The Political Economy of Sugar in Southern Africa – Introduction

In this introductory paper we review historic and contemporary development of sugar cane production across the southern Africa. We argue that the region’s sugar industry provides a useful lens through which to understand current dynamics of corporate capital and agricultural production in Africa. We identify three distinct elements of political-economic analysis: first, the operation of logics of capital investment in different settings; second, the nature of state policies and politics in different national contexts; and third, local processes of production, accumulation and livelihoods, including effects on labour and social differentiation. The paper draws on the empirical cases from seven southern African countries presented in this collection. It highlights the rapid concentration of corporate control by three South African companies over the past decade, but also a diverse set of outcomes contingent on local context. This is particularly evident in the nature of ‘outgrower’ sugar cane production which is found in all cases but constituted in different places by quite different social categories in terms of wealth and scale of production. We argue that common stereotypes of corporate investment as either ‘win–win’ or as a ‘land grab’ rarely apply. Rather, the nature and outcomes of ‘outgrower’ systems needs to be understood as a manifestation of context-specific political-economic relationships between corporate capital, national governments and a variety of local holders of capital, land and labour.


Introduction
The past decade has witnessed an upsurge in interest in the relationship between corporate capital and agricultural production in Africa. This is frequently characterised by acutely polarised debates, as exemplified by the 'land grab' literature. At issue is the role of capital investment in large-scale agriculture, the role of the state, and the balance achievable between costs, such as displacement of existing land users, and benefits such as improved agricultural output and higher incomes from agricultural employment. 1 Yet the impacts and consequences of such capital investments are highly varied. Sugar, as a dominant feature of such investments in southern Africa, is a useful lens on this and wider debates surrounding them, as explained below. The cases in this special issue -from seven countries across the region -show how negotiations with capital are highly contingent and context-specific. Looking across the cases, we argue that we can understand the varied outcomes of expanding sugar production in southern Africa through an examination of the intersections of three political-economic processes. First, the operation of the logics of capital and investment in different settings; second, the playing out of state politics in different national contexts; and third, local conditions and livelihoods, including the role of labour and social differentiation that generate different patterns of production, accumulation and investment. Across the cases, we see the re-emergence of 'outgrowers' as a central feature of new production arrangements. Yet these take enormously varied forms across the region. There are, as a result, diverse consequences for livelihood opportunities, labour conditions and gender relations. Making sense of these dynamic and complex processes reveals important implications for understanding the relationships between capital, agrarian change and economic development, under different labour regimes in southern Africa's fast-changing commercial agriculture. 2 So why is sugar in southern Africa a good lens on these themes? This collection presents a multiplicity of examples of large-scale commercial agricultural investments, allowing the assessment of such investment to be located within a wider political economy, linked to a particular and important commodity. This complements other more general work on 'land deals', which emphasises the diverse processes of land acquisition, financing and investment involved. 3 Sugar cane production covers more than half a million hectares spread across seven countries in southern Africa, and total cane harvested in the region has grown 80 per cent over the past 20 years, with significant implications for land and water use in the region. Unlike other recent corporate-led agricultural investments, sugar cane has a long history in the region, linked to long-term state support involving financing, infrastructure development and political backing. Since its origins in Natal, outgrowers linked to core estates and mills have been central to the production system. A close examination of such arrangements that have now expanded across the region allows for a critical interrogation of the 'win-win' narrative often promoted around this business model. Sugar cane production in southern Africa may be interpreted as both commercial and 'developmental': much of this growth has taken place in the poorest countries of the region, driven by commercial investments by South Africa-based companies due to the takeover, but supported by governments and donors as a means of bringing employment and skills needed by a modern agricultural and industrial sector. For national governments in the region, commercial investment has offered an opportunity to rehabilitate degraded infrastructure and productive capacity, increase direct employment and supply opportunities for local land holders in addition to augmenting its tax base while raising scarce foreign exchange from export earnings. For corporate investors, the expansion of production in 'least developed countries' in southern Africa not only provided attractive terms of access to extensive land and water resources, but has been further buttressed by short-run access to the European Union's internal market. This offered higher prices for sugar until recently, and opportunities to take advantage of the speculative stage means it is subject to political and economic decisions taken by major producers such as Brazil, India or Thailand. In the last few decades, Brazil has come to dominate the global production, producing over 700 million tonnes of cane, double that of India, and dwarfing the production from southern Africa, where the largest producer, South Africa, produces around 18 million tonnes, and Swaziland, Zambia or Zimbabwe, only four to five million tonnes each. 12 The global market has changed significantly in the last 20 years or so. European exports have declined significantly, as the European regional market has expanded, while Brazil's have grown massively, with around 25 million tonnes exported annually, responding especially to growing consumption in fast-growing economies, including China. 13 Domestic demand in Brazil is large, and a state-supported focus on biofuels and ethanol production provided a huge boost, with exports of biofuels expanding too. However, rapid expansion, including investment by large global agribusiness corporates, has resulted in large debts in the industry, and changing global economic conditions -as well as exchange rates -are putting huge pressures on Brazilian sugar businesses.
The southern African sugar industry operates as a small player in this global context. It must rely on a largely regional market, particularly with declining support from European trade agreements, and protect itself from 'dumping' of cheap sugar from elsewhere. In addition to favourable climatic conditions in much of the region, its relative competitiveness depends critically on continued long-term, state-supported infrastructural development for irrigation, growing demand in both domestic and regional markets for sugar products including ethanol, and capital-intensive technological capacity for sugar cane processing. Although operating in a very particular context and on a small scale compared to Brazil or India, southern African sugar production is one of the comparatively rare internationally competitive agro-industries in sub-Saharan Africa, making it an especially significant route to understanding the contemporary drive to commercialise agriculture through large-scale investments.
In this introductory paper we will briefly chart the historical development of sugar cane production in southern Africa and outline current dynamics, introducing briefly the different papers in the issue. We seek to understand the particular form sugar production takes by looking at the intersection of three dimensions. First, the logics of capital and the investment, production and market strategies of (mostly) South African companies; second, the national political economies across the seven country cases, and the particular roles taken by different states (sometimes in relation to international donors); and third, the local conditions, including the availability of land and water, and in particular differentiated 'smallholder' livelihood strategies, and their role as outgrowers. Different types of outgrower arrangement -a central feature of production-marketing systems across the region -are in turn explained in relation to these intersections. By taking a regional and comparative approach, we aim to illuminate the complex regional geographies of sugar, the relationships between state politics, agrarian dynamics and capital, and the diverse impacts on land, livelihoods and economy.

A Brief History of Sugar in Southern Africa
Situating contemporary dynamics within a longer history of sugar in the region is essential to understand current shifts between estate-based plantation production and different forms of planter or outgrower arrangements. Differing across the seven countries, the inherited structures of land ownership, labour regimes, market arrangements and investment patterns, as well as the close relationships between external investment capital and state politics, profoundly influence today's patterns. The following sections therefore provide a very brief history to set the scene.

From Colonial to National Regimes
Sugar production in southern Africa originated with plantations established on Mauritius at the same time as the Caribbean slave plantations that marked the early European colonial expansion during the eighteenth century. Sugar planting and processing technology was subsequently transferred by immigrants from Mauritius in the mid-nineteenth century 14 to Natal. There the authorities recognised the potential contribution of sugar to the colony's economy and supported its expansion through imposition of import controls and arrangements for indentured labour to be brought from India to work on the Natal plantations. 15 By the 1890s, sugar plantations using local conscripted labour under the prazo system 16 had also been established in central Mozambique, dominated by the British capitalist J.P. Hornung's Sena Sugar Estates. 17 Following the end of the Boer war, the growing industrial and urban market in South Africa justified the Natal government's expansion of sugar production through loans for construction of new sugar mills, 18 and demarcation of farms in Zululand for white settlers, drawn from Natal's urban middle class as well as war veterans. This marked a break with the integrated model of production where cane production and milling were organised within a single estate. Instead, capital employed in cane production and in milling formed two separate constituencies, 'planters' and 'millers', competing with one another for profit margins and for influence over government policy.
The tensions between these two, in which the millers' powers of effective monopsony were balanced by planters' unions' greater political sway with the government's Board of Trade and Industry, shaped the regulation of the industry in South Africa until the end of apartheid. Two main features of this were enshrined in the Sugar Act of 1936 and its subsequent amendments. The first related to a pricing system known as the Division of Proceeds, which allocated the national sugar revenue according to average costs incurred by miller and planter sections (with further provision of additional redistribution mechanisms for smaller planters and millers) 19 . The second related to the creation of a dual market: a protected home market characterised by centralised, quantitative controls on production together with quotas allocated to individual producers on the basis of pro rata shares of national production; 20 and an export market of fluctuating world prices into which the remainder was sold by a single industry agency. The industry's exposure to the latter resulted in periods of expansion in response to high international prices, followed by periods of chronic 'overproduction' and depressed prices.

Expansion and Modernisation
The 1930s saw expansion of cane growing areas by the Tanganyika Planting Company in 1931, and by the Rhodesia government in partnership with private interests at Triangle, and a refinery (Rhodesia Sugar Refineries) established by Sena Sugar at Bulawayo in 1935. 21 However, it was not until after the Second World War that expansion of sugar production in southern Africa took off, prompted by preferential market access to Britain for colonies and former colonies under the Commonwealth Sugar Agreement (CSA) of 1951. A further plantation was opened in 1956 at Kagera in Tanganyika, and UK sugar refiner Tate and Lyle established a sugar cane plantation in 1953 at Chirundu, on the south bank of the Zambezi, and a refinery at Ndola, as well as taking over Rhodesia Sugar Refineries. The Colonial (later Commonwealth) Development Corporation (CDC) established plantations in Swaziland in the 1950s at Ubombo and Mhlume, the latter in partnership with South African company Huletts in 1958.
Further expansion in the 1960s was shaped by South Africa's exit from the Commonwealth in 1961 (thus losing its 200,000-ton sugar quota under the CSA) and the Unilateral Declaration of Independence (UDI) by the settler regime in Southern Rhodesia in 1965. In both territories, sugar cane expanded following similar patterns: government subsidy for irrigation infrastructure in lowveld areas (Malalane (formerly Malelane) in the East Transvaal, and Hippo Valley and Triangle in Southern Rhodesia), to be used for white settlers on medium-to large-scale farms and with significant private sector investment (TSB 22 in Malalane, and Anglo American and Tate and Lyle in Hippo Valley/Triangle). These irrigated developments took place against a backdrop of more general expansion of sugar cane plantations and factory capacity within South Africa as a response to favourable international sugar prices. 23 However, UDI also saw Tate and Lyle move its cane production north to Nakambala to form the Zambia Sugar Company, and subsequent foreign investment in sugar production became increasingly focused outside the minority-government territories, and in the newly independent southern African states. Lonrho established the Sugar Corporation of Malawi and in 1969 took over the Ubombo estates in Swaziland. 24 CDC established the Vuvulane outgrower scheme in Swaziland in 1962, and also, in partnership with the World Bank International Finance Corporation and Dutch commercial partners, the Kilombero Sugar Company estate and outgrower scheme in Tanzania. 25 In contrast to the South African industry based on rain-fed production in Natal, post-war investments in sugar cane production in the wider southern African region were primarily based on irrigated production, with consequently higher potential productivity, but also higher costs, than the Natal growers. In all such cases, external capital was combined with state support, often involving considerable investments in irrigation infrastructure, as well as the provision of land.

Nationalisation, International Price Volatility and Smallholder Production
The development of sugar through the 1970 and 80s was inevitably influenced by wider recessionary tendencies in the international economy that followed 'shocks' engendered by oil price peaks and falling commodity prices. Two brief but unprecedented rises in the international price of sugar, first in 1974 and then in 1980, stimulated expansion in many countries in southern Africa and elsewhere. The effects of the subsequent prolonged decline in prices during the 1980s were increasingly differentiated according to whether producers had access to European preferential markets under the Lomé (1975Lomé ( -1999 and Cotonou (2000) trade agreements between the European Community and the African, Caribbean and Pacific (ACP) group of countries. The importance of these agreements would be further emphasised by the final breakdown of the International Sugar Agreement 26 in 1985.
Against this international market background, the sugar industry was also heavily influenced by political developments in the region. First, the regional economy was disrupted by the underlying tensions of South Africa's minority government, particularly as newly independent states opposed to apartheid explicitly sought to establish, through the Southern African Development Coordination Conference (1980), economic cooperation and security independent from South Africa. The establishment of this 'frontline' was met with a South African campaign of destabilisation and insurgency that resulted in huge human and economic damage, notably in Mozambique, where agricultural production collapsed. In terms of sugar production, Mozambique had been the largest producer outside South Africa, producing 300,000 tonnes per year in the mid-1970s, but only 25,000 tonnes per year a decade later.
Outside Mozambique, the politics of sugar production were increasingly differentiated according to the industry's role in the national economy, and in particular whether it was oriented primarily towards national or export markets. In Zambia and Tanzania, postindependence governments sought to maximise government revenues, often through state ownership of the industry, while simultaneously 'socialising' sugar production by promotion of schemes to include small-scale outgrowers, as well as keeping sugar prices low for consumers. In both countries, pan-territorial price controls were associated with the state acquiring a controlling stake in privately owned sugar companies. Despite this 'nationalisation', however, the management of the industries was contracted to private sector partners: Tate and Lyle in Zambia, and Handelsvereeniging Amsterdam (HVA) in Tanzania. The sugar industry in both countries was thus publicly owned, but privately managed. Moreover, further expansion of sugar cane production, such as Tanzania's establishment of a second Kilombero mill ('KII') at Ruembe, remained dependent on foreign funding. 27 From the second half of the 1970s to the mid-1980s, in a context of oil price volatility, plummeting commodity prices and consequent balance of payment deficits and currency instability, the nationalised industries were subject to increasing taxation. There was a proliferation of parallel sugar markets, and scarcity of foreign exchange to pay for capital equipment and spare parts. Together with growing pressure from international financial agencies for reduction of state budgets, these pressures would ultimately see Zambia relinquish consumer price controls in 1982. Tanzania followed in 1988, when prices were allowed to reflect the costs of production plus a target rate of return. Following liberalisation of prices, operating profits nearly trebled in Zambia, while Tanzania witnessed a more than sixfold increase in prices up to 1992, with tax revenues also increasing. 28 In contrast to the Zambian and Tanzanian industries, which primarily supplied their national markets, export-led sugar production in Malawi and Swaziland (predominately run by Lonrho), grew rapidly, largely due to preferential access quotas to the markets of the EU (20,000 tonnes for Malawi and 120,000 tonnes for Swaziland) and US (17,000 tonnes for Malawi). In both countries, government involvement was also strong. In Malawi, Lonrho's SUCOMA enjoyed an effective monopoly and close relations with President Hastings Banda, and influence over government-set sugar prices. A second mill and estate at Dwangwa was established in 1978 by a consortium comprising Lonrho, Press Holdings (Banda's personal company) and the state Agricultural Development and Marketing Corporation (ADMARC). CDC remained an important funder, for this and for further expansion of outgrower production managed by the newly formed Smallholder Sugar Authority. Production and exports doubled, and sugar contributed 16.5 per cent of all Malawi's foreign exchange in 1976, as well as being an important source of patronage for the president and his party. 29 In Swaziland, where almost all production earned the high prices paid by the EU market, the Swazi monarchy took a direct 40 per cent stake in Ubumbo in 1973 through the king's personal trust Tibiyo Taka Ngwane, in return for allowing the expansion of a further 2,000 hectares (ha) of sugar cane. Further growth in royal sugar interests followed in 1977: a 50 per cent stake in Mhlume was acquired from CDC, and the Royal Swaziland Sugar Corporation was established with a new mill and 9,000 ha for sugar cane production at Simunye financed by an international investor consortium. The expansion saw Swaziland overtake Zimbabwe as the region's largest producer outside South Africa, with sugar accounting for 43 per cent of the Swaziland's exports and 50 per cent of its foreign exchange. 30 During UDI in Southern Rhodesia, loss of export markets due to sanctions was initially offset by an expanding local market for sugar, and diversification at both Hippo Valley and Triangle into cattle ranching, citrus and animal feed production; and by the mid-1970s, government price support and alternative export routes for 50 per cent of the sugar output meant that overall production had surpassed pre-UDI levels. 31 Independence in 1980 saw Zimbabwean sugar gain formal access to EU and North American markets, as well as those of southern African importers (Botswana and Mozambique). However, prolonged drought during the 1980s curtailed water supply for irrigation and limited output. Triangle had initiated ethanol processing by 1980, but lack of sufficient molasses and the low prices paid by the National Oil Company of Zimbabwe limited its development. As in Zambia and Tanzania, government policy to maintain low domestic sugar prices ultimately led to opposition from the industry, whose political influence had been enhanced by the incorporation of African outgrowers in 1980 at Mkwasine Estates where 120 outgrowers were settled on 10-ha irrigated plots. Their numbers had grown to almost 200 in 1989, and in all likelihood formed a key lobby for the removal of government price controls in early 1990. 32 In South Africa, while the 1970s and 1980s marked a moment of maximum isolation from the rest of southern Africa, some trends in the sugar industry were quite similar. In particular, expansion of production (initially due to high international prices in the early 1970s) increasingly featured cane production by small-scale African outgrowers. The industry launched the 'Financial Aid Fund', a revolving credit scheme oriented towards promoting small-scale production in Bantustan areas. For the milling companies, support to small-scale growers via subsidiary 'development companies' proved particularly advantageous, since it provided a means to leverage government funding (via Bantustan agricultural development agencies) and also to claim their own 'development company' expenditures as costs to be accounted in the Division of Proceeds -effectively at the expense of the white planters. The low international sugar prices in the 1980s meant that an increasingly indebted industry succumbed to rationalisation, particularly the elimination of transport subsidies and the exit of many white planters. Government eagerness to legitimate Bantustans with the promise of 'development' resulted in further expansion of production by small-scale black growers to substitute that of white planters who were dropping out, while miller facilities grew progressively more capital intensive. 33

Contemporary Political Economy Dynamics
This history of sugar production has shaped the contemporary political economy. International capital, through a variety of routes, is well embedded in the southern African sugar sector, although with varying relationships with state politics and finance. External market conditions and the preferential access to the European market have also massively influenced the trajectory of investments, while political imperatives across the region pushed for the incorporation of 'smallholders' through a variety of outgrower schemes. These look very different to the classic Natal model of large-scale planters, and have increasingly been key to both the interests of capital and the state, and so are essential in understanding the contemporary political economy of sugar in the region.
A regional analysis by Alex Dubb follows this introduction. This examines data available for the operations of the South African company Illovo, and analyses patterns of accumulation by corporate capital across the region. It suggests that these can be explained by the intersection of two different types of relationship, broadly described as those of production and those of trade, and their varying configuration in different national contexts. Following this cross-country analysis is a series of country-based papers.
Across the seven cases, outcomes have been varied. While the logics of capital, in a relentless search for stable profit, are evident, the particular negotiations and articulation within national political economies have been as varied as the contexts themselves. This has resulted in different livelihood outcomes, and different ways that local production and labour have been incorporated through diverse 'outgrower' schemes.
Blessings Chinsinga examines the latest episode of Malawi's engagement with sugar capital through a detailed analysis of the 'Green Belt Initiative', a government programme aimed at boosting irrigated production in which processes of acquisition of 'customary' land and its reallocation as irrigated outgrower plots have heightened political tensions over social differentiation. production, but also to the state that receives so much revenue from the sector. The form of outgrowing taken, however, is varied -ranging from small-scale plot holders in irrigated schemes such as in Malawi and Swaziland, to relatively larger-scale plot holders in the land reform areas of Zimbabwe. In the Zambia and South Africa cases discussed in this special issue, 'outgrowers' often do not actually do the growing, but are incorporated into 'shareholder' arrangements, as absentee tenants. Detailed investigations of these arrangements challenge the simplistic 'win-win' narratives around outgrower-estate relationships sometimes promoted as part of an advocacy of large-scale investment in commercial agriculture. 34 In all cases there are winners, and losers, but who these will be very much depends on the context, and the particular negotiations between capital and the state. While the papers on outgrowers all emphasise the importance of labour in production, and the relationships with the corporate 'estate' production, two papers focus on issues of labour on the estates themselves. Alicia Lazzarini examines the labour recruitment strategies in Xinavane in Mozambique, and how these reproduce historic patterns of labour control that belie the industry's narrative of technical modernity. Bridget O'Laughlin examines the health hazards associated with sugar cane production and the fractured politics of mobilisation around such risks in the same estate, showing how these reflect the wider challenges of collective action around labour conditions in the sugar industry, given the marginalisation of casual, often migrant, rural labour. Her paper also highlights the importance of including workers' health and the nature of non-paid 'household' labour in any understanding of the relationship between capital and labour.
The role of South African capital, and the dominance of three South African companies, Illovo, Tongaat Hulett and TSB, is a common theme across the papers. South Africa's democratic transition in 1994 and its reintegration into the regional economy -now the Southern African Development Community (SADC) -initiated a period of rapid consolidation in the southern African sugar industry, such that the entire regional production has become largely controlled by these three companies.
With the end of apartheid came not only the removal of many political barriers to South African investment in neighbouring economies, but also a convergence of economic liberalisation within South Africa and in many of the other members of SADC. In effect, South African companies were able to raise money more easily than in the past, while many SADC governments had embarked on programmes of privatisation of state assets, prompted by structural adjustment programmes insisted on by international donors and finance institutions. The result was a rapid acquisition of sugar mills and plantations. Illovo acquired Lonrho's sugar interests in Malawi in 1997, followed by Kilombero Sugar (Tanzania) and Ubombo (Swaziland) in 1998, and 50 per cent of Maragra (Mozambique) in the same year. It bought Zambia Sugar Company in 2001. Illovo itself was then taken over in 2006 by Associated British Foods (ABF), the largest UK sugar producer, which initially took a 56 per cent stake, which it increased to full ownership in 2016. Tongaat Hulett acquired the Mozambican sugar plantations and mills at Mafambisse and Xinavane in 1999, and progressively increased its stake in the Zimbabwean sugar operations at Triangle and Hippo Valley, taking full ownership in 2005.
As Illovo and Tongaat Hulett expanded northwards, the third South African sugar producer, TSB, remained strongly focused on its established base in Mpumalanga, but took a larger share of South African production, acquiring a third mill (at Pongola) and establishing a partnership with Royal Swazi Sugar. In 2012 TSB took over a concession for 30,000 ha of irrigated sugar cane and a new mill and ethanol production plant at Massingir, in Mozambique's Limpopo-Elefantes floodplain, 35 entering the regional competition 36 .
Each of the cases therefore reflects on the engagement of elements of South African capital, often in combination with other sources of international finance, with local politics and economy. In the following sections, while continuing to introduce the cases, we examine the three key intersecting explanatory themes introduced earlier: the extension of capital and company strategies; the role of the state and national political economy; and local conditions and differentiated livelihoods. Overall, we highlight the diverse ways in which outgrowing has become central to the southern African sugar sector -ways very different to the original Natal planter model. However, before turning to these three themes, in order to offer some context, we start with an overview of issues of land and water access, and patterns of production across the countries.

Land, Water and Production
Getting access to land and water is crucial for any sugar enterprise. As we discussed earlier, the acquisition of land and the construction of extensive and expensive water infrastructure in irrigated schemes was central to the establishment of sugar estates, particularly post 1945. To do this required massive state backing, and financial and political support. Today the context is different, but the needs are the same.
The principal means through which sugar companies have expanded their operations in recent years is through taking over and rehabilitating existing sugar cane plantations and processing factories, often previously state-owned (in Mozambique, Tanzania, and Zambia) but also by acquiring assets being sold (Lonrho in Malawi, Anglo American in Zimbabwe). This has reduced the costs, as well as the dangers of being accused of 'land grabbing'; although not always, as such state farms may have been occupied by peasant farmers for decades. 37 Following acquisition, an initial phase of recuperation has usually raised production within the existing plantation area and factory infrastructure, making use of existing water resources. This is indicated in the rapid growth in production in all countries, except Swaziland and South Africa (see below). However, the papers in this issue suggest that there is now a second phase of expansion of sugar cane area under way, where new land -much of it under 'traditional', 'communal' tenure -is being acquired, and where new water infrastructure and sources are required. As discussed, much of this is focused upon -and legitimated by -production by small-scale outgrowers, but linked to core, nucleus estates, and land and water demands are significant. 38 Access to this new land and water for sugar requires a reconfiguration of land tenure, a notoriously contentious issue in southern Africa, 39 and one which has proven difficult to reform by legislation, as seen in the impasse over Malawi's Land Bill, 40 and the constitutional objections to South Africa's Communal Land Rights Act. 41 In practice, governments circumvent the problems by zoning areas as strategic to national development, and therefore subject to special planning powers and resulting in significant public investment in infrastructure -and often leveraging concessional loan funding, as with the EU's Accompanying Measures for Sugar Protocol (AMSP) countries support. Examples are discussed of Malawi's Green Belt Initiative 42 and Tanzania's Southern Agriculture Growth Corridor (SAGCOT) initiative, 43 both involving substantial investment in sugar. The key strategy at local level is to compensate people displaced from the land with opportunities to grow sugar cane with a guaranteed market, as in Illovo's planned allocation of 10 per cent of every 1,000 ha of irrigated sugar cane to smallholders in the Green Belt Initiative.
Access to water is central to these extensions of sugar cane area. In some instances, new land is in wetter, low-lying areas (such as in Malawi and Tanzania); in other cases new land requires new dams, canals and irrigation infrastructure. The original establishment of the colonial estates involved often massive engineering feats, costing huge sums and backed by the state. Most states in the region do not have sufficient resources for such investments today, but aid and loan programmes, under 'corridor' investments or other programmes are looked to. Constructed as 'public-private partnerships', for example under the G8's New Alliance for Food Security and Nutrition initiative, public funds are essentially bankrolling long-term infrastructure investment from which private companies can build profitable activity. 44 Questions of land and water access and equity immediately arise. The distributional consequences of large investment initiatives are often not discussed. For example, where water is limited, the appropriation of large quantities in dams, and its direction to particular estates and outgrower areas has consequences for those who miss out. Water and land scarcities can be created, while resources are diverted to new investments. 45 Table 1 summarises some basic production indicators for sugar production in southern Africa, much of it based on irrigated land. The indicators show the strong recovery of sugar production in Mozambique, Tanzania, Zambia and Zimbabwe since 1992, not only in terms of area harvested and overall cane production, but also in terms of cane output per unit area.
The data are also striking in showing the higher productivity in these countries compared to South Africa. The preponderance of rain-fed production in the traditional Natal sugar heartlands of South Africa, compared to the almost universal use of irrigation elsewhere, is undoubtedly a key factor in these productivity figures. It suggests at least one reason for the northwards expansion of South African sugar companies, and the decline, in both relative and absolute terms, of production within South Africa that contracted by about a third from 2002 to 2012. 46 This drop in South African production has been compensated by increases elsewhere in the region, so that overall regional output has remained quite level for the past decade, while production within South Africa dropped from 70 per cent of the region's sugar cane output in 1992 to 57 per cent in 2012. Figure 1 shows corporate dispersal of sugar operations across southern Africa. The data on sugar output from the 23 mills operated by the three South African companies (Illovo, Tongaat-Hulett and TSB) are assembled in Table 2 from company annual reports and government agencies. This is necessarily approximate, as sugar output capacity is constantly revised when milling and refining factories are upgraded and management seeks greater efficiency in factory operations. The three principal South African sugar producers account for nearly 90 per cent of the entire sugar output of the southern Africa region (3.9 of 4.4 million tonnes of raw sugar produced).

Capital, Markets and Profits
With this domination of sugar production by a few companies, what is driving expanded investment in the region? The economic logic of Illovo and Tongaat Hulett is illustrated by an assessment of the contribution of sugar operations in different countries to these companies' overall operating profit from sugar production (Table 3; see also Dubb, in this issue 47 ). The data include profits from 'downstream' manufacturing, including refining, distilling and electricity generation using inputs from sugar cane processing, which tend to be more highly developed in the sugar industry in South Africa. Yet the majority of operating profit from sugar production is generated by the companies' activities outside South Africa, with some operations generating disproportionately more profit than their share of the companies' overall sugar production. 48 Shifts in the European market have played a key role in the recent development of Southern African sugar production. From the late 1980s to 2007, Southern Africa exported around 400,000 tonnes of sugar to the EU each year. From 2007, Southern Africa's sugar exports to the EU grew rapidly, and by 2011 had reached around 880,000 tonnes per year. These increases were driven largely by Swaziland and coastal Mozambique, despite the near total reduction in exports from South Africa (see Figure 2).
Protection from the volatility of the international sugar market has certainly been central to the importance of access to the European market to southern Africa. This is exemplified by the halving of sugar prices from about US$600/tonne ('raw', or unrefined, sugar) in 2009 to about US$300/tonne in 2014. But over the same period the protected price of sugar in the EU market, to which many southern African sugar producers had export quotas under the Lomé Convention, was reduced from US$524/tonne to US$335/tonne as part of the removal of subsidies to the European sugar beet industry. 49 The European market nonetheless remains of central importance, particularly for Swaziland, because southern African producers in the ACP group (excluding South Africa) continue to be able to export to the EU on a 'duty-free and quota-free' basis. This is particularly true for countries where sugar production far exceeds  national consumption and stands as a key agro-industry, most notably Swaziland. Whereas access to 'super' surpluses from a lucrative and protected European market under the Lomé Convention once formed a key area of political competition and negotiation for sugar-producing ACP countries, its reform has instead seen Southern Africa's position shift increasingly towards that of a 'low-cost' supplier. Given growing international competition, from Brazil, India and elsewhere, along with uncertainties about the future of the EU market, there are significant incentives not only to minimise costs and expand production, but also to seek and consolidate alternative markets. Certainly, the supply implications of EU market reform were a key motivation for European sugar companies to acquire stakes in southern African businesses, including the purchase of Illovo by ABF, and also the involvement of French sugar company Tereos in rehabilitating Sena Sugar in Mozambique. However, the European Commission projected in 2013 that over    Malawi  11  17  24  48  39  -----Zambia  10  23  22  25  30  -----Mozambique  2  5  0  6  2  10  19  41  29  18  Swaziland  9  13  13  8  14  5  3  6  5  8  Tanzania  4  7  0  5  1  -----Zimbabwe  -----27  38  7  44  36  South Africa  56  34  39  8  14  58  39  46  21  37  USA  7  -2  -------the next decade imports of sugar into the EU would fall by half, 50 and further reported in May 2016 that imports from ACP countries had already fallen by 22 per cent between 2013 and 2015, 51 suggesting that the longer-term market for southern African sugar will lie elsewhere, with greater reliance on domestic and regional markets. Many national consumer markets for sugar in southern Africa are growing strongly and are seen as an important source of sales growth. Prices will be lower than previously in the EU, but are likely to be higher than in the world market, if they can be 'protected' from cheap imports, blamed in recent years for reducing sales and profits in South Africa, Zimbabwe and Tanzania. The sugar industry in southern Africa continues to lobby hard for protection of its 'home' markets and, in 2014, Tongaat Hulett's chairman noted with satisfaction that: 'Going forward, measures implemented by … governments will substantially curtail imports'. 52 Specifically, in South Africa, the reference price was increased from US$358 per tonne to US$566 per tonne, while in Zimbabwe no raw sugar imports would be allowed and sugar imports would be taxed at 10 per cent plus US$100 per tonne. 53 The integration of southern African production with European refining and consumption is increasingly in competition with the potentially more lucrative prospects of consolidating domestic markets.
Diversification of income is a key element in managing market risk. Ethanol production from sugar has been a long-standing strategy to generate greater value from sugar, strengthened over the past decade by EU and US mandates to include biofuel ethanol in gasoline fuels for  vehicles. 54,55 Until recently, much of the focus of ethanol output has been on export markets, but the widespread adoption of legal requirements for ethanol blends within southern Africa (South Africa, Tanzania, Mozambique and Zimbabwe have made announcements over the past three years) means increasing scope for sale of ethanol within the regional market. A requirement for one per cent ethanol in South Africa's market for petrol fuel is estimated to be equivalent to 180,000 tonnes of exported sugar, 56,57 approaching one per cent of total sugar output in South Africa. Certainly, investment in ethanol production appeared to be accelerating, with new distilleries opening in Tanzania (Illovo), and attracting new investors in Mozambique (TSB) and Zimbabwe (Green Fuels; see Scoones et al., in this issue 58 ), although the collapse in oil prices in 2014-15 may slow the pace of such developments for a time.
Electricity generation for sale to national grids is also an increasing avenue for revenue generation. Mills typically generate 90 per cent or more of their own electricity requirements by burning the cane residue following sugar extraction. Improving efficiency, and the possibility of retrieving more biomass by not burning the cane in the field before harvesting, raise opportunities for producing surplus electricity for sale. All three major sugar producers are increasingly using their sugar production as a feedstock for a biochemical industry with a diverse range of products. This is already evident in Illovo's 'more than sugar' pitch to investors in its 2014 annual report, on the basis of a range of sugar cane-derived 'downstream products' for both food and non-food industries. Land is also a significant part of wider business portfolios. Acquired in the past for sugar cane production, often with state support, areas are now being sold off as part of real estate deals. 59 The expansionary investment drive in search of profit (as well as subsidised market access) underlies the business strategies of the core actors, but how this plays out is highly dependent on national contexts, and the role of different states in the region, as we discuss next.

National Political Economies: The Role of States
The state has been present as an essential actor in the establishment and growth of the sugar industry throughout its history in southern Africa. This continues to be the case, as the papers demonstrate. Sugar is a major contributor to export earnings (18 per cent in Swaziland, 9 per cent in Malawi, 6 per cent in Zambia) and to formal employment in agriculture. Where the state has privatised previously state-owned industries (as in Mozambique, Tanzania and Zambia), it has often retained a minority ownership stake. The state has been a crucial partner in regulating market conditions for the sugar companies; for example by legislating minimum prices in national markets and biofuel blend requirements in petrol, and in allowing -or notnew entrants to invest in sugar production. And the state continues to provide land (including through land redistribution) and support infrastructure investment, if not in the same way as the massive colonial interventions in irrigation schemes. The papers provide details of such state investments, alongside examples of the state's role in providing a political framework for investment in land and water through 'zoning' of areas for development (with attendant impacts on land and water rights); for example in Tanzania's SAGCOT corridor 60 or in Malawi's Green Belt Initiative. 61 This may involve public investment in infrastructure, such as roads, dams and canals, as well as tax and investment incentives. In the case of the EU's ten-billion-euro AMSP programme, designed to compensate ACP countries for the reduction in EU sugar prices, national governments play a central role, together with sugar companies, in enabling expansion of sugar cane areas for small-scale outgrowers. In Zimbabwe, for example, such support has been central to the creation of outgrower arrangements following land reform. 62 Finally, governments are ultimately involved in mediating between sugar companies and their local labour force, whether in terms of minimum wages for labourers, health and safety issues, contract terms for outgrowers or dealing with public disorder and violence arising from disputes (see Lazzarini and O'Laughlin, both in this issue). 63 The negotiations with capital by states often centre on the developmental gains that such engagements can bring. Industry discourses are replete with claims of 'empowerment', 'partnership' and 'social responsibility', and the extension of outgrower schemes, in all their variations, has been a response to this. As James and Woodhouse show in respect of South Africa, the 'economic empowerment' discourse is central to the Nkomazi story, linked first to outgrowers and more recently to beneficiaries of land restitution. 64 Similarly, Scoones et al. show that in Hippo Valley, Zimbabwe, the link between new outgrower arrangements and land reform is central, although for a rather more elite group. 65 In contrast to the Nkomazi case, in Hippo Valley the new outgrowers are becoming an increasingly important part of the overall production system, whereas in Nkomazi there is a crisis of production and livelihoods among the smallholders, and the estate is increasingly taking back production. Despite these varied dynamics, the claims by companies -and their extensive corporate social responsibility messages 66 -emphasise the mutual benefits between capital and smallholder production, with no mention of the tensions, conflicts and often contrasting outcomes.
For states in southern Africa, the sugar industry plays a strategic role in national economies, representing for some countries a very significant proportion of overall economic activity. Such industries require long-term investments, yet are highly risk prone, subject to the vagaries of weather and markets. However, for many states in the region, such operations are 'too big to fail' and the state therefore plays a critical role in providing support, both economic and political. 67 As a result, the industry is also embedded in state politics, and with this a politics of patronage. This national politics of the sugar industry is explored in a number of papers: from the politics of land reform in Mpumalanga 68 and Masvingo, 69 to the investment politics around the SAGCOT corridor in Tanzania, 70 to the 'food security' rhetoric of Malawi's Green Belt initiative. 71 These examples mirror others from the region, including the manoeuvring of the Mozambican state to bring investment and outgrower opportunities to undermine political opposition in areas where opposition parties are strong. 72 Such investments are therefore very much part of state-business alliances, linked to an elite politics, framed by the context of the neoliberal era when states must seek investment and finance from all sources. The result is an intertwining of South African investment (and international finance), with development 'aid' (through in particular EU support), with wider state politics and patronage relations. These bear with particular weight upon the transformation of land and water rights.
Sugar companies' ability to secure land and water access is highly dependent on their ability to work with the state, as discussed earlier. This requires high-level elite pacts between largescale business and political elites. Sometimes these are uneasy, requiring continual renegotiation, but the value of sugar for investment, tax revenues, and employment is so large that political differences are usually quickly overcome, and pragmatic deals are struck. Testament to this ability is the success of the sugar companies in maintaining sugar production in both South Africa (TSB) and Zimbabwe (Tongaat Hulett), despite the large-scale transfer of land ownership through land reform. 73

Livelihoods: Labour and Outgrowers
As the previous sections have shown, the political economy of sugar in southern Africa has to be understood in relation to the logics of capital and the strategy of core companies, as well as the relationships this has with states, influenced by state investments, the allocation of land and water, and wider subsidy and market access regimes. But in the end local conditions have a huge impact on what actually happens on the ground. The livelihood strategies of the growing numbers of small-scale producers engaged in sugar production as outgrowers, together with labour that supports both estate and outgrower production across the region, make a big difference.
Labour on the estates and in the outgrower farms, across our cases, is often migrant, seasonal, poorly paid, and with bad health and safety and employment conditions. Some tasks, such as burning and cutting cane, or working near the furnaces in the mills, can be highly dangerous, and result in significant negative health impacts. As Bridget O'Laughlin argues in her paper, because of the casualisation of employment, and the lack of organisational capacity within such workforces, it is difficult for workers or their unions to address their working conditions, as has been the case historically on commercial farms in Africa, despite moves to address work standards, International Labour Organization targets and so on.
In all cases there is a clear analytical separation between the outgrower as owner of the production and the farm labour employed to undertake the main operations, such as weeding, application of fertiliser, irrigation and harvesting. In South Africa, although some outgrowers employ a small number of permanent workers (e.g. for irrigation), farm work is almost all carried out by casual, and in some areas migrant, labour, much of it supplied for seasonal work by labour contractors (e.g. for cane-cutting). In Mozambique, also, migrant labour is important in sugar cane estates, as is labour recruited from 'communal areas' to work on Zimbabwe's outgrower farms. Employment patterns are also gendered, with men and women taking on different roles. Cane-cutting, for example, is conventionally a male task, while weeding is undertaken by women.
The organisation of labour on outgrower farms varies enormously. Sometimes it is a combination of family labour with some hired in piecework from nearby areas; sometimes a resident workforce is present, living in the case of Hippo Valley in Zimbabwe in labour compounds, part of the infrastructure of the former white farmers' farms. 74 Such arrangements are similar to those of the estates, where the often very large workforces are accommodated in large estate-run compounds or settlements.
These areas are subject to many controls and forms of disciplining, much of it with gendered connotations. Labour organisation and control is thus a direct reflection of the role of capital in postcolonial societies like Mozambique, where there are echoes of earlier colonial orders, but with new configurations emerging too. 75 These all have gendered dimensions. In the estates and labour compounds, for example, gender-based violence is common, particularly in contexts where there are large itinerant populations of workers. Equally, extreme forms of gendered, sometimes racialised, hierarchy are evident in the ways estates are organised, reflecting the spatial, social and political imprint of new forms of engagement with capital that shapes dayto-day lives and livelihoods of workers, and others in these often remote estate fields and towns.
Given the conditions, why then do people flock in numbers to work on estates and outgrower farms, and how can investors get away with poor labour conditions? The attraction of any form of employment in often remote and marginal parts of a country is significant. The alternatives are minimal, given low productivity in farming and lack of employment in urban areas. The benefits of access to education, health care and other services are important for many. Yet the intense global competition in the industry means relentless cutting of costs to increase efficiency. Across the world, this is having a massive impact on labour, and it is often the most poorly paid who are laid off, as a result of such efficiency drives, and through processes such as mechanisation. For example, in the five years from 2009, Illovo's permanent and part-time workforce in southern Africa was reduced by 25 per cent (around 10,000 jobs), while production continued to increase. 76 The result is changing labour regimes linked to sugar. This involves an increased casualisation of the workforce, and a displacement of work to external contractors and outgrowers, who are not subject to the same corporate requirements in labour standards. We can see therefore a differentiation between more marketised, industrialised approaches, with intensified mechanisation and labour specialisation, and more conventional, paternalistic, 'domestic government' arrangements in outsourced, outgrower areas, where resident labour, sometimes in old compounds, is employed. 77 Supporting these labour regimes there is also a class of 'footloose', fragmented labour seeking livelihoods through fragile, temporary piecework. 78 Mobile, frequently migrants from other areas, they are often a highly vulnerable group, and with no guarantee of a livelihood. Such populations are involved in key tasks such as canecutting, and are an important, often forgotten, class of labour in the southern African sugar industry. Across all labour regimes there is extensive gender differentiation. In increasingly industrialised labour regimes, skilled employment is at a premium, and women often lose out. In outgrower systems, men are usually the farm owners, sometimes working closely with their wives, while permanent and temporary workers have wage differentials, often again favouring men. Meanwhile, most of the mobile labourers are men, often migrants.
Depending on the labour regime, and the relationship between capital and labour, opportunities for resistance vary. Sugar estates have been subject to strikes, often linked to the destruction of cane through burning. Forms of unionised labour organisation also exist on estates, but casualisation and outsourcing undermine the efficacy of such collective action, and the permanent workers who are members of unions are often fully aware that working conditions on outgrower farms can be far worse. Unions thus tend not to focus on the conditions of those in casual and temporary work, and instead focus on wages for the already more privileged permanent workforce. 79 Thus, in the context of ever-increasing competition, especially from producers elsewhere in the world using more mechanised operations (particularly in harvesting), the commercial imperatives driving estate managers are increasingly harsh, making the outsourcing of costs and risks, and so the shift to outgrowing and more informal labour arrangements, a clear response.
While South African sugar cane production has always been marked by the planter-miller relationship, outside South Africa the early history of sugar was dominated by integrated estates, containing both plantation and sugar milling under the same ownership. As we have discussed, the need for expansion to increase profits and spread risks, as well as the increasing political imperative to legitimise the industry by distributing revenues more widely, has led to increasing emphasis on outgrowers as suppliers of sugar cane, a pattern seen elsewhere in commercial agriculture. 80 In most southern African countries, outgrowers now supply a proportion of sugar cane ranging from 13 per cent, for TSB's South African mills, to 20 per cent in Swaziland to 25 per cent in Zimbabwe, and up to 43 per cent for Illovo's Tanzanian mills. Further expansion seems likely to increase the proportion of sugar cane from outgrowers, as international aid financing (e.g. the EU's remaining AMSP funding) is targeted to support investments that generate outgrower opportunities. Outgrowers can make money from sugar production, as indicated by experience in Zimbabwe, 81 and (where not burdened by debt) in Swaziland (see Terry and Ogg, in this issue). 82 Yet, the 870 'A2' outgrowers in Zimbabwe, farming plots of 20 ha or more, have little in common with the 8,000 outgrowers allocated about 1.5 ha each by the Kilombero expansion in Tanzania. 83 Malawi's Green Belt Initiative envisages new outgrowers with 20-50 ha each; 84 closer, therefore, to the Zimbabwe model, but with the same dilemmas of elite capture. Among TSB's South African growers there is also evidence of a shift in size of holdings from 7 ha originally allocated in the 1990s to 16 ha, among those able to buy land from those willing to give up their plots. 85 The wide range of size of sugar cane holdings among outgrowers poses questions about the category. Not all outgrowers are the same, and across the cases discussed in this issue, they pursue very different livelihoods and hold very different class positions. There are important lines of differentiation between those farming larger areas (e.g. 20 ha and more) and those with ten ha or less. The former are likely to be full-time farmers, or absentee farmers with business and/or non-farm employment, who employ farm managers to run the sugar cane farm. 86 For these more elite outgrowers, earnings seem to be high enough to generate profits, and, most significantly, to cover the need for working capital to pay for production costs. In some instances, they have also provided opportunities for accumulation by women as well as men. For those with smaller areas -the more classic and considerably more numerous 'smallholders' -outgrower earnings from sugar cane are less likely to provide this cover, and they may be subject to increasing pressure for 'smallholder aggregation' in blocks.
This model, proposed by Illovo in Tanzania and TSB in South Africa, and central to the Magobbo scheme in Zambia (see Matenga, in this issue), 87 involves employing a manager to farm small-scale holdings as a block. The individual 'outgrowers' then receive a dividend or share of the overall profit from the block, and possibly a rent or lease payment as well. A similar arrangement has emerged in Tongaat Hulett's sugar cane area in Xinavane in Mozambique, 88 and is also an option offered by Illovo to new outgrowers taking up sugar cane holdings in Malawi. 89 An alternative, and a more traditional 'extension' model, is for the sugar company to supply services (input supply, business planning, contract services for cane harvesting etc.) to outgrowers on an individual basis (as in Zimbabwe and South Africa -see Scoones et al. and James and Woodhouse), 90 but usually only to larger producers.
The extent to which these different models of outgrower production apply varies across the cases, and reflects different deals negotiated between companies, states and outgrowers. Outgrowers are an increasingly important feature of this triangular relationship, and political pacts between interests are essential for stability. It is no surprise therefore that companies have found arrangements with relatively larger, elite outgrowers (more akin to the traditional Natal planter model) desirable, or have encouraged a more manageable 'shareholder' block arrangement in other areas.
While more experienced and entrepreneurial outgrowers appear capable of developing part of a more integrated agricultural production system, including the production of crops other than sugar cane (as in Zimbabwe, see Scoones et al. 91 ), there does, however, remain a tension between the 'efficiency' logic of maximising sugar cane yield per hectare and mill capacity utilisation -in effect the logic of 'estate' production -and the more individualised logic of managing land and water to meet a more diverse set of livelihood goals. On the one hand, the perishability of cane and the need for rapid transfer between field and mill binds outgrowers to the structural monopsony of the cane milling companies, while for many outgrowers sugar cane offers well-developed credit and input supply markets and guaranteed market for the crop that alternative crops seldom do. It is here, perhaps, that the key advantage of involvement in corporate sugar cane production lies, and provides the basis for an often tight, mutual reliance between outgrowers and companies, supported by states across the region.
With the vagaries of the international market ever present, and the changing fortunes of large corporate agribusiness capital in the region so deeply intertwined with this, we cannot predict whether this corporation-state-outgrower relationship will persist. But for now, in all its variety and differing political dimensions, this relationship dominates the southern African sugar sector, and is central to understanding its contemporary political economy.

Conclusion
This special issue focuses on the activities of three companies across seven countries in which they operate 23 sugar cane mills. Despite the overarching pattern of expansion of sugar production across southern Africa, the papers document a variety of outcomes, in which the logic of capital and investment is influenced both by state politics and national contexts, and by local livelihoods, circumstance and contingency. These three factors combine to result in a highly varied set of outcomes across the case studies. There is no one political economy of sugar in southern Africa, but many depending on the interactions of these processes.
So how should we understand sugar in southern Africa? Is the sugar industry part of a new developmental frontier in the region, transforming investment, market opportunities and livelihoods with a 'win-win' model, centred on linking core agro-industrial investments with outgrowers, as the industry (and other advocates) claim? Or is it a predatory form of capital, backed by elites and international finance, where production and market risks are transferred to vulnerable smallholders; where land and water resources are 'grabbed'; where a colonial model of exploitative estate production is at the centre, and profits are grabbed through monopoly power? The experience in southern Africa suggests that these stereotypes rarely apply. While the logic of capital results in a relentless pursuit of profit, state agency and national politicaleconomic context influence outcomes, as do local conditions, where particular negotiations, resistances, and accommodations matter. The result is diverse patterns of production and profit, together with different livelihood outcomes for very different types of 'outgrower'.
Sugar production in southern Africa is highly dependent on a small number of large commercial players, mostly with a South African base. Such companies must nonetheless manage the vagaries of international markets, and they must also negotiate for labour, land and water with a large number of national governments, and, to a variable extent, with local 'traditional' authorities and communities. We have identified some themes that run through the papers in this special issue: the dynamics of production and profit; the market context; the labour regime; the role of states and public investment; the implications for access to land and water; and the consequences for livelihoods. The papers that follow explore these themes in diverse contexts, and illuminate some of the key dynamics of current corporate agricultural investment in southern Africa. Together, the papers offer an opportunity to analyse the diverse and contingent influences of capital in a heterogeneous regional context, and therefore to gain an understanding of the intersections of capital, agrarian change and development.