Since the inception of development economics the challenges of rural development have attracted great academic and political attention. Yet, in view of persistent and largely unchanging poverty levels – and poverty being an overwhelmingly rural phenomenon (IFAD 2001) – it becomes apparent that these problems remain largely unresolved to this day, and they continue to constitute some of the most trenchant issues of economic development, particularly with respect to Africa.1 This paper argues that without a fundamental understanding of the agrarian question, the attempt to derive any conclusions on rural development is doomed to be incomprehensive and incomplete.
Unfortunately, such inconclusiveness currently is a widespread condition amongst the profession of development economics, and the publication of the recent World Development Report 2008 Agriculture for development serves as a lucid and illustrative example for this contemporary trend. A wide range of scholars (for example, Havenevik et al. 2007, Rizzo 2009, Woodhouse 2009) have pointed to and deplored its many ‘apparent analytical and empirical incoherence[s]’ (Oya 2009, p. 232), culminating in highly contradictory and inconclusive policy recommendations. This article aims to trace this confusion back to some of its theoretical origins in the agrarian question, as well as to reconcile them with new illustrative survey data from the West Usambara Mountains in Tanzania.
Connecting the challenge of rural development in Tanzania with the agrarian question
With respect to the pressing issues of rural development, Tanzania is a typical case in point. The country remains one of the poorest economies in the world with an estimated per capita GDP of just over US$500 (IMF 2009) and a Human Development Index ranking of 159 out of 177 countries (UNDP 2008), despite massive development efforts by the government and international donor community who continue to celebrate Tanzania as a ‘donor darling’ (Harrison et al. 2009) and ‘the story of an African transition’ (Nord et al. 2009).
And indeed, Tanzania has achieved rather respectable consistent GDP growth figures of 5–7.5% over the past decade. However, most commentators agree that the majority of Tanzanians have not benefited as poverty reduction failed to follow suit. Between the two most recent national household budget surveys in 2001 and 2007 (URoT 2002, 2008b), no statistically significant decrease of the share of people experiencing poverty could be observed – however speaking in absolute terms, it was found that about 1.5 million more Tanzanians now live in poverty compared to 2001.
As is the case in so many African countries, agriculture is the main pillar of the Tanzanian economy, contributing about 26%2 to national GDP and employing more than 80% of its working population (data: URoT 2008a, World Bank 2008). It is especially the rural areas that suffer the most from widespread income poverty – about 83% of the country's poor live in rural areas – with nearly 38% of the rural population falling below a basic needs poverty line and 18.4% below the national food poverty line (URoT 2002, 2008b). Per capita food availability has been in constant decline since its recorded high in 1979, since then having fallen by 29% to this day. It is undoubtedly here in rural Tanzania that the country's main development challenges lie, and this has been echoed in recent donor commitments and government policies. Tanzania's latest nationwide development strategy is called ‘Kilimo Kwanza’ – Agriculture First.
What is the major handicap that the rural population faces? To many politicians and academics, the answer to this question seems clear: the agricultural sector's lack of market orientation and exposure, as President Jakaya Kikwete recently made quite clear when describing the sector: ‘It is subsistence agriculture – living from hand to mouth!’ (as cited in World Bank 2007a). On the face of it, this picture is confirmed by national surveys and academic research alike. The latest Integrated Labour Force Survey 2006 reported a share of wage or salaried employment of only 9% of the economically active population (79.8% of the total population), and allegedly only 11% of rural households have at least one member that is engaged in wage employment (URoT 2007), therewith confirming the position that African rural societies primarily consist of small-scale farming, self-reliant households. Repeatedly, numerous academic scholars further endorsed this notion by theorising the ‘peasant mode of production’ (Hyden 1980, 1983, Waters 2007). This image of African farmers has been very formative not only for public perceptions on rural Africa, but also for the design of national and international agricultural policies and poverty reduction strategies which focus on the small-scale farmer.
However, in stark contrast to this, a labour market survey conducted by the author in the West Usambara mountains in 2008 found that 78% of all participating households had at least one member that had engaged in wage-labour activities, and 43% resorted to hired labour in their production process in the observed season. These empirical findings and their general interpretation, which will be discussed in detail in the second half of this article, are also backed by a small number of other studies (for example, Sender 1975, Sender and Smith 1990), which in general have been relatively marginalised in the wider debates of rural development and policy responses in Tanzania. Furthermore, most discussions of the ‘agrarian question’ remain deeply theoretical and abstract and this is likely to be one of the reasons why they find little resonance with the wider development community. For this reason it is believed by the author that it is essential to substantiate possibly dry theory with palpable evidence. Even though its full completion is next to impossible, we are convinced that such an attempted ‘bridge’ between the abstract and the concrete is an important tool, and this article is intended as a first contribution to this effect.
To all those who are genuinely concerned with economic and social progress in Tanzania and Africa in general, this lack of discussion should be deeply troubling. The (agrarian) question about farmers' mode of production must lie at the very core of any rural development discourse, given that it is concerned with the systematic determinants of farmers' reactions to political and economic stimuli of any kind. Consequentially, the rural mode of production has a direct impact on the rural economy's dynamics, progression and indeed development. Ceteris paribus and greatly simplified, a subsistence farmer's decisions about production are based on the trade-off between the ‘cost’ of personal drudgery s/he is willing to impose on himself and the aspired extent of family consumption. As long as the peasant farmer has direct control over the means of subsistence and production, ‘there is no compulsion to exchange in order to reproduce, no pressure/necessity to compete, thus no requirement to accumulate especially by way of innovating in order to survive’ (Brenner 1977, p. 49). The result of course is a mostly static situation that does not adhere to any systematic or intrinsic mechanism for improvement, advancement and investment. A capitalist farmer, on the other hand, is in a situation of direct competition for market share and profits, not least in order to be able to bear the cost of wage labour inputs. Unlike his/her peasant counterpart, he/she is subject to a systematic compulsion to be an efficient producer, and therefore to invest, to expand and to accumulate. The wage labourer – in the double sense ‘free’ from the means of subsistence and free to sell his labour power – faces competition (for jobs) too, and therewith is compelled to be as hardworking and industrious as possible in order to avoid being segregated from his sole source of income. But more importantly, ‘[o]nly under conditions of free wage labour will the individual producing units (combining labour power and the means of production) be forced to sell in order to buy, to buy in order to survive and reproduce, and ultimately to expand and innovate in order to maintain this position in relationship to other competing productive units’ (ibid., p. 32). The agrarian question, as described in more detail in the following section, seeks to tackle these exact socio-economic processes of transition between the peasant economy and a capitalist agricultural sector.
The advent of capitalist relations is therefore a necessary (but clearly not sufficient) condition for development, as only through the capitalist imperative can there be a fundamental trend reversal and a systematic end to stagnation in rural Africa. ‘For capitalism differs from all pre-capitalist modes of production in its systematic tendency to unprecedented, though neither continuous nor unlimited, economic development’ (ibid., p. 30), and ‘whatever Marx thought about the origins of capitalist social productive relations, he was quite clear that their establishment was indispensable for the development of the productive forces, i.e. for capitalist economic development’ (ibid., p. 26). Conversely, it is – by definition – absurd to assume that capitalist development, as achieved by every advanced economy so far, can be based on egalitarian (subsistence) communities of small-scale farmers and economic ‘backwardness’. Regrettably, in contemporary economic science and development studies these concerns and relations have been unjustifiably eclipsed. Yet, they are at the heart of the agrarian question and any rural development context, consequently having been the unresolved concern of scholars across the world for more than a century.
It is essential for any serious understanding of rural development to reconnect issues of rural and agricultural economics, as well as agrarian change, with their theoretical roots in the agrarian question. For this, in the next section I will attempt a summary of how the important historic and theoretic developments of the agrarian question together with their subsequent antagonistic theoretical and ideological approaches have culminated in the disorientation and confusion of contemporary thinking on rural development. Following this, qualitative and quantitative survey data collected in 2008 will be used to exemplify the living reality of farmers in the West Usambara Mountains in Tanzania today and show how these correspond to the different concepts discussed in the preceding section.
The origins of the agrarian question
The agrarian question in its original form was penned by Friedrich Engels as Die Bauernfrage in Frankreich und Deutschland 4 (Engels 1972 [1894]). Engels posed it with an exclusively political concern in mind: How to capture political power in European countries where capitalism was developing but had not yet replaced pre-capitalist social relations as the overwhelming agrarian reality? Or – to put it more bluntly – how can the (essentially pre-capitalist) peasantry be rallied and enthused for the social revolution? In the course of this endeavour, Engels immediately realised a perhaps surprising reality about the peasantry of his time: though not yet fully captured by the capitalist mode of production, it was already considerably differentiated between big, middle and small peasants. He defined small peasants as ‘the owner or tenant […] of a small piece of land no bigger than he and his family, as a rule, can till, and no smaller than can feed the family’ (ibid., p. 488, translated from the German by the author). For Engels therefore, the small peasant usually did not have to resort to non-family labour, and, if at all, only occasionally in the form of labour supply. Middle and big peasants on the other hand refers to ‘peasant property bigger than those that can be cultivated by the family itself, hence […] who “exploit day labourers”’ (ibid., p. 496). Less surprisingly, Engels concluded that socialist endeavours of his time should focus on winning over the small peasantry for their cause.
Of course, from today's perspective the relevance of this reading has mostly disappeared. However, despite its now marginalised political value it did immediately serve to inspire a vigorous intellectual and academic interest and debate, which has lasted for more than a century and has stimulated a successive, arguably more sophisticated approach towards the agrarian question. This ‘second instalment’ was first systematically analysed by Karl Kautsky in his Die Agrarfrage (1899), as well as by Lenin in his treatise on The Development of Capitalism in Russia (1956 [1899]). It can be split into two distinct, but connected concerns:
- a.
Why is a large peasantry apparently able to persist and reproduce itself to some degree within social formations that are characterised by an otherwise inexorable expansion of the capitalist mode of production and society?
- b.
In what form and to what extent is capitalism developing in rural areas, and what are the barriers that may impede it?
Kautsky and Lenin therefore agree that the European peasantries already succumbed to the overarching capitalist market system, however, the capitalist mode of production, where capital and labour are fully separated, had not fully taken control of the rural society yet. In this sense, ‘the peasantry are [sic] not antagonists of capitalism, but, on the contrary, are its deepest and most durable foundation’ (Lenin 1956, p. 173). Consequently, according to this theory capitalism would further encroach on the rural society which over time would lead to the gradual dissolution of the middle peasantry and the complete subsumption of the poor peasantry as wage labourers by rich peasants, who after all are more competitive given ‘the law of greater productivity […] of the big peasants’ (ibid., p. 56). To Lenin and Kautsky this is the only possible solution.
With regard to the process of development, a crucial and consequential question is what factors might contribute or impair this otherwise inexorable process. Lenin wrote that ‘[o]n the question of whether the disintegration of the peasantry is progressing, and if so at what rate, we have no precise statistics’ (ibid., p. 183). However, he offers a few explanations of what factors decide on the rate of capitalism's advancement, such as migration as a major accelerating factor to the disintegration of the peasantry, or (semi) feudal relations in which peasants are tied to their land and subjected to extra-economic coercion as a major impediment to capitalist accumulation. As a general rule, we can assert that any condition that prevents the direct producer from being ‘freed in the double sense’ – free to move and choose their master, but also freed from the means of subsistence (that is, land) – will retard the disintegration of the peasantry, and therewith the transition to capitalism, or in other words, the solution to the agrarian question.
In their time, Kautsky and Lenin provided important answers as to the state and future of the peasantry. Their approaches, however, did not remain unchallenged, since many authors regarded the archetypal and pristine peasantry as the superior unit of agricultural production and social organisation. They therefore sought to establish a development path that took the idealised peasant household as the role model unit, rather than requiring its demise. This, the neo-populist school, was led by the Russian agrarian economist Alexander Chayanov who developed the theory of the peasant economy (Chayanov 1966). His model of the rural economy is very antagonistic to the Marxist approaches discussed beforehand, as it disregarded peasant differentiation based on class relations. Instead, he built an economic model that identified a peasant household's production and consumption decisions based on its demographic life cycle. In this, the individual peasant household would base its production and consumption decisions solely on the trade-off between its consumption demands (determined by the number of household members) and the level of drudgery and self-exploitation necessary to meet these (determined by the number of working household members). This cost–benefit calculation, which shifts pursuant to the household's demographic situation, led Chayanov and his followers to deduce that peasants constantly operate with superior efficiency compared to capitalist farms, which they regard as relatively wasteful, inefficient and unsustainable. They aspired to promote a form of development and advancement built on the shoulders of the efficient peasant households who were to lift ‘Мамушка Россия’ (‘Mother Russia’) to greater prosperity. Naturally, neo-populists disputed the Marxists' concepts of peasant differentiation and encroachment of capitalism, leading to an intense debate. Lenin, in turn, ‘insisted that all efforts to preserve the middle against the extremes as a means of recapturing this mythical past and saving the peasantry from the horrors of industrial capitalism would only serve to retard the process of “depeasantising”, to institutionalize capitalism in its least developed form, and actually worsen the condition of the smallholder’ (Mueller 1980, p. 3).
No matter how antiquated it may seem to some readers today, ‘[t]he debate between Lenin and Chayanov, or between those who embrace a position broadly similar to Lenin's or Chayanov's, has lost neither its relevance nor its force in relation to today's poor countries’ (Byres 1986, p. xxi). Although nowadays they commonly wear different labels, the old debates still have a profound influence on contemporary debate on rural development, and so does the wider issue of the agrarian question.
The unresolved legacy of the agrarian question and its continuation today
Unfortunately, most contemporary students of development economics and related social sciences would hardly have heard about ‘the agrarian question’ as such, given that this term is generally reserved for the notoriously unfashionable discipline of Marxist political economy. Nevertheless they usually will have come to discuss it in some way or other, though most likely under different headings, paradigms and ideologies. Many aspects of both the Kautsky-Lenin treatment of the agrarian question, but especially of Chayanovian neo-populism, had an important impact on subsequent theories of economic development, although few of these were able to apply similar depth and grasp to reach to the core of the problem. Examples include contemporary descriptions of rural development such as the ‘livelihood diversification’ approach (Ellis 1998, 2000), the ‘de-agrarianisation’ and ‘de-peasantisation’ theses (Bryceson and Jamal 1997, Bryceson et al. 2000). The contemporary followers of the neo-populist school and the ‘urban bias’ thesis (for example, Lipton 1977, Hyden 1980, 1983, Waters 2007) also have their thematic roots in the original debates of the early twentieth century.
In the African context, and Tanzania in particular, this influence was vibrantly notable during and after the independence struggles of the 1960s and early 1970s. Political movements like Nyerere's Ujamaa socialism, as well as academic debates, were conducted in a similar vein and spirit. The important ‘Kenya debate’ (Kitching 1980, Leys 1996), which deliberated over the possibility of industrial-capitalist development in a post-independence Africa, and even more so the ‘Dar es Salaam’ debate (Shivji 1975, Tandon 1982), which focussed on the nature of Africa's rural class and the ‘African’ mode of production, represent landmarks in the translation of the agrarian question and similarly structural development challenges into a specifically African context.
However, these early progressive discussions were arguably muted in their infancy, as the neo-populist school gained a rapidly growing following, especially following the rediscovery of Chayanov's works in Western academia in the late 1960s. Probably the most prominent proponent of neo-populist development economics is Michael Lipton, who developed the theory of ‘urban bias’ in the late 1960s and early 1970s, and eventually produced these in the most articulate form in 1977. Urban bias refers to the perceived extraction of surplus off the ‘rural class’ by the ‘urban class’. Neo-populists generally regard urban bias as highly detrimental to the growth of developing countries, as it ostensibly circumvents resources away from those sectors and units – agriculture and small-scale farmers – which are deemed to hold a comparative advantage. An important pillar of this approach is a reinforcement of the Chayanovian conception that smallholder farming is more efficient than other forms, especially capitalist agriculture. This direct confrontation with Kautsky's ‘technical superiority of the large farm’ (1899, p. 92) and Lenin's ‘law of the greater productivity and, hence, of the greater stability of the big peasant farms’ (1956, p. 56) is based on a number of arguments, most notably the statistically observed negative correlation between farm-size and productivity (or efficiency).5 Other arguments to support the viability of peasant farming include the view that smallholder agriculture provides self-sufficiency and food security (Naerstad 2007), that it is more environmentally friendly and sustainable through its use of ‘traditional’ and indigenous knowledge or technology (Netting 1995, Horrigan et al., 2002, La Via Campesina 2007), and most prominently that small-scale farming offers crucial answers to the challenges of poverty alleviation (for example, Rosset 1999). Ultimately, this approach tends to culminate in a more and more common view which can only be described as the romanticising of peasant and subsistence farming as a perfect ‘mode or production’ in harmony with nature and people.
Since their inception, these Chayanovian arguments have been vigorously debated. The inverse relationship represents a point of constant contention. Many scholars have debunked its myth by showing it to be an example of spurious correlation (Rudra 1968), that its empirical basis is generally weak (Sender and Johnston 2004), that the relation only holds in a static context but is likely to collapse with a dynamic perspective (Dyer 1998), and that the perceived high rate of efficiency is simply a reflection of the farmers' incisive poverty, which compels them to ‘self-exploitation’ and operation at maximum labour-intensity (Dyer 1996). Arguments for the poverty reduction potential of small-scale farming have been equally contested, primarily by showing that it is usually not the small farmers who suffer the most from abject poverty, but actually the landless, quasi-landless and especially the female-headed rural households struggling for access to land who are often worst affected (Cramer and Pontara 1998, Barrett et al. 2001). This of course renders support to farmers ineffective as a poverty reduction policy, given that the rural poor mostly rely on off-farm activities to make ends meet (Cramer et al. 2008).
Despite these serious contentions, the neo-populist ideology has exerted a great influence on development theory in general and development policies in particular. Starting in the early 1980s, the ‘urban bias’ doctrine gained a foothold and, together with a dominant conviction that industrial policies had failed, eventually led to the formation of an unprecedented movement that combined neo-classical market fundamentalism with neo-populist celebrations of smallholder agriculture, which was spearheaded by the World Bank (1982). The logic was clear: given their relative land abundance, African countries' comparative advantage lies in agriculture, which therefore must benefit greatly from market liberalisation. Furthermore, according to the neo-populist conviction, the small-scale farmer is the most productive, efficient and therefore competitive unit of production, implying that s/he too must prosper in free market competition, supposedly creating an ideal mechanism of poverty reduction and growth. The neo-liberal policy towards rural development was born, which Byres later aptly debunked as ‘neoclassical neo-populism’ (Byres 2005).
However, as evidenced by many accounts and much data, the neo-liberal approach did not yield the desired results, and – rather ironically – if anything was a direct contributor to and driver of increased peasant differentiation (Ponte 2002, Havenevik et al. 2007, Woodhouse 2009) and in the mid to late 1990s its critics regained momentum. In particular, they took issue with the predominant assumption of neo-populist thought that the vast majority of rural societies in developing countries were organised as smallholder peasant farming. Instead, they described a great variety and dominance of activities undertaken by rural dwellers, which clearly lay outside the mere realm of the family farm, such as small businesses, wage labour employment, outmigration and reliance on remittance flows. These trends are attached to a range of new labels, like ‘livelihood diversification’ (Ellis 1998, Ponte 1998, Ellis and Mdoe 2003), ‘de-agrarianisation’ (Bryceson and Jamal 1997), and even ‘de-peasantisation’ (Bryceson 1999, Bryceson et al. 2000).
Generally, a slow but discernible movement away from the neo-populist school and, especially in the case of the latter, a certain rapprochement towards the Lenin-Kautsky understanding of rural societies is currently noticeable. It nevertheless is also clear that, unlike their Marxist ‘predecessors’, most of these new stories have so far failed to achieve a sufficient degree of structural understanding or explanation of these trends, and they are yet to deliver a satisfactory systematic account that is able to place the observed trends of ‘livelihood diversification’ into a coherent framework that allows conclusions to be drawn about their actual role within the development process. Part of this has to do with the lack of acknowledgement of the considerable degree of differentiation and class formation that is naturally occurring alongside the process of livelihood diversification, nay, that lies at its very core. So far, most proponents of ‘livelihood diversification’ and ‘off-farm employment’ stories continue to treat the rural community as generally homogenous individuals, of which – for some reason or other – some happen to run businesses and some others happen to be wage labourers, without being able to understand the fundamental causes of these tendencies.
That said, this new paradigm has caused considerable upheaval within the field of rural development. As mentioned in the introduction to this paper, this became impressively evident with the publication of the recent World Development Report Agriculture for development (World Bank 2007b), which displayed a rich array of the most vague and equivocal assertions, as well as internal contradictions. Many commentators noticed and exposed many of these, suggesting as it were that the Bank's economists started to wave the metaphorical white flag by rather openly displaying a great amount of confusion and disorientation over the current processes of agrarian change in rural Africa. On the one side, the report acknowledges the different paths out of poverty achieved by ‘moving beyond the farm’ (World Bank 2007b, pp. 202–221), but then it devotes the bulk of its chapters to the question of how smallholder agriculture must be supported effectively (mainly through opening their access to markets). This squaring of the circle becomes even more unintelligible when the authors take up the cudgels for agribusiness, and – as Rizzo (2009) astutely points out – manage to dismiss and support the inverse relationship on one and the same page (World Bank 2007b, p. 91). In accordance with our argument in the previous paragraph, the Bank does accept the notion of diversification of farmers with regard to their individual livelihoods, but is not able to deduce from this that farmers as a logical consequence also have to be increasingly differentiated, that is, they occupy different positions in the socio-economic and political power relations. It therefore appears as if World Bank economists desperately try to cling to the championing of this dated smallholder ideal, despite their own creeping realisation of its glaring incompatibility with reality.
The recent reappearance of contributions on the topic from African scholars goes much further than the mainstream approach. In recognising and applying holistic approaches to the debate they have devoted their attention to issues of land and (neo-)imperialism, and – intimately connected to these – the national and the agrarian question (Mafeje 2003, Moyo and Yeros 2005, Shivji 2006, Moyo 2007, Amanor and Moyo 2008, Shivji 2009). Acknowledging the distinct (class) relations between actors – nationally and internationally – they raise the question of the static and dynamic position of the African peasant within a system of global markets and wide-ranging neo-liberal hegemony, whilst clearly identifying their neo-imperialist character. Connecting to the ‘articulation of modes of production’ debate of the late 1970s (Rey 1973, Foster-Carter 1978, Wolpe 1980, Freund 1985), they characterise the process of accumulation in Africa as being ‘disarticulated’, in essence meaning that primitive and capitalist accumulation is occurring continuously, and yet it does not induce any further processes of transition, industrialisation or more generally the development of the productive forces. Instead, the created surplus value is being expatriated or used to fuel the excessive consumption levels of minuscule elites (Shivji 2009). In this regard, they identify neo-liberal policies and international financial institution (IFI) conditionality as the main reason and culprit for this, as they incapacitated the state to fulfil its role of curtailing the extent of exploitation and the resulting stagnation.6 Unlike the approaches mentioned above, they tend to see livelihood diversification as a clear manifestation of deeply entrenched poverty and destitution, rather than an ‘opportunity for a new agriculture’ (World Bank 2007b, p. 58). The result is ‘the prevalence of semi-proletarianisation – worker peasants – alongside the retention of large peasantry, or of small cultivators’ (Moyo 2007, p. 10) which is often argued to be in a situation of deadlock without the implementation of radical agrarian and ‘land reform in the classical and historical sense of addressing the agrarian question’ (ibid, p. 30). Unfortunately, the unanswered question remains how this can be achieved in the current age of neo-liberal domination, and – as we intend to demonstrate in the following section – more importantly the current, semi-proletarianised state of the peasantry is not necessarily as static as is often assumed.
However, even despite the current lack of pragmatic solutions from either side, all this stirring up of the erstwhile neo-populist ‘consensus’ must certainly be welcomed, as it manages to jolt the long-unquestioned dominance of neo-populism. With these current collisions of antagonistic approaches on how rural socio-economic relations have to be analysed and increasing disagreement over the resulting policy conclusions, it is obvious that the agrarian question in today's development discourses is alive and kicking. To illustrate this, we shall now return to the original example of Tanzania and an attempt to reconcile the theoretical debate with farmers' realities on the ground.
An analysis of the agrarian question in the case of the West Usambara Mountains, Tanzania
In 2008, we undertook field research in the West Usambara Mountains (Lushoto district), Tanzania, in order to analyse the socio-economic relations between farmers in this area, laying a particular focus on the formation of rural labour-market relations. For this, a survey with a sample size of 151 households was undertaken and complemented by in-depth collection of qualitative data by means of numerous semi-structured interviews and focus groups.7
The region stretches over an area of about 3500 km2 of which about 2000 km2 are arable land, and at the time of the last national census in 2002, the population of Lushoto district was 420,000, implying a population density of about 120 people/km2. Altitudes range between 1050 m and 2250 m, agricultural activities depend on two rainy reasons, with the short rains from November to December and the ‘long rains’ from March to May. The region features great ecological and geographic variety, with three agro-ecological zones and highly varying rainfall between 600 and 2000 mm per annum. Depending on the landform, that is, slope angles, slope position, rain shadows or valley width, conditions for agriculture vary greatly. Today, the major cash crops are horticultural ‘fast crops’ (Ponte 1998) such as tomatoes, cabbage and other vegetables. Traditional (export) cash crops such as tea and coffee are still being cultivated in some areas, but coffee production especially has declined sharply. Main food crops are maize, beans, cassava, bananas and potatoes, however, the distinction between food and cash crops is generally misleading, as virtually all crops are both traded and consumed by individual households. Several villages in the region (some of which are included in the sample) are important vegetable suppliers for Tanzania's urban areas and therefore relatively wealthy. The majority of villages, however, are not able to benefit from this interregional trade due to their geographic or transport conditions. Overall the region takes an average economic position compared to other rural areas in Tanzania and provides an interesting case for the analysis of socio-economic dynamics and agrarian change.
The overwhelming majority of sample households (HHs) had direct access to land and were relying on farming for their reproduction in some way or other. They could therefore be identified as farmers, and as a general rule they did indeed identify themselves as such. However, at the same time, the prominence of ‘livelihood diversification’ is immediately recognisable. During the observed season (2007/8), out of the 151 participating HHs:
- •
Fifty-five percent had at least one adult member who worked for a wage at the time of the survey, and another 23% had at least one adult member who had worked as a wage labourer in the past.
- •
Forty-three percent of HHs hired labour at some stage of the production process and nearly half of these also supplied wage labour in the same season.
- •
Thirty-two-and-a-half percent of HHs presently operate at least one off-farm business of sorts, and another 15.2% had run a business in the past.
- •
Thirty-nine percent of HHs had at least one dependent member who temporarily migrates or has migrated in seek of income opportunities, and an additional 28% of all HHs had experienced migration in the past.
- •
In total, 73% of HHs did receive some form of off-farm income in the observed season.
Evidently there is an active labour market in place. However, this is not uniform, but clearly divided in two qualitatively and quantitatively different segments. The first is the market for casual labour (kibarua in Swahili). Kibarua employment is characterised by being very insecure and irregular hard manual work, with very low wages of about 1000–2000 Tanzanian shillings (TSh) (around US$0.75–1.5) per day, often paid in form of task/piecemeal remuneration. The other segment constitutes the market for formal labour of all kinds, most of which requires particular skills in the form of education or training, though this is not a necessary condition. In contrast to kibarua work, formal employment – even if sometimes hard manual work – is usually much more secure, earns considerably higher incomes, usually comes with extra benefits such as pension schemes, and is mostly situated outside the agricultural sector.
In quantitative terms, the kibarua market vastly surpasses the market for formal labour – 81.4% of all individuals engaged at that time in some form of wage labour employment were kibarua (mostly, but not exclusively doing agricultural tasks). Taking into account that another 3.4% were holding unpaid public offices, in sum only 15.2% of all wage workers were able to secure formal employment. This market segmentation is a crucial aspect in the solution of the agrarian question. For this, it is important to understand that the kibarua market is predominantly situated in the rural economy. Kibarua is the ‘indigenous’ or ‘home-grown’ form of labour, based on the communal economy, for example cultivating another farmer's shamba (field), brick-making, domestic service and many more which are all organised in some form of kibarua employment. On the other hand, formal employment is more or less exclusively9 extended from the urban economy in some form or other (for example, teachers and police officers employed by the government, bus-drivers in towns, and so on). Virtually all formal jobs in the study area are related to the urban sphere, either through direct employment as public agents in the villages (mainly teachers or extension officers), or through periodic commuting.10 From the local perspective, formal salaries are very high in comparison and they enable the recipients to gain wealth and ‘capital’, as in turn this often is being used for agricultural production and the hiring of kibarua wage labour.
In this regard, our example of the West Usambaras is especially informative, as it represents a particular case with a marked decline of ‘external economic interests’, that is, in the form of (sometimes labelled ‘imperialist’) foreign investment or more generally large-scale, capital-intensive, export-oriented agricultural production.11 It therefore shows a potential indirect transmission mechanism of capitalist relations into the rural economy through the urban sphere, which might take a dominating role in the absence of any other direct involvement of external large-scale capitalist producers. Formal employment, as an extension of the urban realm, therefore directly provides one element of increased peasant differentiation in the local village community, as it is one direct expression of the encroachment of capitalism – and ultimately contributes to the resolution of the agrarian question.
This is further exemplified by the low social status of kibarua work, which becomes very apparent during conversations on the topic. The Swahili word kibarua has a very negative connotation, as it stems from the early colonial era and slave trade during which many men and women were forced to work for the colonial authorities.12 Many people are obviously ashamed and hesitant to admit that they work as kibarua. This general sentiment became the most evident, when people were asked whether they had been employed as kibarua. Several times respondents intensely proclaimed ‘Nashukuru mungu, sijaajirwa!’ – ‘I thank God, I have never been employed!’13
This is clear evidence for the struggle against proletarianisation. This struggle does not even faintly occur against formal employment, as that is without doubt regarded by the villagers as a clear accomplishment and ‘way out’ of village life with all its privations.14 These trends and struggles are of a systemic nature, and they cannot be simply explained by unassigned notions of ‘diversity for survival’ and ‘diversity for accumulation’ (Ellis 1998, p. 7), but must be seen as part of the wider issue of peasant differentiation. Furthermore, it is not sufficient to analyse them purely in terms of their class relations as they often are fought out within the household.
Gender and intra-household power relations are extremely important in shaping the process of proletarianisation and also which groups particularly have to endure its adversities. Typically and for a long time, the vast majority of people pushed into the labour markets were women – especially those freed from ‘marital obligation’ – due to their lack of access to the means of subsistence (Sender and Smith 1990, Oya and Sender 2009). Sender and Smith in particular used their empirical findings derived from fieldwork in the West Usambaras in 1986 to explain the often observed and paradoxical scarcity of labour supply. They concluded that married men – mainly through their control over land – were able to expropriate their wives' labour power, therewith being able to abstain from the labour market themselves. It was mostly the unmarried, divorced or widowed women that typically constituted the small supply of labour in the local economy. However, our findings point to the possibility of a trend reversal in this regard. In our sample, only 39% of all kibarua labourers were women, and no significant difference between male- and female-headed households in terms of their labour supply could be determined. Beyond that, the former scarcity of labour clearly was reversed and next to all people who depend on wage employment are struggling to find jobs. These findings suggest that due to the increased pressures on their lives due to the steady encroachment of capitalism and further concentration of land, many men are likely to have lost their power to control women's labour, and were forced to enter the labour market themselves. This generally indicates that the state of semi-proletarianisation is much less stagnant as many observers claim, and supports the notion that the capitalist transition is furthermore proceeding.
The overall level of differentiation becomes clear when not just labour supply, but also its demand side is taken into the picture. On average, hiring HHs are more than four times wealthier than non-hiring HHs, and their average land holding is twice as large. More interesting is that their incomes derived from crop sales are nearly seven times higher than those of HHs that rely on family labour only. This is a clear indication for accumulation that is based on the hiring of kibarua. The resulting class structure of the observed communities becomes obvious. However, to determine its exact composition is anything but trivial, as it may depend greatly on the applied criteria and definition of class, which has to go beyond the self-evident aspect of labour market participation. To illustrate this point, Table 1 lists possible class distributions across the sample of 151 HHs, according to different sets of class definitions.
| Definition set | A | B | C | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Count | % | Income (TSh) | Land (acres) | Count | % | Income (TSh) | Land (acres) | Count | % | Income (TSh) | Land (acres) | |
| Rich farmers | 36 | 24 | 1,347,001 | 3.72 | 43 | 28.5 | 1,576,164 | 3.73 | 19 | 14 | 2,585,750 | 5.18 |
| Middle farmers | 29 | 19 | 1,015,117 | 2.16 | 22 | 14.5 | 461,609 | 1.65 | 16 | 10.5 | 561,359 | 1.68 |
| Poor farmers | 49 | 32.5 | 307,621 | 1.43 | 48 | 32 | 307,603 | 1.42 | 44 | 29 | 314,556 | 1.40 |
| Simple commodity producers | 34 | 22.5 | 286,083 | 1.76 | 35 | 23 | 286,724 | 1.77 | 69 | 44.5 | 446,076 | 2.02 |
| Subsistence farmers | 3 | 2 | 0 | 0.67 | 3 | 2 | 0 | 0.67 | 3 | 2 | 0 | 0.67 |
| Total/average | 151 | 100 | 680,336 | 2.18 | 151 | 100 | 680,336 | 2.18 | 151 | 100 | 680,336 | 2.18 |
| Notes: A = differentiation based on general labour market participation; B = differentiation based on kibarua labour market participation only; C = differentiation based on kibarua labour market participation only, together with additional qualifying criteria. |
In definition set A, farming HHs are categorised strictly according to their forms of labour market participation: rich farmers are those HHs that hire labour, but do not sell their labour power; middle farmers are those HHs that both hire and sell their labour power; and poor farmers are those who only sell labour power, but do not enter the labour market as employers. In addition, two other strata have been added that do not fit directly into the above definitions. Firstly the ‘subsistence farmers’, that is, those three HHs that are not participating in market exchange, neither in the labour nor in the commodity market. The other, more interesting category, we call ‘simple commodity producers’. These do not participate in the labour markets either, yet they do produce crops for the market. It is especially this last ‘class’ that should be the most recognisable and appealing to neo-populist scholars.
As can be seen in the table, this results in a relatively even distribution of the different classes (with the exemption of the ‘subsistence farmers’, which are included merely for the sake of completeness). This is not surprising given the relatively crude categorisation of definition set A. Nevertheless, a consistency check of comparing income and land holdings averages does seem to confirm ex-ante expectations, for example, that employers do have higher incomes and that poor farmers have less land at their disposal compared to simple commodity producers, which would explain why the former have to resort to wage-labour income and the latter do not. Moving to definition set B, the distinctions become slightly more revealing. Class definitions in set B are identical to those in set A, however, set B does only incorporate labour market participation on the kibarua market and disregards the market for formal employment. The rationale for this follows directly from the above argument that – strictly speaking – formal employment does not emanate from the rural economy, and secondly, that from the villagers' perspective it rather constitutes, as they say, a form of ‘accumulation’ (which clearly can be described as such only in a strictly practical, but not theoretically consistent sense). Compared to set A, this leaves the classes of the poor farmers and simple commodity producers relatively unchanged as they basically had no access to formal employment in the first place, however, the extent of differentiation between middle and rich farmers has increased, allowing for this collateral form of ‘accumulation’. Set B nevertheless still incorporates some inconsistencies, since it accounts for any level of labour market participation, meaning for instance that it would regard any farmer as ‘rich’ as long as he hires any labour, regardless of his income or overall reliance on wage-labour inputs in the production process.
For this reason, an extended set of definitions (C) is necessary which comes considerably closer to the original definition in the Kautsky-Lenin sense. It includes that ‘rich farmer’ households should have a minimum income (for example, in this case at least the average annual household income: TSh680,000 ≈ US$525) and minimum wage expenditures (at least TSh50,000). Accordingly, poor farmers only qualify if they fulfil criteria of minimum dependence on wage incomes, and conversely middle farmers should chiefly be family reliant and net sellers of labour power. In doing so, the class differentiation within the sample becomes the most clear and separating. However, at the same time the, in this sense, ‘residual class’ of simple commodity producers has grown considerably, now making up nearly half of the whole sample. Although this is not immediately surprising, it illustrates how sensitive survey data can react to an analysis of socio-economic strata and dynamics. This does not represent a class division of any real sort, since the majority of the ‘simple commodity producers’ of set C are of course regularly engaging in the labour market, to however petty an extent, and they are without doubt exposed and subjected to the forces of capitalist markets and relations. This simple example therewith neatly illustrates the ultimate antagonism between the abstract and the concrete.
If anything, this shows that the borders between peasant differentiation and an undifferentiated peasantry depend on the particular interpretation of particular classification criteria of the relevant groups, and the hopelessly positivist nature of such an attempt in the first place. It also shows how arbitrarily set definitions can greatly alter the outcome of the analysis, and that whichever set is chosen a multitude of important factors are doomed to be omitted, like migration, remittance flows, off-farm business, use of advanced inputs and technologies to name only a few.
Finally, and to bring home this point, it is also worth mentioning that a classification based on income quintiles – as is common in many economic analyses – is hardly able to yield any illuminating insights about the socio-economic relations within society, as it does not reveal anything about the systematic reasons behind ‘livelihood diversification’. One illustrative example for this: a query for the number of ‘HHs with off-farm incomes’ releases a nearly even distribution across all five quintiles, devoid of any explanatory power and creating the misleading illusion of homogeneity.
But, with the present analysis at hand, we can conclude that peasant differentiation and capitalist relations are a clear reality in the Usambara Mountains, and for that matter also in many other rural areas of Tanzania – as indicated by a few other studies (for example, Ponte 2000, 2002). However, this is not to say that the ‘neo-populist’ element of smallholder farming with little recourse to wage-labour inputs and off-farm incomes has no relevance at all to the study of rural African societies, as a visible but dwindling element of that is obviously still existent. The consequential question is how much longer will this be the case and what factors might currently be impeding the completion of the capitalist transition, which unfortunately goes beyond the scope of this paper.
Concluding remarks
In this paper, we have linked the old debate on the agrarian question with contemporary thinking on rural development and present-day conditions for farmers in rural Tanzania. The goal was to provide an argument for the relevance of often-forgotten debates and experiences in the now developed part of the world. It also revealed both the great importance, but also the great difficulties of attempting to create a more systematic account of rural development, rather than limiting oneself to merely stating the obvious (for example, ‘livelihood diversification’) without producing meaningful analysis.
It was shown how differentiation is a concrete reality in rural Tanzania, as it is an obvious reality in any process of capitalist development. This of course is a sobering insight, especially for those concerned with the farmers' wellbeing. Differentiation means increased inequality, especially in comparison to the often-upheld ideal of fairly egalitarian peasant communities. The Gini coefficient of the collected survey sample is a sinister .65, compared to the country's national Gini coefficient of .38 (URoT 2002). This of course cannot be taken as a representative account of inequality, yet it does provide an indication of the sobering conditions many farmers find themselves in today. The extensive incidence of kibarua employment despite the strong struggle against this is a further expression of this reality.
These considerations notwithstanding, it is clear that the kibarua market nowadays is central to the agricultural sector, as on the one hand, it provides casual wage incomes for most households, and on the other hand, it virtually is the only provider for agricultural labourers for larger farmers. The larger (capitalist) farmers, especially in the two richer villages depend greatly on kibarua labourers in their production process. The hiring of kibarua allows them to produce labour-intensive vegetables on a large scale, nowadays to such an extent that the West Usambaras are a major vegetable-supplying region for the Tanzanian cities. These farmers have achieved considerable wealth through horticultural farming, and they are the richest members of our sample. Their workers on the other hand, often coming from the surrounding villages on a daily basis, form the lower strata of the society and are considerably poorer. Although exploitative in its very nature, kibarua labour provides these farmers with enough income that lets them and their families survive despite their constantly depleting agricultural assets. It therefore can be regarded as the most direct form of poverty alleviation, despite the obvious element of exploitation. There must be no ignoring of the fact that capitalist relations are ever expanding in rural communities in Africa and across the developing world, and indeed one might conclude that the capitalist imperative itself – in true dialectic form – does destroy and create the farmers' means of subsistence and livelihood.
Arguably, what will determine the outcome is whether the current – in a manner of speaking – ‘incomplete capitalism’ will persist and become a vehicle for deeply entrenched poverty and stagnation, or whether it will evolve into a fully fledged ‘modern’ and competitive version that is capable of bringing about agricultural growth, surpluses and rural wellbeing to create long-lasting development for rural Tanzanians. Either way, we have to note that development is not a smooth process of gradual improvement and enrichment, but a highly frictional affair. This has been impressively shown by Cramer (2006) for the extreme case of civil wars, but it is equally true for the everyday struggles of farmers to keep up their so-called ‘traditional’ way of life which is no longer compatible with the globalised capitalist world. It probably is a very similar struggle compared to the one that European farmers went through during the European transition from feudalism to capitalism (Brenner 1976, Hilton 1976, Wood 1999). The hope can only be that unlike so many of its African peers, a comparatively peaceful country like Tanzania will be able to manage these frictions without coming off the rails as well.