DEBATES ON LAND GRABS AND LAND IN A MOZAMBICAN CONTEXT

The BAGC initiative started in 2009; the same year as ‘land rushes’ both in Africa and at a global scale induced a broad debate – with international media headlines reporting on the rise and fall of a large-scale land deal in Madagascar, involving a South Korean firm and the national government. The ‘spectacular fall’ of the Ravalomanana government in Madagascar was one factor spurring international attention (Wolford et al. Reference Wolford, Borras, Hall, Scoones and White2013: 190); another was the size of the land area involved in the planned deal, reported to be 1·3 million hectares (Cotula et al. Reference Cotula, Vermeulen, Leonard and Keeley2009: 3). The international debate presented both critical analyses and mounting evidence on land acquisitions and land deals of a potentially enormous scale and speed (White et al. Reference White, Borras, Hall, Scoones and Wolford2012: 624). Authors drew attention to the convergence of crises in several fields – finance, energy and climate – in combination with the dramatic increase in world food prices in the period 2007–2008 (Scoones et al. Reference Scoones, Hall, Borras, White and Wolford2013: 472). This convergence was also used to explain why powerful transnational corporations searched for available ‘idle’ or ‘empty’ land to be used for future food and biofuel production (Borras et al. Reference Borras, Fig and Suárez2011).

From the beginning, the debate on Land Grabs was particularly concerned with corporate and transnational takeover of smallholders’ land. In an African context, large-scale takeovers also represented historical continuities with earlier colonial land rushes (Alden Wily Reference Alden Wily2012; White et al. Reference White, Borras, Hall, Scoones and Wolford2012), and FDIs were associated with ‘neo-colonial scrambles for land and resources conducted by predatory investors’ (Wolford et al. Reference Wolford, Borras, Hall, Scoones and White2013: 191). An overriding concern was the size of the land areas involved. Though cautions were made on the limited reliability of available data on land acquisitions (Cotula et al. Reference Cotula, Vermeulen, Leonard and Keeley2009: 40), aggregated information of uneven quality on deals and hectares tended to be presented as facts (Edelman Reference Edelman2013: 485). Considerable uncertainty about how much land had actually been ‘grabbed’ soon led the debate into a new phase, with calls for empirical studies of the complex relationships between investors, state actors and other institutions involved (cf. Fairbarn Reference Fairbairn2013).

The existing literature on transformations in land tenure in contemporary Mozambique mostly engages with two phases in the country's modern history. After gaining independence from Portugal in 1975, and with a government established under the leadership of FRELIMO (the Mozambican Liberation Front), only a few years passed before the country was immersed in a violent civil war. When this devastating war ended with a peace accord in 1992, the situation concerning land access and landownership was unsettled, involving competing and overlapping land claims (Myers Reference Myers1994). Under FRELIMO's socialist development agenda, an agricultural sector consisting of large-scale state farms had been prioritised. As an agricultural development model it met with resistance from local people, and has been seen as one of the factors behind the protracted civil war (Geffray Reference Geffray1990; Lubkemann Reference Lubkemann2008). Towards the end of the war, the IFIsFootnote ⁵ presented a conditionality package to the Mozambican government. It was implemented through extensive structural adjustment and policy liberalisation in the country (Abrahamsson & Nilsson Reference Abrahamsson and Nilsson1995; Hanlon Reference Hanlon1996; Pitcher Reference Pitcher2002; Taylor Reference Taylor2003). The liberalisation policies also spurred the privatisation of state-farm land. The state-farm sector was in 1991 estimated to cover at least 600,000 ha (Myers Reference Myers1994: 609). Privatisation was implemented through a ‘chaotic alienation of land, including land grabbing and speculation’ (West & Myers Reference West and Myers1996: 29). Land rights were granted to private enterprises and government officials, while former state-farm land was rarely given to smallholders. After this process of privatisation and restructuration of the Mozambican economy, the major foreign private investors in agriculture and agro-processing were mainly British, South African and Portuguese (Pitcher Reference Pitcher2002: 208; cf. Hall Reference Hall2012); while FRELIMO government officials had also acquired substantial land areas.

With the Mozambican peace accord in 1992, millions of displaced people returned to their areas of origin in the countryside, and the land question was one of the urgent problems to be addressed. Research on land issues in this period (e.g. Myers Reference Myers1994; Gengenbach Reference Gengenbach1998; Waterhouse & Vijfhuizen Reference Waterhouse and Vijfhuizen2001; Negrão Reference Negrão, Santos and Trinidade2003) addressed the widely recognised need for a new national land legislation, and research findings fed into broader debates among political and civil society actors. A new Mozambican Land Law was enacted in 1997, aiming to secure access to land for the large majority of people in the countryside, including people who had been displaced during the war, while also providing regulated access to land for private investors. Thus it succeeded in settling many of the contentious issues concerning access and competing claims to land of the early 1990s. In the debate on Land Grabs starting in 2009, Mozambican land tenure relations have again been shown to involve conflicts (Ribeiro & Matavel Reference Ribeiro and Matavel2009; Hanlon Reference Hanlon2011; Tanner Reference Tanner, Kleinbooi and De Satgé2011; Cotula Reference Cotula2012; Hall Reference Hall2012; Chichava et al. Reference Chichava, Duran, Cabral, Shankland, Buckley, Lixia and Yue2013). While new deals involving transnational and foreign investors have been in focus, some studies also relate to the internal Land Grabs of the 1990s (e.g. Fairbarn Reference Fairbairn2013; Milgroom Reference Milgroom2015), when ‘government officials or others well-connected politically’ (West & Myers Reference West and Myers1996: 40, fn. 23) were able to acquire much of the best agricultural lands. It can be argued that former state-farm land for some time may serve as an ‘available-land reservoir’ for the Mozambican state when allocating land for new investment initiatives, as well as for a well-connected elite making more direct deals with external investors. When this reservoir runs low, land currently used by smallholders may easily come under increasing pressure and be targeted for FDI ventures justified by national-development interests. This can lead to new conflicts over agricultural land, as well as new challenges to Mozambican policymakers envisioning land use either as a basic resource sustaining precarious rural livelihoods, as a target for investments in large-scale agricultural development, or as a field for testing ‘hybrid’ alternatives in-between (Smart & Hanlon Reference Smart and Hanlon2014).

THE AGRICULTURAL GROWTH CORRIDOR CONCEPT

The Agricultural Growth Corridor concept was first presented by the international fertiliser company Yara in 2008, at a Private Sector Forum organised as a side event to the UN General Assembly in New York. Yara launched it as a ‘partnership concept’, with an overall aim to ‘develop underutilised land areas in Africa’ as part of the company's ‘Africa engagement’ (Yara 2016). Yara is at present one of the world's largest producers of mineral fertiliser; a production which started in Norway in 1905. Operating under the name Hydro, the fertiliser company over time diversified its portfolio both in Norway and internationally. In 2004, the fertiliser production split off and established itself as a separate company, Yara International ASA.Footnote ⁶ The same year, former Secretary-General Kofi Annan made a call for an African Green Revolution, and encouraged the creation of new partnerships to end the long-term neglect of African agriculture. Yara saw Annan's call as a timely opportunity, and came up with the Agricultural Growth Corridor concept as a private-sector response, presenting it at successive high-level events between 2008 and 2011.

As a partnership concept, the Agricultural Growth Corridor was directed at private-sector investors, as well as national public partners in Africa, but also addressed the international community of donors. Yara had already worked with the Alliance for a Green Revolution in Africa (AGRA), established in 2006 with the aim of dramatically improving African agriculture. When Yara launched the Agricultural Growth Corridor concept in 2008, it was as a partnership concept linked to a specific development model, the value-chain, within a Green Revolution framework. From the early 2000s, the value chain model has been used in business company management, as well as in investment discourses. Serving both as an operational approach and as an analytical tool (Ouma et al. Reference Ouma, Boeckler and Lindner2013), it was soon adopted by donors in international development, including agricultural development agencies. As a market-oriented, input-output model, the value chain is meant to represent the processing of a commodity from production to consumption.Footnote ⁷ In agriculture, introducing a value-chain approach implies turning subsistence-oriented farmers into (more) credible suppliers of marketable commodities. This requires promotion and facilitation of increased uses of agricultural inputs such as certified seeds and mineral fertiliser to increase productivity. Yara had obvious interests in increasing demands for fertiliser, envisioning new markets among both larger and smaller-scale farmers in Africa. In Yara's view, the ‘underutilised land areas in Africa’ needed ‘large-scale investments to shift the agricultural landscape and transform subsistence farming into viable businesses’ (Yara 2016). To get there, the proposed concept involved using PPPs as a vehicle to develop value chains in Agricultural Growth Corridors in selected regions.

The Beira Corridor in Mozambique was on the agenda when Yara, in collaboration with AGRA, in January 2009 organised a roundtable at the World Economic Forum (WEF) in Davos.Footnote ⁸ The Mozambican Prime Minister Luisa Diogo attended the presentation, and confirmed that the Government of Mozambique would be willing to come in as a public partner. In Mozambique, the Land Law of 1997 provided possibilities of regulated access to land through long-term leases (up to 99 years) for both national and international investors. In 2009, after more than a decade of growth in the national GDP, investments in domestic agricultural development were still insignificant. Among my interviewees, a former FRELIMO Minister stated clearly: ‘In Mozambique, nobody wants to invest in small-scale family agriculture’.Footnote ⁹ In the words of a businesswoman from the Beira region: ‘People here put capital into property – abroad. It's a risky environment for agriculture’.Footnote ¹⁰

The Mozambican Government had established an Investment Promotion Centre (CPI) to support and assist investors, both foreign and domestic, and further handle the approvals necessary to make FDIs in Mozambique (U.S. Department of State 2012). In 2009, the Government proceeded to create a Special Economic Zones Office (GAZEDA)Footnote ¹¹ to work specifically with the establishment of Industrial Free Zones and Special Economic Zones, following the trend of other governments around the globe ‘to link extractive frontiers to metropolitan areas and foreign markets’ (White et al. Reference White, Borras, Hall, Scoones and Wolford2012: 629). The Mozambican Government had also set up a Centre for the Promotion of Agriculture (CEPAGRI) to facilitate public-private partnerships in the agricultural sector, and give advice to foreign investors in cases where foreign direct investments would involve large-scale land acquisitions.Footnote ¹² The concept of Agricultural Growth Corridors could furthermore be linked to the already identified Development Corridors within the so-called Spatial Development Initiatives (SDIs), launched in the larger region in the 1990s.

The SDI-corridor projects had focused on transport infrastructure and non-renewable natural resource extraction for exports; the general idea being ‘to attract export-driven investments and stimulate public-private partnerships to areas with under- or unutilized potential’ (Meeuws Reference Meeuws2004: 62). The corridors connected inland regions with exploitable resources to the ports on the Indian Ocean coast in Maputo, Beira and Nacala, or were based on major waterways, such as the Zambezi and Limpopo rivers. Several of the corridors could trace their history back to transport and migration routes connecting the colonial coastal ports of Portuguese East Africa to the British colonies in the interior; to the mineral-based industries in Transvaal (South Africa) for the Maputo Corridor, to the expanding large-scale agriculture in Southern Rhodesia (now Zimbabwe) in the case of the Beira Corridor, and in British Nyasaland (now Malawi) for the Nacala Corridor. Later, the Beira Corridor had been particularly important as an outlet to the sea for the land-locked neighbour Zimbabwe. With the fall of apartheid in South Africa and the 1992 peace accord in Mozambique, the South African government soon became a driving partner for the regional SDIs, with a particular interest in developing the Maputo Corridor. By 2009, the other Corridors still lacked large-scale investment prospects. Thus, the Mozambican Government no doubt saw a convergence between the Agricultural Growth Corridor initiative and its own interests in promoting the Development Corridors. The distinctive feature of the Agricultural Growth Corridor concept in the SDI context was, however, its focus on agriculture.Footnote ¹³ That is, promoting economic growth through linking the development of commercial agriculture to other large-scale investment projects in a spatially delimited area surrounding an axis of transport infrastructure.

The professionals who were recruited by Yara in 2009 to develop the concept of an Agricultural Growth Corridor considered commercial agriculture in this part of Africa to be ‘at the very earliest “greenfield” stage of development’ (Palmer Reference Palmer2010: 5). Greenfield investments is part of FDI terminology, referring to investments involving the establishment of new production facilities in a foreign country (Eun & Resnick Reference Eun and Resnick2007: 404). In the development of the Agricultural Growth Corridor concept, investment in ‘greenfield agriculture’ is in principle associated with high risk and low profitability (Palmer Reference Palmer2010: 8), indicating that private investments require public facilitation and funding. Conventional FDI discourse (Lee & Ries Reference Lee and Ries2016: 206) also indicates that in many developing countries ‘a critical level of aid is required to encourage greenfield investment’. Directing greenfield investment to an Agricultural Growth Corridor further presupposes the creation of a ‘special space’ – with strategies for facilitation and exceptions from at least some constraining rules and regulations. The PPP is here a vehicle aiming to facilitate this special-space treatment.

The ‘greenfield space’ can be said to bring in connotations, or an ‘imaginary’ (Shankland & Gonçalves Reference Shankland and Gonçalves2016: 36), of starting from ‘a clean slate or tabula rasa’ (Kennedy & Sood Reference Kennedy and Sood2016: 42). For local people, the same space will normally be perceived in terms of a ‘thickness’ of uses and structures, of power and meaning ‘superimposed upon or knotted into one another’ (Geertz Reference Geertz1973: 10), referring to history and conflicts, interests, rights and claims, as well as often conflicting aspirations for the future. For Yara as an agribusiness company, the cultivable land areas along the Beira Corridor basically represented resources with a potential for larger-scale agriculture, which would in turn result in increasing demands for agricultural inputs. The existing infrastructure axis of road and rail could be upgraded to carry loads of mineral fertiliser from the port of Beira to the inland region, beyond central Mozambique into Zimbabwe and Zambia. The main road in the Beira Corridor further connected the inland province of Tete, with its new mega-projects in coal mining, to Beira port (Figure 1). With a rapidly expanding coal extraction industry, there were expectations that new demands for food products would provide a market for larger-scale agribusiness initiatives in the region. Finally, Yara had its own investment plans. A key element in BAGC was the construction of a fertiliser terminal as part of a larger development plan for Beira port.

BAGC FROM CONCEPT TO PRACTICE

Based on the interest expressed by the Government of Mozambique at the WEF roundtable in Davos in January 2009, Yara had the agricultural growth corridor initiative concretised into a report with a blueprint for investments: Beira Agricultural Growth Corridor: Delivering the Potential (BAGC Report 2010). Three UK-based consulting-and-development companies were involved: Prorustica, InfraCo and AgDevCo.Footnote ¹⁴ Operating at the public-private interface, these companies were familiar with the priorities, models and discourses of both public and private actors in the field of agriculture and development. Yara itself was particularly concerned with the time factor, and wanted a complete blueprint for investments to a short deadline.Footnote ¹⁵ The BAGC blueprint report was ready – on time – in January 2010, when Yara organised a first Partnership (PPP) meeting in Maputo. Immediately after this meeting, Yara had the proposal launched at WEF in Davos. Only a month later, the Norwegian Minister of Development Erik Solheim visited Mozambique and attended another launch, at the projected site for Yara's fertiliser blender at Beira port. At this event, the public sector in the PPP was further represented by CEPAGRI (for the Mozambican Government) and the Norwegian Embassy in Maputo, while the private sector was represented by the local business association ACIS in Beira, and by Yara as lead partner in the PPP.

When I visited Beira port only a month after the launch, there were no signs of activities linked to the BAGC initiative or Yara's plans to construct a new fertiliser terminal. A few people working in the harbour had vague memories of a meeting with high-level people, but they had apparently never come back. When I met representatives from the local business association, ACIS, they said it was difficult to communicate with Yara representatives, and they wondered about their own role in the new partnership. ACIS had contributed to the process: ‘For Davos, they needed statements for support. Then there was pressure to have a proposal before Solheim came, so that he would endorse it. But the proposal is not that tangible …’Footnote ¹⁶

It turned out that after the rush to present a ready-made blueprint for the successive launches in early 2010, Yara received studies of soil mechanics and water-logging at the projected construction site for the fertiliser blender. These revealed that the costs of construction on the projected site were much higher than expected.Footnote ¹⁷ At the same time, other initiatives within the company resulted in the sale of Yara's retail marketing businesses in southern Africa.Footnote ¹⁸ With a quick shift in priorities, Yara by mid-2010 turned to a corridor project considered to involve less risk and no conflicting business commitments, the Southern Agricultural Development Corridor of Tanzania (SAGCOT). Moving swiftly, Yara was able to present a blueprint for the SAGCOT initiative in January 2011, at that year's WEF in Davos. When the Norwegian Minister of Development Solheim visited Tanzania in February 2011, he pledged further support from the Norwegian Government for the development of the SAGCOT initiative. Replicating parts of his speech at the BAGC launch in Beira less than a year earlier, the Minister asserted that: ‘The goal is increased growth, increasing food production, and to contribute to a green revolution in the country’, and continued: ‘This can be a model for other countries’ (Norwegian Government 2011).Footnote ¹⁹ Less than four years later, in September 2015, Yara opened a new fertiliser blender in Dar es Salaam, as part of the ongoing SAGCOT corridor project in Tanzania. What happened to BAGC?

LAND, PARTNERS AND MODELS FOR COMMERCIAL FARMING

In the original presentations of the Beira Agricultural Growth Corridor a recurring discursive figure was ‘over 10 million hectares of arable land available’ with ‘less than 3%’ as yet commercially exploited (BAGC Report 2010). The figure 10 million ha highlighted the agro-investment opportunities for FDIs in the Beira Corridor. Government officials in Mozambique would generally hold that there was a lot of land available in the country, and the official view was that pressure on land was not a problem. To what extent and under what conditions land is actually ‘available’ for investments has, however, increasingly become an issue of debate and contestation in Mozambique (Åkesson et al. Reference Åkesson, Calengo and Tanner2008; Tanner Reference Tanner, Kleinbooi and De Satgé2011; Shankland & Gonçalves Reference Shankland and Gonçalves2016). Borras et al. (Reference Borras, Fig and Suárez2011: 215) uses the case of the large-scale biofuel project ProCana in the south to assert that ‘the assumption about existing, available marginal lands is fundamentally flawed’. In principle, the Mozambican Land Law secures rural people's established user and access rights (Kaarhus & Dondeyne Reference Kaarhus and Dondeyne2015). On the other hand, all land is in legal terms the property of the State; which also involves a state apparatus operating as an economic actor promoting investments in land, assisting investors in finding appropriate land, and ‘relocating individuals currently occupying land designated for development’ (U.S. Department of State 2012).

During my fieldwork in 2010, representatives of the Provincial association of small farmers in Manica, UCAMA,Footnote ²⁰ were sceptical of new investments in commercial agriculture in their area, and concerned that BAGC would present a threat to members’ rights and interests in land. They held that all land in the Corridor area with reasonable access to transport facilities was already occupied; but also held that ‘… when you leave the accessible areas, there is lots of land’. The 10 million hectares figure was unknown at the local level. As presented in the BAGC Report (2010: 15), this figure was a product of the assessment presented in Table I.

Table I. BAGC area land assessment.

Source: BAGC Report (2010:15).

The BAGC assessment defined the Beira Corridor area as covering the three provinces Sofala, Manica and Tete.Footnote ²¹ It characterised 44% of the total land area in these provinces (10 million ha) as ‘arable land’ based on technical criteria such as soils and climate. The assessment found 300,000 ha of the 10 million to be of interest to investors in commercial irrigated agriculture, taking into account access to transport infrastructure. Based on a value-chain approach, the blueprint identified a number of products with commercial potential: maize, wheat, rice, soya, citrus, mango, banana and sugarcane (BAGC Report 2010). Only 90,000 ha was classified as suitable for agricultural ventures to develop value chains for these products.

As an agricultural development initiative, BAGC was based on assumptions about the benefits of economies of scale. In the broader literature on agricultural development, scale has been a much debated and contested issue.Footnote ²² In a background paper delineating a conceptual framework for BAGC, Palmer claims that: ‘… small farmers cannot generate enough value to pay for the cost of installing infrastructure with its high fixed costs’ (Palmer Reference Palmer2010: fn 12). But the BAGC Report (2010) makes it clear that neither can large-scale commercial ventures in ‘greenfield agriculture’ be expected to pay for these costs. The solution presented is access to patient capital, defined as ‘long-term capital made available by the international community on concessional terms’ (Palmer Reference Palmer2010: 11). The BAGC concept was based on both public investments and donor funding being made available to develop the necessary infrastructure to make this region attractive to FDI investors. A critical analysis of the discourse promoting BAGC shows that it highlighted the opportunities for FDI in land, with PPPs as vehicles to develop commercial agriculture. While the high-profile BAGC launches sought to attract interest among FDI investors, the BAGC Report (2010) pointed to the need for mobilising ‘patient capital’ from public and donor sources in order to make investments attractive for private investors. However, very soon Yara's own withdrawal from a role as an active investor in the initiative demonstrated that the key convenor among the private sector partners itself rather operated with ‘impatient’ capital.

From the start, representatives of some of the main donor agencies in Mozambique were sceptical about BAGC. One said: ‘BAGC cannot come here with just big concepts. There is nothing here!’ A representative of another agency held that: ‘What we lack is an overall strategic dialogue! In the Ministry, few have real contact with agriculture.’Footnote ²³ BAGC was based on a vision of convergence of private-sector interests and public-policy models for development. At the same time, the public-private-partnership model provided for a flexible organisation, permitting shifting positions for those who joined as partners in the alliance. In more concrete terms, the BAGC approach involved linking larger-scale commercial farming ventures to selected local smallholders in promising value chains. Two main models were delineated: (1) The outgrower model, with a commercial farm as a hub providing services, such as irrigation or storehouse facilities; (2) The serviced farm blocks (also called ‘in-grower’) model, with a serviced and preferably irrigated area managed by a commercial leasing company, sub-leasing farm blocks of different sizes to both larger-scale and small-scale farmers (BAGC Report 2010: 22). Both models fall under the more general category of contract farming, referring to pre-agreed supply arrangements between producers and buyers in a value chain. Historically, sub-contracting has been regularly used in Mozambique, and local farmers have been involved in pre-established arrangements to grow and deliver specified amounts of produce – e.g. sugar cane or cashew – of a specified quality at an agreed date (Vermeulen & Cotula Reference Vermeulen and Cotula2010: 39; Buur et al. Reference Buur, Mondlane and Baloi2011). As a general model, contract farming addresses some of the problems of risk and scale in agricultural value chains aiming to supply regional and global markets. As such it involves mechanisms to comply with established quality standards and timely delivery. On the other hand, contract farming entails its own problems and challenges. These may be related to enforcement of agreements, power-relations in decision-making, responsibilities for outcomes, as well as conflicts in relations of dependency; though in practice a broad variety of arrangements prevail (Smart & Hanlon Reference Smart and Hanlon2014). In this context, the BAGC Report (2010) presents a business-oriented discourse pointing to opportunities, while problems and obstacles are left out of the picture. Seeking to promote FDIs in Mozambican agriculture, Yara had commissioned a report with a blueprint for pilot projects that could deliver as ‘Fast-Track opportunities’ (BAGC Report 2010: 8), targeting investors with more ‘impatient’ capital.

PILOTING BAGC IN FDI BACKWATERS

The BAGC partnership never developed into the envisioned coordinating framework for large-scale investments in agricultural development in the Beira Corridor. Yara's shift to SAGCOT in Tanzania in early 2011 was no doubt an important factor in this regard. But a few donors stayed on. A representative of the British Department for International DevelopmentFootnote ²⁴ during an interview in 2010 held that: ‘Yara's vision is limited’. The Norwegian Embassy in Maputo set up a Catalytic Fund to finance a number of the agricultural business initiatives that the BAGC Report had identified as Fast-Track pilots. By 2011, both DfID and the Dutch Embassy in Maputo had joined as funding partners in the BAGC Catalytic Fund. The donors contracted the UK-based AgDevCo to manage the Catalytic Fund – as a preliminary arrangement until a national investment company could be established (Carnegie Consult 2015). With support from the Norwegian Embassy, a BAGC institutional structure was set up, including: (1) A Partnership registered as a civil society association in Mozambique; formally operating with a board and a secretariat; (2) The Catalytic Fund, with its own board and an investment committee, and AgDevCo serving in the key role as fund manager.

After Yara's withdrawal as a lead partner in BAGC, the geographical centre of the initiative shifted from the port of Beira to the upland province of Manica and its capital Chimoio.Footnote ²⁵ Most of the ‘pilots’ that received funding from the Catalytic Fund were also located in Manica province. They were small- and medium-scale agribusinesses, with BAGC entering as one among several sources of funding and support.Footnote ²⁶ The pilots were involved in various forms of agricultural production, including seed multiplication and wild honey collection, and small-scale soya, goats and pig-meat processing. Only four of the agribusinesses would eventually operate in the value chains identified in the BAGC blueprint.Footnote ²⁷

The ‘investees’ who were directly involved in agricultural production were basically operating on former state-farm land. These state farms had been established on some of the best agricultural land in the province of Manica. At Independence this land was occupied by colonos who had been recruited from Portugal in the 1930s and the early 1960s, and granted land – and support – from the colonial power to set up modern family farms in so-called colonatos (Newitt Reference Newitt1995: 466). From 1989 onwards, the re-privatisation of state farms in the Manica area resulted in this same land being granted to an emerging business elite of higher-level party officials and public servants. These were usually absentee landowners with residence in the national or provincial capital. Some of the farms were for a period worked by former white farmers from Zimbabwe, who moved into Manica province as a result of the country's Fast Track Land Reform and the farm invasions starting in Zimbabwe in the year 2000. However, many of the Zimbabweans encountered serious problems in running commercial farms on an economically sustainable basis in Mozambique (Hanlon & Smart Reference Hanlon and Smart2008: 71–88; Hammar Reference Hammar, Derman and Kaarhus2013). A few stayed on; one of them asserting: ‘Here you need time, to get things done.’Footnote ²⁸

For the BAGC initiative, the former state-farm land, now private-held or leased, represented a reservoir available for new ventures in commercial agriculture. Another resource drawn upon in the fairly down-to-earth development of the BAGC pilots from 2011 onwards, was the farming and business management experience possessed by a few remaining ex-farmers from Zimbabwe and some resident South Africans. In contrast to the South Africans, the Zimbabwean farmers themselves had at this point very little or no financial capital left. BAGC set out to promote and fund emerging agro-investments at a moderate scale; supporting this fairly cosmopolitan agri-business environment in the Chimoio area, without local smallholders’ land being directly targeted – or affected. Still the business initiatives that were supported by the Catalytic Fund were expected to provide income opportunities for associated smallholders, especially through contract farming. According to the information presented on the Beira Corridor website in 2015, Catalytic Fund projects had ‘directly benefitted … over ten thousand smallholder farmers’ (Beira Corridor 2015). Two years later, a DfID Project Completion Review reported that approximately 100,000 smallholders had been ‘engaged’ – if not in contract farming, through getting access to improved seeds, fertiliser or markets.Footnote ²⁹ In 2017, around half of the pilots are still operating; some are also profitable in business-economy terms. The people behind these relative successes have invested impressive amounts of commitment, struggling to operate in a local context marked by the problematic security situation resulting from the outbursts of violent political conflict between 2013 and 2016, and the general uncertainty resulting from the national political-economic crisis and the devaluation of the national currency.

MODELS, ROLES AND PRIORITIES IN PUBLIC-PRIVATE PARTNERSHIPS

In an early phase of the BAGC initiative, both Yara and the Norwegian Government played important roles; one as a ‘catalyst’, the other through provision of ‘seed money’. The role played by the Norwegian Minister of Development when demonstrating political support at the BAGC launch in Beira and pledging continued Norwegian funding for the initiative has several interpretations. It can be seen as an example of support to a global agribusiness actor – Yara – from its ‘home government’, but also as a signal of the shift in aid away from untied development cooperation. It further signalled a shift in priorities from public-sector institution building towards facilitating ‘real growth’ in the private sector. With this policy shift, the means also changed, resulting in a stronger emphasis on public-private partnerships. Yara initially assumed a key role in the BAGC initiative; then its role shifted to only a nominal partner in the course of 2010. The company's initial role was no doubt based on its interests to create an expanding market for commercial fertiliser, but also on ambitions to develop its profile as a socially responsible global agribusiness company. Using high-level events to promote an African Green Revolution through PPPs in agricultural growth corridors, Yara set out to mobilise influential actors as partners in its quest to achieve the envisioned ‘transformation of African agriculture from subsistence farming to profitable entrepreneurship’ (Yara 2011).

When accepting to join the public-private BAGC partnership, the Mozambican Government already had some experience with development corridors as SDI initiatives realised through PPPs. As these SDIs had been conceived on the background of neoliberal policy reforms and the down-sizing of the African state initiated in the late 1980s, the public partners in the SDIs primarily had the role of facilitating and ensuring an enabling environment for market actors (Taylor Reference Taylor2003; Söderbaum & Taylor Reference Söderbaum and Taylor2008a ). South African capital had contributed to infrastructure projects such as the major cross-border toll road (Maputo-Witbank) when the Maputo Development Corridor started to materialise in the late 1990s. The construction of the large aluminium smelter (MOZAL) in the Maputo Corridor was also a result of foreign investments, while the Mozambican Government contributed by defining MOZAL as an Industrial Free Zone with corresponding tax exemptions. Within the Southern African Development Community (SADC), the Maputo Development Corridor was considered a success with a ‘strong emphasis on commercial viability’ (Söderbaum & Taylor Reference Söderbaum, Taylor, Söderbaum and Taylor2008b : 42). The other development corridors in Mozambique were in the early 2000s considered less successful (Nuvunga Reference Nuvunga, Söderbaum and Taylor2008); a Trade and Transport Facilitation Audit made for the World Bank stating that Beira was ‘still waiting for a mega project like MOZAL in Maputo’ (Meeuws Reference Meeuws2004: 73). In 2009, Yara's plan for a fertilizer blender at Beira port was not precisely a mega project, but still represented FDI at a scale that the Mozambican Government could probably not decline.

Until BAGC entered the scene, development corridors in Mozambique had in practice focused on infrastructure development and prioritised the development of extractive and processing industries in ‘enclave spaces’ (Kirshner & Power Reference Kirshner and Power2015). Private-sector capital had through the PPPs secured investments in corridor infrastructure. This was a feasible approach when the private actors involved in joint ventures were multinational corporations in extractive industries and mineral processing. The Agricultural Growth Corridor concept, by contrast, aimed to involve public partners in more active roles. The government was expected to take on more responsibilities, dedicating public funds for infrastructure development. Combined with donor funding, this was expected to make the corridor attractive for private investments in commercial agriculture. Given the established SDI framework – and a more generalised lack of political prioritisation of the agriculture sector after the disintegration of the state farms around 1990 – it is not obvious that the Mozambican government was prepared to fill this role. Mozambican authorities furthermore considered another of the planned SDI corridors to have greater immediate potential. This was the Nacala Corridor, linking the deep sea port of Nacala – one of the best natural harbours on the Indian Ocean – to the fertile lands in the province of Nampula, and through Malawi to Tete province. By 2009, no development strategy for the Nacala Corridor had yet appeared (World Bank 2010: 22). In 2010, when I had the opportunity to interview several Mozambican officials about the BAGC initiative, I found that it was met with reserved enthusiasm. In an interview in the Ministry of Transport and Communications in Maputo, the official said politely that BAGC was an interesting initiative: ‘But, I must admit that Nacala has a better natural harbour …’.Footnote ³⁰

In the following years, the Nacala Corridor alternative soared as a large-scale – and soon widely contested – complex of development plans and PPP alliances. It included infrastructure development in Nacala port; a transport line from the harbour to the coal extraction industry in Tete province, where the Brazilian company Vale was heavily involved in coal extraction; in addition to the ProSavana Programme with plans for agricultural development along the Nacala Corridor. The historical roots of ProSavana was the cooperation between Japan, the USA and Brazil to transform the Cerrado grasslands on the Brazilian Plateau into a highly productive agricultural frontier, producing soybeans for the global market (Wolford & Nehring Reference Wolford and Nehring2015). In 2010, the Japan International Cooperation Agency (JICA) and the Brazilian agricultural research institute (Embrapa) entered a new Japan–Brazil–Mozambique Tripartite Cooperation on Tropical Savana Agriculture. The ‘Cerrado model’ itself was contested within Brazil, especially among civil society organizations such as MST – the Brazilian Landless Rural Workers Movement (Oliveira Reference Oliveira2013). As an influential member of the international union of smallholders, Via Campesina, MST had close connections with the Mozambican UNAC, and these civil-society networks and alliances contributed to make ProSavana subject to public debate and protests at a scale that the BAGC initiative never encountered.

Though sharing a number of development-model characteristics with ProSavana, the BAGC initiative in the Beira Corridor continued as a low-key PPP, without mobilising civil-society contestation. With Yara's withdrawal of investments plans in Beira port in 2010, there were – in contrast to ProSavana – no major FDI investments connected to BAGC. Neither did Mozambican authorities come up with special incentives or exemptions from state rules and regulations. The initiative basically counted on support and project funding made available by European donors. Their involvement created opportunities for a total of 17 mostly small-scale agribusinesses located in the Corridor area. It also created opportunities for development actors operating as intermediaries and facilitators at the interface between donors, private companies and public institutions. Among the consultancy-and-development companies originally involved in designing the BAGC blueprint in 2010, AgDevCo would from 2011 onwards assume an increasingly important role in the development of the agribusiness initiatives supported by the BAGC Catalytic Fund. It was able to expand its operations to a number of African countries in addition to Mozambique, also acting as a more independent investor in the agriculture sector, through providing ‘long-term, flexible risk capital, which is structured to help early-stage businesses reach profitability and scale’ (AgDevCo 2017). The company's professional role performance no doubt contributed to several of the BAGC pilot projects considering themselves as AgDevCo clients rather than as BAGC pilots; while the established PPP partnership itself was struggling with disagreements, confusion and mistrust (Carnegie Consult 2015). In 2015, the BAGC donors decided to ‘exit the BAGC Partnership and withdraw from the 14 remaining investments of the BAGC Catalytic Fund’.Footnote ³¹ It was further agreed that DfID should oversee an exit plan for the ‘catalytic’ investments managed by AgDevCo.

BY WAY OF CONCLUSION

As a ‘greenfield’ initiative to promote large-scale commercial agriculture, BAGC was introduced into a high-risk environment characterised by economic and socio-political constraints that have shaped the opportunities to develop the region's agricultural potential. Historically, the local population of small-scale farmers in this region were never the primary beneficiaries, neither of colonial-era agricultural colonato arrangements, of later large-scale state farm establishments, or of the privatisation of the same farms at the end of the Civil War in 1992. At present, Mozambique has established a legal framework, not only to protect rural people's user rights to land, but also to secure consultations with local people in cases where new investments in land and natural resources may affect their access rights and livelihoods (Tanner Reference Tanner, Kleinbooi and De Satgé2011). The extent to which this legal framework in practice serves to protect local people's rights and interests has been questioned (Hanlon Reference Hanlon2011; Milgroom Reference Milgroom2015). Still, it sets a standard. Nevertheless, when potentially large-scale investments aim to increase agricultural production, and promise to bring in expertise, technology and finance, the Government tends to see national-development interests as opposed to smallholders rights – and local-development priorities. In the global Land Grab debate, as well as in the protests against ProSavana within Mozambique, we see that large-scale FDIs in agricultural land may result in broad-based contestation. What comes up as a key question in this context is what type of capital investments is most likely to become subject to forceful demands for public accountability (Shankland & Gonçalves Reference Shankland and Gonçalves2016: 43). And here we talk of accountability demands that go beyond the companies’ own ambitions in terms of socially responsible agribusiness, and even beyond international organisations’ responsible business standards (OECD-FAO 2016).

In April 2016, Yara bought the regional distribution company Greenbelt Fertilisers, including a fertiliser blending plant at Beira port. Already in February 2017, this fertiliser facility was closed down; but in July 2017 Yara signed a new Memorandum of Understanding with the Government of Mozambique about a new fertiliser plant in the northern Cabo Delgado province, using gas from the planned offshore projects in the Rovuma Basin. Thus Yara linked its renewed presence in Mozambique to the extractive-economy sector, rather than to high-risk and potentially contested Agricultural Corridor development initiatives.

In the narrative presented here, we see that new public-private partnerships may bring up unresolved questions of accountability. Public-private partnerships aiming to promote agricultural development can in the Mozambican context easily exacerbate existing power imbalances and strengthen extractive-economy trends. When PPPs are associated with large-scale FDIs in agricultural land, as in the case of ProSavana, responses may take the form of open contestation. If agriculture in this region is particularly high-risk and low-return, the term ‘patient capital’ – a key concept in the Agricultural Development Corridor approach – itself points to some of the inherent dilemmas of neoliberal globalisation as it is played out in initiatives to develop commercial agriculture in Africa. In a PPP of unequal partners, the question of who remains with the risks, and who is left to pay the debts, becomes acute when some partners leave, subject to their own shifting priorities or the imperative of short-term profits. Others have to stay – with the risks, debts and uncertainties involved.

On the other hand, if FDI directed towards extractive-economy ventures in mining, oil and gas more easily escape widespread demands for public accountability, what implications do we see for investments in agriculture? Can we draw the conclusion that FDIs should be de-linked from agricultural-land development? In that case, national as well as international public actors need to increase their direct support and commitment to agricultural development – at appropriate scales, and in ways that take local knowledge, local investment priorities and local time horizons seriously. The experiences represented by the remaining actors in the BAGC Catalytic Fund may provide valuable lessons informing such a reorientation.