize on both peasant and state farms, was around 7200. Yet November and December were the months when the fewest casual days were worked. On the other hand the most casual days were worked in March, when according to the plan there should not have been so much field labour required. Despite persistent shortages of field workers at appointed times, the overall wage-bill for casual labour was actually much higher than planned: around 39.2 million meticais vs. 27.4 million allocated. Crop yields were not surprisingly much below the plan and labour costs per unit of production much higher. Behind this technical question of productivity lay three inter-related lacunae in Frelimo’s strategic vision of how state-farms would socialise relations of production.

The first failure in Frelimo’s state-farm strategy was that it did not address the transformation of rural labour markets, which were assumed not to exist. It was presumed that underemployed labour was available in the pre-feudal subsistence-oriented family sector and could be politically mobilised to provide it. Implicitly women were expected to cover the loss of labour by intensifying their own work with no impact on levels of food production, nutrition or health.

The cropping patterns planned for the state-farms exacerbated the seasonal labour demands of former settler-farms. In Frelimo’s strategic vision, economies of scale were key for state–farms. Large contiguous units could be worked by heavy machinery and even sprayed aerially. Frelimo policy was therefore to group scattered settler farms together in large units, reinforcing the mono-cropping patterns of conventional plantations. In 1981, a planner in the Ministry of Agriculture calculated the possible number of permanent workers on state-farms on the basis of current cropping patterns (Quental-Mendes 1981). Except for a horticultural farm all showed one or more sharp peaks in planned labour demand. Many of the state-farms shared the same agricultural calendar as the surrounding peasant households that were asked to furnish seasonal labour, inevitably leading to conflicts in weeding and harvesting periods.

Planning procedures minimised the implications of seasonality for recruitment, procurement and investment in housing by projecting a virtual stable labour-force. For budgeting purposes, the plan calculated that CAIA needed on average about 3000 workers a month. In discussing the problems of labour shortage, the director of CAIA repeated this estimate to us, dropping the key phrase ‘on average.’ Recruiting 7000 workers during some months and providing food and accommodation for them is quite different from recruiting 3000. Some of the casual workers we met at CAIA were sleeping in abandoned pigsties using potato sacks for blankets. Short rations were a complaint on almost all farms.

Shortfalls in seasonal labour supply were attributed to a political failure of consciousness on the part of the peasantry and thus addressed through political mobilisation. In theory ‘mobilisation’ was supposed to persuade people to do voluntary labour. Conscription was contrary to Frelimo’s critique of colonialism and its celebration of emancipation. In practice it was often otherwise. In the case of the Lugela plantations in Zambezia, in 1980 workers were recruited in two ways. For immediate and small-scale problems, the director of the plantation informed the secretary of the party on the plantation who, accompanied by militia, would call on village secretaries in areas near the plantation to ask for workers, including women and school-children. Larger and longer-term requests for young men were channelled through the District Administrator who then allocated these requests to administrators at the local level. One administrator sent out militias to run document checks on young men to see whether or not they had paid their taxes. Generally they had not, so their documents were seized and they were sent off to a plantation until they had earned enough to pay their taxes.

Labour shortages at particular times of the year were so pressing that niceties broke down. A state cotton farm in Nampula in 1979 negotiated directly with local officials for seasonal weeding labour, sending trucks to distant areas to pick up women and children. On the large state-farms, one block might even steal workers from another block, just as the settler farmers had sometimes stolen each other’s workers. At CAIA, there was a low-lying swamp block where it was unpleasant to work. The block-chief would send some militia and a truck to pick up workers from other blocks of CAIA. In accord with military idioms, this practice was euphemistically called ‘going on patrol.’

When all else failed, a major Frelimo political figure might step in. Samora Machel, citing the Cuban sugar harvest, mobilised factory workers and students from Maputo for the rice harvest at CAIL. In 1982, an important minister visited Angonia just as the new agricultural year was about to begin. In a public speech, she called upon the party and the government to mobilise the population to provide workers for CAIA, people’s property, when needed. After all, she argued, such a populous district as Angonia should have been able to provide 3000 workers (note the recurrence of the planner’s average).

Criticism of the supposedly underemployed peasantry for failing to recognise the needs of the nation hardened with the continuing decline of state-farm production. In 1986, the governor of Nampula announced in a widely reported speech that the local administrators would have to recruit workers for all the state and private firms (by 1986 capitalism was renascent). Moreover, ‘producing cotton and cashew nuts is not a favour; it is an order of the state.’[11]

Diversification of rural livelihoods by the end of the colonial period meant that many were willing to take on wage-work in agriculture, but they expected decent conditions of work and housing and a wage that would allow them and their families to buy the inputs and some of the same commodities people in cities wanted – staples like oil and sugar, clothing, soap, fuel, notebooks. Here they were stymied by the goods famine in the countryside, which undercut the value of the money-wage.

Here lies the second failure of state-farm strategy – not recognising the importance of local commodity markets for both state-farms and the peasantry. Given the scarcity of consumer goods, priority went to urban areas and selected sectors in rural areas: state-farms, cooperatives, schools and the army. The state farms were supposed to trade with each other, a sugar plantation exchanging with a nearby fruit and vegetable farm (Cardoso 1991). But not only did the peasantry need consumer goods, state-farms needed staples to feed their workers, which in many areas meant buying peasant surpluses. These were not forthcoming, not because the peasantry were subsistence producers but because the state-farms offered no commodities in return. In the end, the state-farms and the army turned to private traders and became important agents fuelling the rise of parallel markets and thus accumulation of commercial capital (Mackintosh 1985).

The third strategic failure was the absence of new forms of management to deconstruct the extreme alienation of manual workers characteristic of colonial capitalist agriculture. Planning did not dislodge capitalist relations of production from the organisation of work or forms of payment or the relations between management and workers. There were three different layers in the labour force. First were the educated technicians and managers, a few foreigners but mainly Mozambicans, who were capable of understanding and responding to communications from the ministry, but did not always know much about local agriculture and even less about the livelihoods of people in the surrounding communities. Then there were the experienced permanent workers. When settlers left, their foremen and tractor drivers often stayed on. They came with ideas about which crops would grow on which fields, about how workers should be supervised and paid, about which machinery was good. There was tension between these two groups but also collusion. Both were paid a monthly wage, had privileged access to consumer goods and often had housing provided by the company. They also had access to the unused inputs and water rights of the state-farms, some of which were employed in their own fields and some they sold to local private farmers.

The third layer of the labour force were the casual workers, identified as peasants on the payroll sheets of some farms, paid by the day on a task- or piece-wage. Women and children were usually the truly casual workers, recruited from day to day, not knowing whether they would be hired, working only in periods of peak labour demand. A substantial part of the male day-workers on most farms became, however, effectively fixed. There was an informal system, organised at the field level, in which foremen recruited permanent teams, setting shifts and marking pay-cards to assure that their (mainly male) workers got a consistent monthly wage and to give them time-off for home-visits. This is why the second line in Figure 1, days really worked (or paid), is almost flat.

This team system was outside the plan and formally unrecognised by the director and the administrative staff. The permanent workers who dominated the production councils and party-cells knew of these informal arrangements of course, but pointed out that they were fulfilling Frelimo’s mandate to protect and advance the rights of workers. Cardoso (1991) noted with some irritation that Frelimo members at the base meddled in management tasks, blocking the reforms he as director proposed and that higher-level party officials had approved. The administrative workers who controlled the time-cards also saw how the informal system worked, but regarded planning and control as formal exercises, intended ‘for the Englishman to see.’ So here again the ‘rural proletariat,’ the permanent skilled workers, were embedded in class relations of both conflict and collusion.

Conclusion: The Poetry of the Future

By 1983, it was apparent that the state farms could not sustain the recovery in yields made in the first few years of independence. Production was falling, debt was sky-rocketing, some land was not even being cropped, and there was not enough foreign exchange to sustain the country’s dependence on imported inputs (Wuyts 1989). Results in the cooperatives were equally disastrous. The fourth party congress in May 1983 decided to privatise part of the state-farm land, to allow cooperatives to transform themselves into associations of small-holders and generally to provide more support for family and commercial private farming.

Some within Frelimo characterised this as a change in socialist tactics, a Mozambican NEP, but others saw it as shift towards capitalism. The key political slogan for the period was ‘Land should belong to those with the capacity to work it.’ Some interpreted this as meaning that the land should be in the hands of the peasants, but others took it to imply that it should belong to those with the capital needed to purchase inputs and hire labour. These were the years when USAID appeared, extending credit for vehicles and machinery in urban Green Zones. One would not say that a socialist revolutionary process was out of the question, but the words ‘a luta continua’ (the struggle continues), took on an increasingly hollow ring. The areas in the countryside contested and then occupied by Renamo, the armed opposition movement, expanded rapidly and Frelimo’s policies towards the peasantry grew increasingly authoritarian. In 1986, Frelimo began serious negotiations with the IMF and the World Bank concerning the terms of a structural adjustment package.

Some attributed the failings of the Mozambican revolution in rural areas to external support from imperialist powers for internal resistance. Contemporary academic analysts and now many within Frelimo itself focus on its attempts to marginalise ‘traditional’ authority politically.[12] This chapter suggests that the answer should begin where Marxist analysis – historical and materialist – always suggests, with the politics of production. This account of the failures of state-farm strategy has made some points that apply to Frelimo’s agrarian strategy but that mirror questions raised by other histories of socialist revolution as well.

First, Frelimo’s class analysis of its strategic beginning point placed most of the Mozambican population outside capitalist relations of production. There was only a small urban working-class, which Frelimo distrusted as a labour aristocracy; almost everyone in rural areas was viewed as part of a semi-feudal subsistence-oriented peasantry for whom the market was not a necessity. Failure to see the broad range of relations of production that subordinate labour to capital has not been just an issue for Frelimo. It underlies the difficulties confronted by Marxists in understanding the politics of the diverse forms of exploitation, oppression and resistance in nationalist struggles. It also surfaces in tensions between Marxism and feminism over concepts such as unproductive labour. In the case of Mozambique, where colonial patterns of labour recruitment and forced cropping intensified women’s already substantial responsibility for agricultural production, calls to work harder in the context of severe goods shortage fell particularly flat.

Second, Frelimo took from the history of Marxist-Leninist movements a millenarian tendency to see socialist revolution as that moment when revolutionaries seize state-power, neutralise the bourgeoisie and then follow a fixed path towards an emancipatory future (unless perturbed or displaced by the need to confront enemy action). Its main strategic concern was with economic development; it had no strategy for addressing the ways that capitalist relations of production were embedded in the organisation of work, including non-commodified work. Its strategy for markets was to limit their functioning, not to reorganise how they continued to matter both for the statefarms and for the livelihood of rural people. Frelimo could not recognise ongoing class struggle in its planning processes and resistance to them as something other than ‘enemy action.’ The irony is that in the interstices of these lacunae there was room for private accumulation of capital.

So Frelimo’s failures are not exceptional. In moments when power has been defied and defeated, other socialist revolutionary movements have also been tempted to envision the future as the suspension of the past, a new world with possibilities not yet experienced. But revolution is more than such millenarian moments and class struggle is not so easily suspended; socialist revolutions are long-term contingent processes of confronting the contradictions of the past embodied in the ways that people produce and reproduce their conditions of existence and in the theoretical images (poetry if you will) they construct to understand them.


Adam, Y. 1987. Cooperativização agrícola e modificação das relações de produção no período colonial em Moçambique. Maputo: Universidade Eduardo Mondlane.

_____ 2006. Escapar os Dentes do Crocodilo e Cair na Poca do Leopardo, Trajectório de Moçambique Pos-Colonial (1975-1990). Maputo: Promédia.

Bettelheim, C. & B. Pearce 1976. Class Struggles in the USSR. First period: 1917-1923. New York, NY: Monthly Review Press.

Burawoy, M. 2009. Working in the tracks of state socialism. Capital & Class 33: 2, 33-64.

Cardoso, F. J. 1991. Estratégias, economias locais e empresas  agrárias: o desenvolvimento rural em Moçambique. Lisboa: Instituto Superior de Economia e Gestão, Universidade Técnica de Lisboa.

Dinerman, A. 2006. Revolution, counter-revolution and revisionism in post-colonial Africa: the case of Mozambique, 1975-1994. London: Routledge.

Frelimo. 1977. Explicação das teses do III Congresso. Maputo: Frelimo.

Hermele, K. 1986. Contemporary land struggles on the Limpopo: A case study of Chokwe, Mozambique, 1950-1985. Uppsala: Department of Development Studies, University of Uppsala.

Isaacman, A. et al. 1980. Cotton is the Mother of Poverty: Peasant Resistance to Forced Cotton Production in Mozambique, 1938-1961. The International Journal of African Historical Studies 13: 4, 581-615.

Mackintosh, M. 1985. Economic tactics: Commercial policy and the socialization of African agriculture. World Development 13: 1, 77-96.

Mamdani, M. 1996. Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism. Oxford: James Currey Publishers.

Marx, K. 1852. The 18th Brumaire of Louis Bonaparte (available on-line at: http://www.marxists.org/archive/marx/works/, accessed 4 August 2015).

Mondlane, E. 1995. Lutar por Moçambique. Maputo: Centro de Estudos Africanos, Universidade Eduardo Mondlane.

Mosca, J. 1988. Contribuição para o estudo do sector agrário do Chókwè. Maputo: Ministério da Agricultura.

Moyana, S. 1986. Produzir algodão e castanha decaju não é favor é ordem do Estado. Tempo 836, 1215.

O’Laughlin, B. 1996. Through a divided glass: dualism, class and the agrarian question in Mozambique. Journal of Peasant Studies 23: 4, 1-39.

Penvenne, J. 1995. African workers and colonial racism: Mozambican strategies  and struggles in Lourenco Marques, 1877-1962. Portsmouth, NH: Heinemann.

Schaedel, M. 1984. Eingenborenen-Arbeit: Formen der Ausbeutung under der portuguiesichen Kolonialherrschaft in Mozambique. Koln: Pahl-Rugenstein Verlag.

Vail, L. & L. White 1980. Capitalism and colonialism in Mozambique: a study of Quelimane district. Minneapolis, MN: University of Minnesota Press.

Wuyts, M. E. 1989. Money and planning for socialist transition: the Mozambican experience. London: Gower.

About the contributor

After leaving Mozambique in 1992, Bridget O’Laughlin taught at the ISS (Institute of Social Studies) in The Hague. Now retired, she is a research associate of IESE (Institute of Social and Economic Studies) in Maputo. Her current research deals with the political economy of rural health in Southern Africa. Her e-mail address is brolaughlin@iss.nl.


  1. This paper draws heavily on my experience as a teacher and researcher at Eduardo Mondlane University in Maputo from 1979 to 1992 and particularly on collective research done at the Centro de Estudos Africanos (CEA) under the direction of Aquino de Bragança and Ruth First.
  2. The one important exception, the stevedores union, was limited only to those contracted by private shipping agencies for work in the boats and acted as a recruiting agency.
  3. Some of Frelimo’s critics have treated this as the dispossession of the peasantry by the state, but at the time this was simply a way of asserting that the land belonged to the Mozambican nation and not to those who claimed it through colonial dispensation.
  4. These departures were often clandestine and quick, reflected in the use of Cuban terminology to describe leaving as a kind of betrayal, a flight (fugiu).
  5. Photocopy, DINAPROC, Maputo, 1978.
  6. Though one high-ranking Frelimo cadre, no longer socialist at all, was once heard to say in reference to his ultimate objective of bringing all rural production into state farms, that the cooperatives with irrigated land on the Limpopo were ‘not our principal enemy.’
  7. We were very reluctant to suggest in our CEA reports that they were not, lest that be taken as justification for the appropriation of more land for state-farms.
  8. This section draws on various sources. The CEA studied a range of state-owned or intervened farms and plantations with students from the Development Course: the UPBL rice farm on the Limpopo, a horticultural farm close to Maputo, tea plantations in Zambezia, a maize and potato complex in highland Tete and cotton and sisal plantations in Zambezia and Nampula. In these cases we also studied family farming and cooperatives in surrounding areas. There are also in-depth studies of two important state farms done by their former Mozambican directors as doctoral theses, the CAIL complex on the Limpopo (rice, wheat, vegetables, some livestock) (see Mosca 1988) and the Maragra sugar estate (see Cardoso 1991).
  9. Various CEA research projects across Mozambique 1979-1983. There is not space to discuss why cooperatives failed here, but the approach – treating them as a space of class struggle – would be the same.
  10. Sidney Mintz’s formulation.
  11. Moyana (1986: 1415).
  12. As Dinerman (2006) emphasises, Frelimo’s self-critique here is significantly partial, obviating the need for any critical analysis of its economic policies.