Restructuring the Swazi Sugar Industry: The Changing Role and Political Significance of Smallholders

Since Independence, the contribution of Swazi smallholder farmers to sugar production has grown greatly. This is in part due to a change in the political support that the farmers have received from the Swazi state. Initially viewed with suspicion and as a challenge to royal hegemony, smallholder sugar cane production is now seen as a crucial means of overcoming rural poverty in Swaziland’s poorest region, the semi-arid lowveld. Provision of irrigation water is vital for successful sugar cane production on the co-operatively managed farms that are necessary to achieve the economies of scale in irrigated sugar production. Threats to their profitability arising from increasingly volatile markets facing Swaziland’s sugar industry since changes to the EU Sugar Protocol in 2005 undermine their capacity for co-operative governance and challenge their long-term viability. EU support to facilitate adjustment to the new sugar market has led to increased sugar cane production by smallholders. The paper argues that the two big dam projects that are at the heart of the new irrigation regions are politically ‘too big to fail’, which has committed the Swazi state to the long-term support of smallholders to ensure repayment of loans to build the dams. At the same time, the success of commercial small-scale sugar cane production poses challenges to the Swazi model of customary land tenure. Although emblematic of autocratic royal power, it also – in the processes of land and water development for commercial smallholder sugar cane production – suggests limits to how that power may be deployed in the changing relationships between Swazi elites and their corporate partners and the diverse classes of rural stakeholders.


Introduction
Since the mid-1950s the Swaziland sugar industry has expanded rapidly. Although the smallest southern African state, by the early 1980s Swaziland had the second largest sugar industry in the region after South Africa, and, by 2014-15, the fourth largest in Africa. Over the past 20 years the sugar industry has remained Swaziland's pre-eminent economic activity. By 2014-15 it accounted for almost 60 per cent of agricultural output, 35 per cent of agricultural wage employment, and contributed about 18 per cent to Swaziland's GDP. 1 In addition, although agriculture contributed less than 10 per cent of GDP, manufacturing, dominated by sugar processing, accounted for over 42 per cent. 2 Whilst the industry remains predominantly based upon large-scale estate production (responsible for 77 per cent of production in 2012-13 3 ), there has been long-term involvement by Swazi smallholders, who have been incorporated into the industry from a subsistence or semi-subsistence agriculture. Since 2000, this has accelerated as large-scale irrigation projects dependent upon two dams have enabled thousands of farmers, who formerly relied on rainfed practices, to consolidate their individual farms into commercial irrigated sugar cane co-operatives. During this period, the area under sugar cane has grown by 28 per cent, most of which is due to smallholder expansion.
A primary aim of this paper is to explore the role of Swazi smallholders in the sugar industry and their relationship with economic, political and social factors encountered in Africa's last absolute monarchy. 4 In particular, we examine the significance of sugar for the Swazi monarchy, which is seeking to re-establish its control over two-thirds of its pre-colonial land from which the 1907 British land proclamation had excluded it. As well as reasserting its control through incorporation of previously settler-owned property, title deed land, as Swazi Nation Land (SNL), the Swazi monarchy needed also to negotiate its control of the institutions of the chiefs 5 who controlled local access to SNL under customary tenure. 6 A second aim of the paper is to the evaluate the role of the sugar industry within Swaziland's agricultural development policies, 7 following the World Bank's rediscovery of the central importance of agriculture as the most effective means of overcoming deep-seated poverty. 8 In recent years, such studies have emphasised the role of market forces and the encouragement of an entrepreneurial spirit among African smallholders as being a crucial element in successful rural development policies, 9 while downplaying the processes of social and economic differentiation that produce winners and losers from the transition to capitalist agriculture. 10 We examine the impact of these political and economic processes with respect to the incorporation of smallholders within the sugar industry over the past 50 years. The approach we take is chronological, but focusing on three questions: First, in what ways has smallholder incorporation differed over time, and how has this been manifested in social and economic relations, including the development of a distinctive labour market within the smallholder sector?
Second, how has incorporation affected political relations between smallholders, the ruling elite, including the king and traditional leaders (the chiefs), paying particular attention to the impact on rights to customary tenured land?
Third, to what extent has the expansion of the smallholder sugar sector reflected international markets and Swaziland's relationships with its major international trading partners, especially the EU?
The paper is based upon a review of secondary materials including academic, official and industry-based reports and plans. It involves a retrospective analysis of past research undertaken over a period of 20 years by both authors in their roles as independent researchers, analysts working within the Ministry of Agriculture, and as internal and external community support workers on the Komati Downstream Development Project (KDDP) and the Lower Usuthu Development Project (LUSIP). One of the authors currently works with Swazi smallholders in the sugar industry and is based in the northern sugar belt, providing a continuing local insight into recent developments, particularly with respect to contemporary political economy.
The remainder of this paper traces the development of smallholder sugar production within Swaziland. It argues that an initial phase of large-scale estate production failed to address widespread rural poverty, prompting political pressure for wider distribution of sugar cane wealth. A combination of factors, including regional water resource management and changing international markets, have also prompted expansion of sugar cane areas via smallholder schemes. However, the major investments that were necessary for this approach to succeed, particularly extending irrigation onto customary tenured SNL, create a combination of political and economic commitments that make smallholder sugar cane production politically too big to fail, linking its fate to the political elite, including the monarchy.

The Geography of the Swaziland Sugar Industry and the Importance of Land Tenure
Between 1957 and 1980, Swaziland's sugar industry represented a classic, capital-intensive agri-industrial complex funded by external capital within the colonial and postcolonial systems. Three clusters of large-scale estate sugar cane production are centred on Swaziland's three sugar mills: Ubombo on the Usuthu river; Mhlume, serviced by an 88-km canal from the Komati river via the Sand River dam; and Simunye, dependent on the Mnjoli dam on the Black Mbuluzi river (Figure 1). All are located in the semi-arid lowveld, where irrigation is crucial for successful production. In explaining the pattern of development within the industry it is necessary to identify the various types of land tenure that exist in Swaziland. Two main types of land tenure resulted from the British colonial administration's efforts in 1907 to end uncertainty over land rights due to overlapping land concessions granted to white colonists by the Swazi king. 'Title deed land' (TDL), held by settlers and companies, diminished from 62 to 28 per cent of total land during the course of the twentieth century due to acquisitions by the Swazi monarchy. SNL was land under customary tenure controlled by the king via the chiefs. SNL is held by the king 'in trust for the nation', 11 and the proportion of land in this category has increased since the early twentieth century (Table 1 A third category, Crown Land, is owned by the Swaziland government. It covers less than 0.4 per cent of Swaziland and is insignificant in the sugar industry. The early sugar cane estates were developed on TDL: British and South African capital developed the Ubombo ranches complex on TDL in the central lowveld supplying the Big Bend (Ubombo) sugar mill. In 1958, in the northern lowveld, the second cluster commenced when the Colonial (now Commonwealth) Development Corporation (CDC) and JL Hulett started production at their Mhlume sugar estate on TDL, processing its first sugar in 1960. After Independence, the expanding industry provided a vehicle for investment and land acquisition by the Swazi monarchy.
In 1968, King Sobhuza established Tibiyo as an investment fund following the reversion of mineral rights to the king 'in trust' for the nation. 16 Because of shortages of domestic capital, Tibiyo sought partnerships with foreign capital, through joint investment in new projects and the acquisition of shares in companies, including those in the sugar industry. The focus was on 12 E. van   the purchase of TDL, which boosted the percentage of SNL, but much of this was no longer under the control of chiefs. This non-customary type of SNL expanded from 2 per cent to 19 per cent of Swaziland by 1997. Much of it was developed for sugar cane, in association with CDC and other agri-businesses. Therefore, the de facto position was the creation of new largescale estates where traditional authorities, other than the king, had little or no direct influence. Therefore, it has provided the material basis for the transformation of the entire social structure of Swaziland by providing the aristocracy with an independent basis for capital accumulation. Through involvement in Tibiyo, Swazis with both royal and non-royal origin found themselves part of a new elite club. 17 The expansion of sugar was a relatively new form of rural development for Swaziland, the first joint deal to develop land with CDC having been in the forestry sector. 18 In the first four decades this policy had little to do with poverty alleviation, although wages, housing and other social services provided by the sugar estates enabled a good standard of living for estate workers compared to other rural dwellers. However, the development of Simunye, 19 the third sugar mill, and associated estates on SNL resulted in the forced removal of approximately 5,000 people. Levin notes that it was surprising how limited resistance to this was. This possibly reflected the sense of powerlessness of people residing on SNL in the face of coercive power exercised by the king through his control of the chiefs. This has taken the form of forced labour, forced financial contributions 20 and forced removal of groups considered to be either a threat to royal power or a barrier to Tibiyo-sponsored development. 21 The Royal Swazi Sugar Corporation (RSSC) was formed as a consequence of Simunye's development and both Tibiyo and the state, through the Swazi government, were major shareholders. Since 1973, when King Sobhuza suspended the constitution, the government has had limited powers. In 1978 the tinkhundla system was introduced, a form of limited democracy under which the king assumed 'supreme power'. 22 Through this system, parliament became restricted to political forces approved by the king, and its members are not allowed to monitor Tibiyo's funds.
Although compensation and suitable land and housing were supposed to be provided as Simunye was developed, 23 in practice little attention was paid to the expelled families, who lost access to valuable land and water rights. The result was impoverishment for many, and a decline in food security, only partly compensated for some by employment on the newly established sugar complex. 24 In all these areas, the key to the expansion of the sugar industry had been the development of water resources, with most of the newly irrigated areas on TDL or Tibiyoacquired SNL. As a consequence of Tibiyo's activities, by 1994, 34 per cent of cane growing and 50 per cent of milling operations had come into Swazi ownership, the majority of which is under direct royal control and not under the supervision of the government. 25 It is against this broader political and economic context that we now consider the evolution of smallholder production of sugar cane.

The Evolution of the Smallholder Sector in the Sugar Industry
Phase 1: Vuvulane Irrigated Farms 1962Farms -1990 In 1962, CDC instigated a smallholder project, Vuvulane Irrigated Farms (VIF), whose 257 26 farmers supplied its Mhlume mill from 1,203 ha of TDL within the CDC concession. 27 By joining the scheme, settler farmers obtained 20-year leasehold title to their farm, a major departure from the traditional relationship between land and farmer as it had evolved on SNL. The farmers were contracted to commit 70 per cent of the land to sugar cane. 28 VIF was supported by the colonial commissioner, who was keen to provide opportunities for 'enlightened' Swazi farmers to benefit directly in the rapidly expanding and profitable sugar industry. This mindset may be traced back to 'the same concept of "improvement" … [as that] used by British colonial agricultural officers in their many attempts to force Africans to adopt modern farming in the 1930s and 40s'. 29 However, VIF was associated with considerable political upheaval. In 1972, one of its farmers was elected to parliament as a member of the Ngwane National Liberatory Congress, which 'represented an alliance between the middle class, the working class and segments of the peasantry'. 30 Despite the economic success of the scheme, 31 this was perceived as a serious challenge to traditional authorities and ultimately the king. Subsequently, no further smallholder developments on TDL were permitted, as the concept of 'extending freehold rights to peasants challenges the authority of chiefs and hence the very foundation stones of royal hegemony'. 32 By 1982, legal title had been transferred to the Swaziland National Agricultural Development Corporation (SNADC) and the land reverted to SNL, although CDC was retained as manager. This triggered further unrest, as farmers were still required to pay rents on what had, from their perspective, become SNL, where traditionally no rent is paid. However, SNADC did not accept this argument and continued to insist that rents be paid. In 1986, Tibiyo acquired ownership. The effect of this was that until the early 1990s, no further attempt was made to incorporate Swazi smallholders into the sugar industry, which continued to be dominated by a small group of large-scale miller-cum-growers and large commercial growers (Table 2)  By 1994, the three miller-cum-growers controlled 65.5 per cent of the sugar cane area. A further six large-scale commercial growers controlled 23.4 per cent of the total. Sixteen smaller commercial estates, each less than 1,000 ha, controlled 8.1 per cent. In contrast, VIF's 257 growers, representing 98 per cent of the growers, controlled only 3 per cent of the sugar cane area. The total output of VIF was insignificant compared to the whole industry and was incorporated into Swaziland's Commonwealth Sugar Quota that lasted until the end of 1974, replaced by the EU's Lomé Sugar Protocol when the UK joined the European Economic Community, the forerunner of the EU.

Phase 2: 'Schedule D' Growers 1991-1999
Changing political and economic circumstances prompted a return to sugar cane expansion through smallholders. Swaziland's Fourth National Development Plan  argued that large-scale commercial developments on SNL such as Simunye were no longer appropriate for future investment in the sugar industry as they had created little employment relative to the scale of investment. 34 The consequence of this decision was to switch the focus for further expansion onto SNL smallholders.
In 1991, the Swaziland Sugar Association (SSA), which since 1967 had controlled the sale of sugar from all Swaziland's producers, set aside 10,000 tonnes of sugar quota for smallholder growers. This was 9 per cent of the Swazi Lomé Sugar Protocol Quota. This smallholder sugar allocation, termed the 'Schedule D' quota, was initially not taken up enthusiastically. Charles 35 cited lack of knowledge of the industry, lack of finance and inadequate technical backing as inhibiting factors, although Funnell 36 and de Vletter 37 demonstrated the willingness of such farmers to innovate. However, the pace of expansion increased after 1993 as the profitability of sugar production became apparent. By 1994, 692 ha of sugar cane had been planted by Schedule D growers, representing 2 per cent of that year's national total (Table 3). Distribution between the three mill groups was uneven: 17 of the 20 farmers' associations (FAs), operating either as smallholder shareholder companies or, in a few cases, individual growers, supplied the Ubombo Mill group, while none supplied Simunye. 38 By 1995, the Schedule D quota had been increased to 15,000 tonnes, reflecting the growing interest being shown. This was in part caused by a decline in off-farm income opportunities within South Africa between 1993 and 1996, during which time the number of Swazis employed  in South African mines declined by 26 per cent. 39 This contributed to growing political and economic tensions, as many of the unemployed returned to their farms. The increasing profitability of sugar, the decline in off-farm incomes and the concentration of ownership of the sugar industry by the political elite changed the political-economic dynamics and stimulated the smallholder sector -a major turnaround from the post-VIF situation. During this period all the major stakeholders began to see their interests align behind the policy of encouraging increased sugar production, from whatever source. 40 However, in contrast to the early phase of VIF, SNL smallholder sugar cane production remained under the control of the chiefs and, via them, the king. Of equal importance was the fact that this emerging group of richer Swazi sugar cane farmers did not create a distinctive political movement and seemed to be supportive of the status quo that was beginning to deliver them substantial economic benefits.

Differentiation among Schedule D Growers
The 692 ha of land that had been allocated a sugar cane quota and which had been planted in 1994 was entitled to produce 7,210 tonnes of sugar at the premium price. The 212 members of the Schedule D FAs earned a net income on sugar of E1,785 (US$624) per ha per annum, compared with an average of E800 (US$280) per ha per annum for dryland cotton, then the most important alternative cash crop for lowveld smallholders. This raised incomes by 123 per cent, 41 exacerbating the social and economic differentiation within SNL. 42 This model of rural development was, moreover, unable to meet the challenges faced by most people on SNL. Farmers had to be close to a perennial water source, a river, canal or reservoir, and generally within 30 km of the mill. Even if they were in the correct location, the complicated process of applying for a quota inhibited individuals and groups from moving into sugar. The financing of the loans was also daunting, given the cost of installing irrigation equipment. In practice, the smallholder sector was heterogeneous. In 1994, excluding the 21 FAs without a water permit, without which irrigation could not be obtained, the Ubombo group contained 83 per cent of the FAs, but only 56 per cent of the total membership, reflecting the fact that seven 'FAs' were composed of a single member and only four FAs contained more than 20 members. Although the average area per FA was 37.8 ha, the average of 4.1 ha of cane per grower was similar to the average of 4.5 ha for VIF growers. 43 Areas farmed by individual FAs varied from 8 ha to 146 ha, with a majority of FAs farming 20 to 50 ha, only one being greater than 100 ha.
Although the Schedule D farmers generated twice as many permanent jobs per ha as on larger estates, wages were lower and there was no provision of housing, education or health services. 44 To a certain extent, these FAs were using poorer neighbours as reservoirs of cheap labour. This exploitation of non-FA members has persisted, and is also apparent on VIF, 45 as most of the better-paid employment tends to be assigned to family members, leaving the seasonal, unskilled, low-paid work to others. This has led to a feeling of alienation of non-FA members. For example, only 16 per cent of non-FA members of the Nyakatfo community claimed they had benefited from the Komati Pilot Project ten years after it had been implemented, whilst 60 per cent claimed no impact and 20 per cent stated it had had a negative impact. 46 Developments such as the Komati Pilot Project and VIF have led to greater class differentiation within rural communities, 47 as a consequence of richer smallholders exploiting poorer neighbours when employing them as agricultural labourers, resulting in accumulation from below. 48 Smallholders and the Swazi State The Schedule D farmers lacked relevant agronomic and managerial skills, and most had little experience of working communally. Consequently, the Ubombo and Mhlume miller-cumgrowers needed to provide them with extension support. Traditionally, the industry had been the responsibility of the Ministry of Economic Development rather than the Ministry of Agriculture and Cooperatives, whose extension workers had no experience of sugar cane production. This reflected the fact that the agriculture ministry's remit was to support farmers on SNL, not TDL. As the number of SNL-based sugar cane farmers increased, this required the agriculture ministry to provide extension support, which proved to be a major challenge 49 and led to greater political involvement in the sector during its next phase of expansion, with the creation of the Swaziland Agricultural and Development Enterprise (SWADE).
The closed structure of the sugar cane market had a profound impact on the relationship of smallholders to institutions of capital and the state. Given that all Schedule D growers farmed SNL and had no title deed to the land allocated to them, they would have found it extremely difficult, if not impossible, to obtain a loan to fund capital development. The structure of the sugar industry was therefore crucial in enabling FAs to gain access to credit. The mill provided a guaranteed market for the product and was responsible for distributing the 'sugar cheque' to the growers (as they called it) and, as a consequence, credit was provided to farmers on SNL. 50 As the number of FAs expanded, development finance institutions (DFIs) such as the Swaziland Investment and Development Corporation, the Enterprise Trust Fund and Swazi Bank began to play a greater role in lending. Under the proviso that the newly established FAs or individual had acquired a quota to grow sugar, had a licence to extract water and occupied land suitable to grow the cane, and that the mill would oversee sugar cane husbandry on the newly created farms, the DFIs were willing to lend.
Although there was a sense of optimism around most Schedule D initiatives, FAs were at risk through poor husbandry, and the variability in yields between them could be high. 51 However, these developments were politically popular because the government and the king were keen that the financial benefits of the sugar industry should be spread more widely, as the scale and depth of rural poverty began to be recognised. 52 Extra sugar also added to the profitability of the mills in which the king, via Tibiyo, was a major shareholder.
The role of sugar in supporting wider rural development was pointed out by UNCTAD who stated that In the Government's opinion, a major benefit of the increased sugar production would be the resulting investment in infrastructure, particularly irrigation facilities. Income from sugar would pay for the investment, which -if sugar production eventually becomes less profitable in, say, 15 years -could be used for other crops. 53 Smallholders were to pay for the infrastructure, but could only access funds by growing sugar cane. Because sugar was the crop with a guaranteed market, it made sense to encourage farmers to enter the industry which would provide, along with government extension services and those from the private sector, the skills necessary for the new producers to become successful co-operative irrigated sugar cane farmers. However, it was also believed that, once established, the infrastructure of dams, in-field irrigation systems and improved roads and bridges, would facilitate a more resilient rural economy, even if sugar cane were later displaced by other crops. Sugar was thus argued to be a means to an end, not the end itself. Concurrently, from the perspective of the millers, including Tibiyo, as they made investments in milling capacity to process extra sugar cane, it was in their interest to keep the new growers in the industry. These concepts became the foundation upon which the next phase of smallholder incorporation was built.

Phase 3: Community-Driven Development: The Komati Downstream Development Project and the Lower Usuthu Smallholder Irrigation Project 1999-2009
By the end of the 1990s the expansion of smallholder production of sugar cane was clearly established as the only option if the industry was to expand. The lack of suitable TDL for sugar cane production and the availability of suitable soils on SNL near the mills, coupled with the negative publicity associated with the expulsion of communities during the creation of Simunye, meant that expansion via large-scale estate production was no longer politically acceptable. The growing awareness of the breadth and depth of poverty on SNL also provided a stimulus to incorporate smallholders into the industry, 54 especially as most schemes were profitable and had raised rural incomes of the FA members substantially. Therefore new developments would have a dual purpose: to promote expansion of the sugar industry and to tackle poverty through commercialisation of the subsistence farm sector. From the perspective of the king, this could be achieved without the need for land reform and therefore without relinquishing the power he held over the farmers via the chiefs. This was only possible if sugar was adopted because no other agricultural product was marketed through a closed system. However, despite the success of most Schedule D farmers, their national impact was tiny: in 1997 they represented only 0.5 per cent of SNL households. 55 Accordingly, the National Development Strategy (1999) priorities included the shift of smallholder farmers from rain-fed maize subsistence farming to irrigation-based, commercial sugar cane production, 56 supported enthusiastically by the king. 57 Overall, the SNL smallholder sector is the largest contributor to the livelihoods of the majority of the population. 58 The major constraint until the early twenty-first century was the lack of irrigation water, as the dry season run-of-river flows had already been fully allocated to existing farmers. This was to be addressed by two large-scale developments associated with dams on the Komati and Usuthu rivers: the Komati Downstream Development Project (KDDP) and the Lower Usuthu Irrigation Project (LUSIP). The lead organisation in these developments is SWADE.
The creation of SWADE in 1999 (originally Swaziland Komati Project Enterprise) was an important event in the development of irrigated sugar cane on a large scale on SNL. SWADE's role was to provide the missing social and technical elements of support identified during the previous decade. These included first, a development philosophy of community empowerment and knowledge transfer. Second, this was complemented with technical and financial advice to facilitate change from subsistence rain-fed agriculture to co-operatively managed irrigated commercial agriculture. The ultimate goal was the creation of independent businesses, placing SWADE's philosophy squarely in the market-led approach to rural development. The scale of the KDDP and LUSIP projects dwarfed any previous SNL initiatives and due to their size held considerable risks, even in the benign international trading regime in which Swaziland's sugar industry had evolved. 59 Kerr and McDonald estimated that the EU Sugar Protocol premium price had resulted in an income transfer worth 6.97 per cent of GDP to the Swaziland economy, compared with 0.56 per cent for Malawi and 0.24 per cent for Zimbabwe. 60 Nonetheless, considerable job reductions had already occurred on the large estates before the KDDP and LUSIP projects commenced, as some of them merged to achieve economies of scale. 61 In the event, in 2003 a successful challenge to the Sugar Protocol by Australia, Brazil and Thailand within the World Trade Organization set in train a reform of the internal EU sugar market that undermined Swaziland's expansion strategy and demonstrated the overreliance of the KDDP and LUSIP on the sugar industry as the main income generator. This in turn exposed the FAs when they were most vulnerable to financial shocks, whilst learning new managerial, agronomic and financial skills.
Swaziland was identified as one of the country's most at risk from changes to the Sugar Protocol. 62 The first stage of the EU reform in 2006 resulted in a 36 per cent reduction in the EU price by 2009-10, hitting Swaziland harder than any of the other countries in the region. In 2007 the EU established a new regime to support African, Caribbean and Pacific group (ACP) sugar exporters, 63 the Accompanying Measures for Sugar Protocol countries (AMSP), to be implemented through the National Adaptation Strategy (NAS) 64 . However, methods used to support different sections of Swaziland's sugar industry led to the creation of winners and losers, even within the smallholder groups on the two projects, 65 and, although the recovery of world sugar prices by 2010 enabled all FAs to survive, those with high levels of debt had earned too little to enable dividends to be distributed to members. The two projects must therefore be considered in the context of this economic shock and the measures put in place by the EU and other agencies to enable the industry to become more competitive in a liberalised sugar market.
The Komati Downstream Development Project Had Swaziland not begun construction of the Maguga dam in 1998, it is likely that South Africa would have claimed this shared water resource by building a dam on the Komati river upstream of Maguga. 66 This would not only have prevented an extension of irrigation onto SNL, but also threatened the Mhlume sugar mill and its sugar cane suppliers. Paradoxically, at the time that the decision to go ahead with the project was taken, there was considerable uncertainty about the viability of similar projects in South Africa due to the enormous capital costs. 67 However, the international context was one of general optimism about the prospects for the industry and a belief that Swaziland could not allow this opportunity to develop the impoverished northern lowveld to slip away.
The KDDP, established in 2000, was originally envisaged to provide irrigation to 7,400 ha of SNL, with 5,500 ha under sugar cane and 1,900 ha to fruit and vegetables. 14,500 people were expected to benefit within the project area. 68 The KDDP is made up of three elements: the Maguga dam, the Agricultural Project Area and the expansion of the Mhlume sugar mill. Whilst the cost of the Maguga dam was shared between the governments of Swaziland and South Africa on a 40:60 ratio, the in-field developments were funded by the farmers. SWADE and the sugar miller assisted the FAs with land development, design and installation of irrigation systems. The FAs in the KDDP area are located near to the Komati river, from which irrigation water is pumped. Later FAs were created that incorporated, amongst others, those communities that had been displaced by the Simunye sugar scheme over 20 years earlier and who, when production of sugar cane commenced, began to supply the Simunye sugar mill.
The cost of the expansion of the Mhlume mill was borne by RSSC. Originally scheduled for completion in 2006, this was postponed to 2013 due to drought and lack of finance. 69 By March 2010, 2,500 smallholders were cultivating 3,095 ha of irrigated land in the Komati Basin -51 per cent of the original target -with approximately 90 per cent under sugar cane and the rest under vegetables and other field crops. 70 Smallholders, most of who are within the KDDP, supplied 52 per cent of sugar cane processed by Mhlume and 25 per cent of that by Simunye. The smallholders were members of 29 FAs, ranging from 37 to 220 members, with an average membership of 86. 71 Compared to 1990, the industry in the northern lowveld had undergone a profound structural change, with important ramifications for the politics of sugar in the country.

The Lower Usuthu Smallholder Irrigation Project
The LUSIP's objective was 'the reduction of poverty and sustained improvement in the standard of living of the population in the Lower Usuthu Basin through commercialisation and intensification of agriculture'. 72 This contrasted with the KDDP's initial objectives, which had principally been wealth generation, though by 2007, it was realised that the KDDP also needed to address issues of deep-seated poverty, resulting in more effort to diversify production, promote home gardens and improve potable water and sanitation to the scheme members. 73 This demonstrated a flexible approach throughout the policy cycle, a factor identified as important in successful rural development projects. 74,75 The LUSIP commenced in 2003 with the start of the construction of three dams forming an off-river reservoir impounding water diverted from wet season flows in the Usuthu river. The project is designed to irrigate 11,500 ha. Of the 6,500 ha to be completed in Phase 1, around 51 per cent was achieved, comprising 3,050 ha of sugar cane and 321 ha of alternative crops. By May 2014, 15,202 people had benefited directly, with a further 5,277 benefiting indirectly. 76 Development of the project was 'extraordinarily difficult' 77 mainly due to the protracted negotiations with traditional authorities to try to bring about a more equitable distribution of the benefits to all in the project area than had been the case in the earlier phases of the KDDP, though the extent to which this has been achieved has been questioned. 78 In addition, significant redesign of the upstream works added to delays, resulting in the final cost being one and a half times the original estimate. 79 The Swaziland government contributed 55 per cent of the total funds, with the remainder coming from eight other funders. Three per cent was contributed by the beneficiaries. 80 The expansion of the Ubombo sugar mill was paid by Illovo, at the time a South African transnational corporation and largest sugar producer in Africa, now 100 per cent owned by Associated British Foods. 81 For both the KDDP and the LUSIP, the vast majority of farmers viewed the prospect of joining the schemes with great enthusiasm, whilst the chiefs were also highly supportive. A new alliance had been struck, with the king, and associated business interests along with traditional leaders, backing smallholder production in the heavily subsidised, supportive market environment of the time.

Phase 4: Financial Support in Response to Decreasing Prices: 2007-2013
In 2006, the KDDP and the LUSIP were at different stages of development as the first EU price shock hit the industry. By September 2005, the KDDP had developed 14 of the 29 FAs. However, even before the EU price reduction, only one was able to service the loans due to an appreciation of the South African rand against the US dollar and the euro, causing the sugar price to fall 25 per cent below the financially viable level, although by 2010 it had recovered to 12.5 per cent above it. 82 As the LUSIP reservoir did not fill up until the 2009 wet season, its smallholders were not affected directly as their farms had yet to commence producing sugar cane. 83 Given the increasingly challenging economic conditions facing FAs, the EU and the Swaziland government introduced grants to enable new farmers to start up production, mainly of sugar cane. An important element of EU support was the 'Smallholder Irrigation Project' (SHIP) that ran from May 2009 until November 2013 84 and, by 2014, had provided grants to 32 FAs (59 per cent) in the southern area and six (17 per cent) in the north. A further 16 FAs under the KDDP were retrospectively provided with credits equivalent to grants received by the later FAs. 85 FAs were provided with grants to cover 70-75 per cent of development costs as part of the EU's AMSP Aid for Trade package. 86 Without this support, few of the FAs on either project would have become financially viable, and continued insecurity undermined trust within the FAs and compromised co-operative management. The government recognised that if the FAs foundered, one of its main instruments for achieving rural development would collapse, with major implications for government finances dependent upon a positive income flow from the FAs to service the loans used to finance the two schemes. The EU and Swazi government grants were therefore recognition of the financial and social catastrophe that would ensue if both projects were to fail, threatening the political alliance around smallholder support that had been built. The political imperative of supporting the smallholder sugar sector was paramount.
The Economic Impact of Sugar Production on Farmers' Associations In 2013, the National Adaptation Strategy Technical Assistance (NAS TA) Business Unit published a study 87 of 89 FAs (out of a total of 105) supplying Ubombo Mill in the south, and Mhlume in the north. The results showed similar sugar cane areas per member (1.64 and 1.66 ha per member) although southern FAs were less than half the size (averaging 39 members and 63 ha per group, compared to 89 members and 147 ha in the north). In each case about half of the FAs had no residual debt. This had transformed the situation facing the KDDP associations compared to 2005, and ensured that the LUSIP associations were not faced with the same huge debt repayments at an early stage of their development. Had the EU and Swazi government grants not been available, the lack of competition between DFIs would have been a critical factor in preventing FAs from benefiting from the recovering sugar prices. The NAS TA study noted that: On the whole, the level of indebtedness of the farming companies is very low … many companies with no or modest debt and a few companies with an excessive debt burden. This situation has been greatly assisted by the EU and [Swazi government] grant scheme and the KDDP rebate. However, due to the higher price paid for sugar from 2011 many of the FAs which were developed under 100 per cent loan finance have also managed to become debt free. 88 The average profitability per FA member was US$3,418, 14.3 per cent higher than Swaziland's GNI per capita of US$2,990 in 2014, 89 and represented a sixfold increase in average incomes within the lowveld since 2007. 90 The most profitable FA earned US$12, 272 per member, four times the national average in 2014, which demonstrates the potential of well-managed FAs to deliver huge benefits to their members. However, the worst performing FA made a loss of US$291 per member in the same year. Operational, rather than financial, factors seemed to be the main weaknesses within FAs. This has been explained by a failure to replant cane ratoons 91 in a timely fashion, leading to a reduction in yield, demonstrating that even long-established FAs face challenges, but of a different type than newly established ones. The NAS TA report concluded that: The most critical factor that determines success is the quality of governance. The best performing FAs comprise a mixture of both small and large farms and demonstrate the common characteristics of being open, democratic companies with memberships that are fully engaged and informed on the workings of their companies. 92

Smallholder Sugar Cane Production and the Swazi State
The political importance of the smallholder sugar sector is apparent at a number of levels within Swaziland's dual political system. Two of SWADE's former managers have become Swazi government cabinet ministers. Individuals from within FAs are leaders in the SSA and sit on sugar planter and mill group boards, ensuring that the smallholder sector is represented in bodies previously dominated by agri-businesses. Within the KDDP, 19 FAs have formed 'the Federation', which has taken the lead in raising production efficiencies, as well as engaging proactively with the Fairtrade Foundation in a sugar scheme. The Federation operates outside the sugar industry's established structures and therefore represents a radical departure from the industry status quo, possibly reflecting a willingness of some FAs to embrace a more liberal market as an opportunity, rather than being constrained by the SSA closed market.
The role of the chiefs remains, as it was during the Schedule D period, essentially reactive. No chiefs are FA chairpersons, even if FA committee members, and they rarely intervene directly in management, other than to mediate in disputes. Since this would happen in any case, the extension of the sugar industry onto SNL has not substantially altered the relationship between chiefs and their subjects. Given the importance of internal governance as a determinant of FA profitability, the role of the chief in settling disputes quickly and encouraging frequent, well-attended meetings with transparent decision-making may be important. Some FAs have agreed to continue the traditional practice of paying the chief a tithe in the form of cash; although it is unclear how popular this is or whether individual FA members would risk objecting to the practice. Where chieftaincies overlap geographically, chiefs may be the cause of instability either between chiefs or their subjects. One such dispute held up the Komati Pilot Project by three years, 93 but all have been resolved within the customary institution.
As with the rise of Schedule D farmers, the spread of FAs on the KDDP and LUSIP projects has not been accompanied by any increase in political activism amongst FA members, either as an identifiable interest group or through the emergence of strong political leadership. Therefore, the FAs have not created a challenge to traditional authorities, which suggests that most members see themselves as winners from the developments, especially when compared to the poverty experienced before the schemes were implemented. This seems to have strengthened their conservativism, and they remain staunch supporters of the tinkhundla system, helped by the financial support they received during the EU transition period.
The challenges facing the FAs now that the majority of loans have been paid off will be different, but still pose a risk to their continued operations. The FA leadership is faced with a number of different pressures in the utilisation of profits. There is growing demand among shareholders to increase dividends as their lifestyles improve. 94 There is also pressure to invest in the farm infrastructure as this comes to the end of its useful life. The complexities of these issues demonstrate the difficulties that face the sugar industry stakeholders. With such a large and rapid increase in small businesses, and with these facing a much more challenging market than was the case for the first 50 years of the sugar industry in Swaziland, it will be more difficult to predict and control events. This has political implications for the king and the government. These economic changes have been achieved without land reform. In theory, the relationship between the king and his subjects has therefore remained unchanged. However, given the higher value of land and the level of investment in it by the farmers, it may be problematic in practice for chiefs to exercise their formal power to remove individual farmers from the land, even in a political system where the king has become increasingly autocratic.

Smallholders and International Markets
Most FAs have become successful producers of sugar cane and achieved significant increases in income, and it appears therefore that the strategy of extending smallholder sugar production to SNL has been justified. For  However, given the changes to the EU internal sugar beet market from October 2015, if supplies to the EU sugar market were to increase, they would be faced with: a situation of increased competition between raw cane sugar suppliers on the EU market … in the context of a substantial decline in EU sugar prices … This will fundamentally change the economics of ACP cane-sugar trade with the EU. 96 As a consequence, ACP producers were urged to seek new diversified markets that would become available within the EU, including higher value-sugars, Fairtrade-certified or valueadded sugar products. The response, according to the ACP-EU joint institution Technical Centre for Agriculture and Rural Cooperation (CTA), should be market driven. 97 These policies have been strongly promoted within Swaziland's NAS and could have a positive impact on FAs, who are already responding. There may also be further possibilities for the Swaziland sugar industry and its FAs. Since the 2006 reform, the EU has changed from being a net exporter to a net importer of sugar. In addition, CTA notes that 'European companies have increased imports of white sugar from ACP states (e.g. Mauritius and Swaziland), as a strategic response to the reform'. 98 This helps to explain the decision of EU-based companies to invest in ACP countries who they judge could take advantage of the more liberal EU sugar market, and the increasing interest in the whole region by sugar refining companies such as Associated British Foods. 99 However, expiry of EU sugar quotas in 2015/16 is predicted to lead to a slight decline in imports from low-cost countries such as Swaziland to the EU market, 100 so there continues to be little certainty with respect to access to the EU market beyond that time, with the likely outcome that Swaziland will become more dependent upon regional markets, which tend to be more volatile.

Conclusion
Two dynamics may be discerned in the Swazi sugar industry. First, from the perspective of investors, the key question relates to future profitability. Even after incorporating a large group of smallholders, Swaziland is ranked as one of the ten lowest-cost producers of sugar in the world. 101 With respect to the smallholder sector, although the NAS TA Business Unit report 102 is positive about the current position of the FAs, it notes that relatively small changes in price and costs would result in a drop in profitability of approximately 44 per cent for FAs. 103 The volatility of incomes in a post-EU Sugar Protocol market is a major challenge to the whole industry, but more so to FAs, where it could have a negative impact on group cohesion and co-operative governance with implications for profitability.
The challenge to the profitability of the FAs has further implications for a second key dynamic, namely the relationship between the king, the government and smallholders. The king and government have gained substantially from ownership and rents derived from the estate sector, including the sugar mills, but the development of the industry suggests a need to co-opt a smallholder sector to sustain political legitimacy. In the late 1990s, the success of the Schedule D FAs, in addition to the increasing difficulties facing the established estates in increasing output, provided an opportunity for the king and the government to support expansion of irrigated sugar cane to SNL. The sugar industry provides the only market that enables farmers without title deed to acquire credit to finance investment on their farms. That this coincided with an increasing awareness of the depth and breadth of rural poverty on SNL made the decision to invest in the KDDP and the LUSIP doubly attractive to the king and the government, as it provided evidence of action to address rural poverty at a time when the king was facing challenges by opponents concerned about the growing inequalities and political repression within the country. The government has taken a major gamble that the FAs will be profitable enough to repay huge loans to finance the dams, although the unpalatable alternative would have been the loss of the water, vital to underpinning agricultural investment and reducing rural poverty on SNL in the lowveld, to South Africa.
The government's strategy has been put at risk by changes to the EU sugar market and it is only through lengthy and expensive intervention by the EU and the Swaziland government that FAs have managed to survive. However, since 2010, there are signs that the FAs can be made competitive, but relatively small changes in prices or costs can have major impacts on profitability. These dynamics raise questions about the reaction of farmers to a sustained fall in the profitability of sugar. If, as was the case under similar circumstances in the mid-2000s, farmers sought to diversify to other crops, 104 this would have serious implications for capacity utilisation in the sugar mills and their ability to supply income to their shareholders, including the king and the government. It would also have implications for the creditworthiness of the FAs, as financing institutions would no longer be able to retrieve credit repayments at the mill as deductions from income from sugar cane deliveries by the FA members. Such a scenario could be seen as a challenge to the power of the king who granted the land to produce sugar and who, in theory, could remove FA members through his continued control of SNL via the chiefs. The practicalities of this are questionable, because many chiefs may now be in a rather ambivalent position as most are either FA members or are receiving enhanced tithes, but also because expropriating the higher value of land and other fixed assets now controlled by FAs is likely to meet stiff opposition and social disruption. Notwithstanding that a large number of FA members are older men who have traditionally been loyal supporters of the king, such support may be dependent upon that system continuing to deliver the economic advancement that has transformed their lives in the past decade. The investments made by FA members might prove to be worth fighting for if they were seen to be at risk from coercive practices that have in the past been used by traditional authorities.
The sugar industry's payment arrangements were key to enabling large areas of lowproductivity lowveld SNL to be transformed from a region where absolute poverty was the norm to one where communities have received potable water, improved sanitation and improved food security. 105 This, coupled with the training provided by SWADE (which focused on holistic business skills development), means that even if the sugar industry eventually declines, longterm benefits should continue, although any alternative use of the land will need to generate sufficient income to reinvest in worn-out irrigation equipment and other capital investment on the farms. 106 The relative size of the sugar industry in the Swazi economy highlights its major role in the politics of land and state formation in Swaziland. In this paper we have identified how the increasing importance of 'smallholder' sugar cane production is a key marker of changes in the balance of those politics. In particular, the development of sugar cane production in Swaziland has seen its transformation from a vehicle for income and capital accumulation uniquely benefiting a small elite associated with ownership and management of estate production to an industry notionally improving the livelihoods of a mass of small-scale farmers. Central to this change is the customary land tenure system of SNL, which on the one hand symbolises autocratic royal power but also, in the processes of land and water development for commercial smallholder sugar cane production, suggests limits to that power and an altered terrain on which Swazi elites and corporate partners must negotiate with the diverse classes of rural stakeholders. How these negotiations are resolved will be one of the key political legacies of the decision to support smallholder sugar production on SNL.