1In the past twenty years, Côte d’Ivoire has risen from being an insignificant producer of raw cashew nuts to becoming the world’s largest producer. In 2017, it produced 745,000 tonnes, surpassing India (700,000 T) and Vietnam (350,000 T), the second and third largest cashew nut producing countries (N’Kalo 29 March 2018; Rongead, personal communication, 9 April 2018). Cashews are a perennial tree crop that grow in the savannah region where agro-ecological conditions, especially a long dry season, are favourable to its growth. As cashews have come to dominate the agricultural economy, cotton area and production have greatly fluctuated. Figure 1 shows the trends in cotton and cashew production over the period 1990–2017. The rapid growth in cashews contrasts with the zig-zag pattern for cotton. In this paper, I argue that this contrasting pattern is shaped to a large extent by the market structures and prices that characterise each sector. I argue that the cashew filière is based on a competitive market structure characterised by intense competition for raw cashew nuts. In contrast, the cotton filière is non-competitive. It is organised around a small number of companies who hold exclusive rights to the cotton produced in ‘their zone’. While cashew prices fluctuate with demand, cotton prices are fixed. The paper examines the agricultural, historical, and political economic dimensions of these different structures and their effects on producer prices and farmer incomes.

2I begin by situating these themes within the literature on market coordination and competition. Subsequent sections focus on the history of market structures and price formation in the cotton and cashew economies of Côte d’Ivoire. The final section presents a case study of farmer perceptions and engagements with the cotton and cashew economies at the village level. The case study suggests that the cashew boom represents an agricultural diversification strategy rather than a wholesale substitution of cotton by cashews.

3This paper draws upon decades of multi-scale field research in Côte d’Ivoire. It brings together long-term research on farming systems in the Korhogo-region community of Katiali with archival research and interviews with key actors in the cotton and cashew filières. The Korhogo region of north-central Côte d’Ivoire is the epicentre of the cotton and cashew booms. Between 1981 and 2014, we conducted eight surveys in Katiali on the crops cultivated by forty households stratified by economic standing. [1] This longitudinal research shows significant shifts in crops grown and their relative importance in terms of crop area. Figure 2 illustrates the most important trends by major crop complexes in Katiali. [2] The graph shows the area in cotton see-sawing, while the cashew area rises steadily to become the number one crop. In 2014, cashews accounted for 36 percent of the cultivated area. Cotton, the second most important crop, covered 25 percent of the crop area that year. This is a major shift from 1981, when cotton dominated the crop area (44 percent) and cashews were insignificant. The following sections advance understanding of this major change in cash cropping in the Ivorian savannah. [3]

Market structures and prices

4Market structures and prices are linked to how different actors in a value chain (filière) organise themselves around the production and exchange of commodities. Poulton and colleagues call such organisational initiatives ‘market coordination’ (Poulton, et al 2003). They argue that markets can be coordinated both formally and informally. Formal coordination occurs when the state creates sector-specific rules and regulations. Private actors can also coordinate markets through their ‘formal collective organisations with decision making powers’ (Poulton, et al 2003, 522). An example of the latter is an interprofessional organisation that sanctions its members for breaking the rules of the game. Private actors also engage in informal coordination when they make decisions among themselves on how to organise and manage a filière. This latter type of coordination is most likely to occur when a small number of powerful actors dominate a sector.

Cotton and cashew production in Côte d’Ivoire, 1990–2017

Cotton and cashew production in Côte d’Ivoire, 1990–2017

This graph shows the changes in the production of cotton and cashew nuts between 1990 and 2017. The rapid growth of cashew nuts contrasts with the zig-zag pattern for cotton. These contrasting results are shaped to a large extent by the market structures and prices that characterise each sector. The cashew nut filière is based on a competitive market structure characterised by intense competition for raw cashew nuts. In contrast, the cotton filière is non-competitive. It is organised around a small number of companies that hold exclusive rights to the cotton produced in “their zone.

5With reference to the cotton economies of six African countries, Poulton and colleagues identify three different market structures based on the type of market coordination and the degree of competition. Concentrated, market-based cotton economies exist where just a few firms dominate a sector (e.g., Zambia and Zimbabwe). Concentrated, local monopoly systems are characterised by a small number of companies who control a zone or concession in which they have the exclusive right to buy cotton (e.g., Ghana, Mozambique). The third market structure, numerous small players, is characterised by the presence of large numbers of ginning companies in a country (e.g., Tanzania, Uganda). Market competition is greatest where numerous players are involved and lowest in concentrated, local monopoly systems.

6The major coordination challenge facing cotton companies and the state in all six countries centres on the provision of agro-inputs (seeds, fertilisers, and pesticides) on credit and the recuperation of these loans after the harvest. In weakly coordinated systems, input loan recovery rates are low due to ‘side selling’—when farmers obtain input loans from one company but sell their cotton to another. Most relevant to this paper, Poulton and colleagues suggest that a relationship exists between a specific type of market structure and price competition. In their case studies, producer prices for seed cotton are highest in concentrated, market-based and numerous small players systems and lowest in concentrated, local monopolies.

7This paper builds upon and extends Poulton et al’s analysis in two ways. The first contribution is comparative and theoretical. I compare the contrasting market structures for cotton and cashews in Côte d’Ivoire with emphasis on their effects on price competition. This comparative study reveals the importance of power relations, particularly the role of the state, in market coordination and competition. My emphasis on power relations contrasts with the technocratic and depoliticised discussion of market coordination in Côte d’Ivoire’s 2013 agricultural reform documents.

8The second contribution is methodological. I take a multi-scale agricultural historical approach to study the interactive effects of broad political- economic and local-level processes on the dynamics of agrarian change. I suggest that the transformation of the agricultural landscape is shaped by the interplay of market coordination and farmer decision-making. This multi-scale historical approach helps to clarify who wins and who loses from agricultural reforms and why certain paths are taken over others. [4]

Cotton development and decline

9Cotton and cashews have distinctive political economic and agricultural histories. During the precolonial period, cotton was intercropped with yam, millet, and maize in the savannah regions of West Africa (Caillié 1830). Its widespread cultivation caught the eye of French explorers and colonialists who viewed African cotton as a potentially cheap raw material that could serve the needs of the metropolitan textile industry. Despite the concerted efforts of colonial administrators to expand cotton production during the first half of the twentieth century, the colonies never supplied sufficient quantities to meet that demand (Bassett 2001, 2002). A major factor in the supply failure was the stiff competition for cotton between local traders and French merchants. Jula traders, who supplied the local weaving industry with its basic raw material, consistently offered higher prices to producers than their French counterparts. Indeed, more than a third of the cotton grown in Côte d’Ivoire between 1917 and 1947 was consumed by the local handicraft textile industry (Bassett 1995). Colonial officials sought to increase cotton production through a combination of coercion (e.g., forced cultivation) and agronomic research (Bassett 2001).

Crop complexes in the Katiali region of northern Côte d’Ivoire, 1981-2014

Crop complexes in the Katiali region of northern Côte d’Ivoire, 1981-2014

This graph illustrates the main trends of the primary crop complexes in Katiali. It shows that the area for cotton cultivation see-saws, while the area for cashew nuts rises steadily to become the number one crop today. In 2014, cashew nuts represented 36 percent of the cultivated area. Cotton, the second most cultivated crop, represented 25 percent of cultivated areas that year. This is a major shift from 1981, when cotton dominated in terms of cultivated areas (44 percent) and when the cultivation of cashew nuts was insignificant.

10Agronomists associated with the Institut de recherche du coton et des fibres textiles (IRCT) experimented with different cotton varieties with the hope of finding higher yielding plants. In the 1950s, they found the Mono variety (Gossypium barbadense) to be highly promising. When intercropped with maize and sprayed with pesticides, Mono yielded 200 kgs of seed cotton per hectare under the supervision of an extension agent. Allen cotton (G. hirsutum, v. latifolium) was first introduced into Côte d’Ivoire from Nigeria in 1927. It was quickly eliminated from experiment station trials because of its extreme sensitivity to parasites. Allen was reintroduced in 1959 following the development of new pesticides (DDT, Endrin, Lindane) that allowed it to flourish in monocropped fields. When planted in fields enriched by commercial fertilisers and cultivated following a strict calendar, the Allen ‘cotton package’ yielded 800–900 kgs of seed cotton per hectare, four times higher than Mono.

11The agro-ecological advance of IRCT’s Allen programme was just the first step. To be successful in farmer’s fields, a new agricultural extension programme was needed. The Compagnie française pour le développement des fibres textiles (CFDT), a French parastatal company created in 1949, worked with the newly independent Côte d’Ivoire government in the early 1960s to promote Allen cotton. CFDT declared that the Allen programme’s success depended on it controlling seed supply and providing fertilisers and pesticides to producers. The best way for it to achieve this goal, it argued, was for the company to gain exclusive control over cotton markets. The proposed vertically integrated system would allow CFDT to provide loans to farmers to adopt the technological package. The company could recover its credit when it paid farmers for their cotton minus the loan amount. Monopsonistic control over cotton markets would also allow CFDT to eliminate its fierce rival, the Jula trader. The Ivorian government approved CFDT’s request to restructure the cotton economy along the lines of a concentrated, local monopoly.

12The ‘CFDT system’ became the model of cotton market coordination in francophone West Africa between the early 1960s and late 1990s. In Côte d’Ivoire, producer prices were set by the government marketing board (Caistab) based on CFDT’s projected extension and marketing costs and income. The company ginned the entire crop and traded it on world markets. Caistab reimbursed CFDT (and later CIDT [5]) if the costs of production and marketing were higher than world market prices. The system persisted until the 1990s when two forces acted upon the Ivorian government to restructure the cotton economy.

13Fractures in the cotton monopsony. Farmers were the first force to change the CFDT/CIDT system. Their cooperative organisations repeatedly mobilised cotton market boycotts in the 1990s to pressure CIDT to increase the producer price for seed cotton and to reduce the costs of fertilisers and pesticides. Farmer militancy eventually earned cotton growers a place at the table where producer and agro-input prices were annually negotiated. This development was a milestone in farmer-state relations and energised cotton growers to take a decisive role in reshaping the cotton economy at the turn of the twenty-first century.

14The World Bank was the second major force restructuring the Ivorian cotton economy. In a series of structural adjustment programmes that spanned the 1980s and 1990s, the Bank insisted on the privatisation of public companies like CIDT as a condition for development assistance. To the objections of CFDT, the Bank oversaw the break-up of CIDT in 1998 into three companies—Ivoire Coton, La Compagnie Cotonnière de la Côte d’Ivoire (LCCI), and the New CIDT, a residual of the old CIDT but now a 100 percent public company. [6] The partial privatisation agreement gave the three companies exclusive rights to the cotton produced by farmers in ‘their zones’ during a two-year transition period (1998–2000) (Gergely 2010). The zoning scheme was scheduled to finish at the end of this period, but the government did not enforce this provision of the privatisation agreement. Cotton growers strongly opposed the continuation of zoning as well as the oligopsonistic structure of the new cotton economy. The cotton growers’ union, URECOS-CI, [7] created its own cotton ginning company, the Société industrielle cotonnière des Savanes (Sicosa), which began ginning operations in 2002. [8] In its brochure, URECOS-CI displayed a map showing the entire cotton growing region as the ‘Zone URECOS-CI’. It seemed that after forty years of monopsonistic control, the cotton economy was about to become more competitive. But the timing of URECOS-CI’s filière changing actions could not have been worse.

15In September 2002, during the first season of Sicosa’s operations, an armed rebellion led by the Forces Nouvelles de Côte d’Ivoire divided the northern and southern parts of the country, leading to a decade-long period of political and economic instability. The ‘crise’ especially affected the cotton economy due to its dependency on the timely delivery of agro-inputs. Insecurity on the roads impeded the distribution of fertiliser and pesticide deliveries. The New Forces also destroyed the cotton experiment station in Bouaké, which halted cotton seed varietal research and development. Cotton companies had to pay millions of US dollars in informal and formal taxes to government and New Forces agents between the port and the cotton growing areas (Bassett 2011). This heavy financial burden was passed onto cotton growers in the form of lower producer prices.

16Producer prices for seed cotton are set annually through a price-setting mechanism under the aegis of Intercoton, the sector’s interprofessional association. The complex price-setting formula considers the operating expenses of cotton companies, anticipated world market prices, the volume of cotton ginned the previous year versus the company’s ginning capacity, currency exchange rates, and cotton grower costs. Cotton ginning companies possess more knowledge of and experience in international cotton markets than producers. As a result, they are more successful in manipulating the mechanism in their favour. For example, when world market prices for cotton were at an all-time high in 2013, cotton companies tweaked the price-setting formula, which allowed them to reap the largest rewards. A comparative study of Côte d’Ivoire and Burkina Faso shows that cotton growers effectively subsidise an inefficient and over- extended cotton-ginning sector through the price-setting mechanism (Bassett 2014). The combination of high input costs and low producer prices make cotton growing a high-risk enterprise. It is not unusual for even the best farmers to get caught in the cost-price squeeze and fall into debt.

17Farmer debt and disillusionment. Field research in Katiali in 2008 revealed the particularly high risks and harsh reality of cotton during the crise. Farm surveys showed that three-quarters of the cotton growers were indebted to cotton companies after selling their crop and paying off their agro-input loans in 2006/07. The situation remained dire the following year when 65 percent of Katiali households remained locked in debt. The number of cotton growers working with Ivoire Coton and CIDT Nouvelle dropped by 57 percent and 69 percent respectively. [9] As farmers abandoned cotton, national production levels fell by 70 percent (Gergely 2010). Cotton ginning companies also felt the squeeze. LCCI filed for bankruptcy in 2006. Its three gins were sold in 2008 to Ivoire Coton and two new companies: the Compagnie Ivoirien du Coton (COIC) and the Société d’Exploitation Cotonnière Olam (SECO).

18The de facto end of exclusive zoning after 2002 did not fundamentally alter the structure of the cotton economy. Farmers still obtained their agro-inputs and extension services from cotton companies and Intercoton continued to set producer prices. The major difference was that multiple companies operated in the same zone and often in the same community. As a result, cotton companies had to compete against each other to attract growers. This competition took the form of enhanced services to farmers. For example, companies were quick to pay producers after they delivered their crop to the gin. They also provided loans to their growers to purchase food during the hungry season, to pay for school fees, and to buy oxen and ploughs. Farmers in Katiali spoke quite favourably of this change in the behaviour of cotton companies, and especially of their ability to decide with whom they wished to work (Field data, 31 December 2012). Cotton ginning companies complained about the increasingly competitive environment. They cited the high costs of providing extension services far from their gins and the difficulty in obtaining cotton to keep their gins running at full capacity (Interviews with COIC and SECO, January 2, 2012). To improve their economic situation, they adopted a number of strategies. Some companies like DOPA (Développement des opérations Agro-industrielles) encroached onto the territories of other companies to secure more cotton. They did so by paying cash for cotton, but at half the official price (Field data, 7 and 30 April 2008). Ivoire Coton, COIC, and SECO lobbied the Ministry of Agriculture for a return to exclusive zoning (Gergely 2010). A related strategy focused on weakening farmer organisations and thus their ability to contest company initiatives like exclusive zoning. Cotton companies undermined the cooperative structure by giving priority to decentralised informal groups over cooperative unions when it delivered inputs and bought cotton (Macrae and Marmignon 2004). This practice led to the multiplication of informal groups, which greatly diminished URECOS-CI’s power to negotiate on the behalf of cotton growers. [10] As the cotton sector unravelled, farmers increasingly invested their limited resources in other crops, especially cashews.

‘Cashews, the coffee of the North’

19The cashew tree (Anacardium occidentale) flourishes in the savannah region of Côte d’Ivoire, where a five- to six-month dry season is favourable to its growth. It is a fast-growing evergreen that begins to produce nuts three years after planting. The tree gained ground in Côte d’Ivoire in the late 1920s when the French colonial state promoted it in the context of savannah region reforestation programmes. Cashew plantations were commonly created in gazetted forests (forêts classées) and in village settings. Between 1929 and 1969, cashew trees accounted for 25 percent of the trees planted in state reforestation projects. More than half (56 percent) of these plantations occurred in village settings (FAO 1981).

20Following political independence in 1960, Côte d’Ivoire’s reforestation agency (Sodefor) continued to promote cashew plantations for soil conservation purposes (Sama and Koné 2002). Greater appreciation of the economic value of the cashew nut grew in the late 1960s in the context of rural development programmes funded by French development aid. After surveying the ‘sad spectacle’ of 8,000 hectares of unproductive cashew trees (Goujon, et al 1973), French foresters recommended a number of agronomic practices that would lead to soil conservation and higher nut production (Lefebvre, Leturco, and Praloran 1973).

21This revalorisation of the cashew tree received a boost in the early 1970s following President Félix Houphouët-Boigny’s tour of the savannah region (Labazée 2002). The president’s visit brought greater visibility to the region’s underdevelopment in comparison to the more prosperous forest region where the cocoa, coffee, and timber sectors were booming (den Tuinder 1978). To address this important development gap, the Houphouët-Boigny government financed a regional development programme to stimulate economic development in the north. The rethinking of cashew growing as a cash crop that would increase household revenues, just like the tree crops of the southern forest region, gave rise to the regional development slogan: ‘Cashews, the coffee of the North’ (Sama and Koné 2002).

22Typical of Côte d’Ivoire’s economic development model in the 1970s, the government created a state-owned company called Sovanord in 1972 to jump start the cashew economy. [11] The company was charged with developing domestic and foreign markets for raw cashew nuts and for creating a cashew processing industry. In 1975, Sovanord financed a cashew processing plant in Korhogo called the Anacarde Industrie de Côte d’Ivoire (AICI). Another state company specialising in the development of vegetables and fruit trees (Sodefel) was charged with selecting and diffusing high yielding cashew trees. With assistance from the government marketing board, Sovanord increased the producer price of raw cashew nuts from 25 to 40 CFA francs between 1975 and 1980. This price incentive encouraged the cultivation of cashew trees by individual households in contrast to the village-scale plantations favoured by Sodefor. The individualisation of production was similar to the transformation taking place in the cotton economy, where cotton fields and yields were rapidly expanding at the household level (Bassett 2001). But the cashew economy faltered and quickly fell. Economic inefficiencies associated with costly marketing and processing operations forced AICI to close its doors in 1980. Sovanord ceased its activities the following year (Sama and Koné 2002).

23One of Sovanord’s difficulties was the stiff competition it faced from local merchants who were also engaged in the raw cashew nut trade. Korhogo-based merchants had established trading relationships with Indian agro- export firms who supplied cashew- processing plants in India with their basic raw material. Rather than work with local merchants, Sovanord marginalised them in its trading and processing operations (Sama and Koné 2002). Local traders responded by competing with Sovanord for the limited tonnage of raw cashew nuts circulating in the north. While state investments in cashews declined, they substantially increased for cotton and rice (Sawadogo 1977). Cotton growing in particular expanded in the context of subsidised fertilisers and pesticides and attractive producer prices throughout the 1970s and 1980s, making it the number one cash crop of the savannah region.

24The resurgence of cashews. Farmers renewed their interest in cashew growing in the 1990s in the context of a changing agrarian economy. First, high world demand for cashew kernels led merchants to pay higher prices for raw cashew nuts. Indian traders worked closely with local merchants to buy cashew nuts for processing in South Asia. The price for a kilogram of raw cashew nuts increased from 40 CFA francs in 1983 to 75 CFA francs in 1993 (Sama and Koné 2002). Following the devaluation of the CFA franc in 1994, the producer price shot up to 200 CFA francs (Field data, 9 July 1995). These more attractive market prices motivated farmers, especially male household heads, to plant more cashew trees. The ability to diversify their agricultural system was made possible by two important conditions. The first was land availability. In contrast to the forest region, population densities in the savannah region are relatively low. Moreover, there is little land titling and private property in rural areas. Customary land tenure systems predominate, which provide the flexibility to convert fallow cropland to cashew orchards. However, immigrants, both national and international, and women find it difficult to access land to plant perennial tree crops. [12] Land conflicts abound in the north, where land authorities generally prohibit immigrants from planting cashew orchards. [13]

25A second condition that facilitated the expansion of cashews was its relatively easy integration into savannah farming systems. The agricultural calendar for cashews does not significantly overlap with subsistence crop and cotton cultivation. With the exception of weeding orchards during the rainy season, the labour demands of cashews, particularly harvesting, drying, and storage, take place after the food crop and cotton harvests. And, unlike cotton, cashew trees require few agro-inputs. [14] Merchants also pay farmers in cash, while cotton growers typically wait three months to be paid. These win- win-win conditions have made cashews an ideal crop for agricultural diversification, with the potential for improving rural incomes and livelihoods. This potential is ultimately a function of producer prices.

26The producer price for cashews was initially set each year by the Autorité de Régulation du Coton et de l’Anacarde (ARECA), Côte d’Ivoire’s cotton and cashew regulatory authority. [15] ARECA’s ‘indicative price’ was neither fixed nor guaranteed. Rather it was considered to be a floor price that buyers were supposed to respect. [16] But ARECA had neither the authority nor the resources to enforce its so-called ‘official price’. Consequently, producer prices were often well below that level. Field research in Katiali showed that farmers received as little as 35–50 CFAF/kg for raw cashew nuts in 2008 when the floor price was 150 CFAF (Field data, 12 April 2008). The year before, merchants paid 15–25 CFAF/kg despite ARECA’s indicative price of 150 CFAF (Field data, 18 March 2008). ARECA’s failure to ensure a minimum price allowed large exporting firms to strongly influence producer prices. [17]

27Despite these unfavourable conditions, farmers continued to expand the area in cashews. They did so in part because of the unravelling of the cotton economy during the 2002–2010 crisis. Savannah farmers hoped that cashew cultivation would help buffer them from the deteriorating conditions and daily deprivations of a divided Côte d’Ivoire. In addition to cashews, farmers also invested in food crops like peanuts, rice, and maize, whose prices were particularly buoyant at this time (Interview with Korhogo grain merchant, 10 January 2009). Depending on their economic standing, some farmers also invested in livestock raising, especially cattle raising.

28The upward trend in cashew production during the early 2000s (see Figure 1) also reflects the time lag between the time of planting and the time when trees come into production. Depending on tree density and orchard maintenance, a cashew tree reaches peak production between its tenth and fifteenth years (Koné 2010).

29In summary, the contrasting political economic and agronomic characteristics of cotton and cashews have led to contrasting market structures. Cotton is an agro-input intensive crop. The cotton package has historically been delivered through vertically integrated filières that have tied cotton growers through agro-input loans to a particular cotton company. URECOS-CI tried but failed to create a more competitive market structure. Unlike cotton farmers, cashew growers are not obligated to seek agro-input loans from a particular firm. Moreover, cashew growers are paid in cash and can decide to whom they sell their crop. In both cases, crop prices are negotiated prices. The big difference is that seed cotton prices are ceiling prices, while cashew prices are theoretically floor prices. The 2013 reforms of the cotton and cashew sectors build upon the legacies of these contrasting political economic and agricultural histories.

The 2013 reform of the cashew and cotton sectors

30The Alassane Ouattara government approved at its ministerial meeting of 22 March 2013 a strategy to reform the cotton and cashew sectors of Côte d’Ivoire. The reform documents contain a common set of goals for both sectors (RCI 2013; CCA 2015). These include improving crop productivity and quality; increasing revenues among value chain actors through more transparent marketing practices and a more equitable distribution of income; increased value added through local processing; and improved value chain governance. The major difference in the reforms is how the problems of market coordination and competition are framed.

31The return of cotton zoning. The cotton reform documents are classic examples of the development discourse that portrays the filière in a state of chaos that requires the intervention of development experts to bring order and stability. The documents describe the cotton economy as in a near state of collapse due to ‘multiple dysfunctions’. These dysfunctions include uncontrolled competition (une concurrence non-maîtrisée) characterised by the presence of more than one cotton company operating in the same area; poorly functioning agricultural extension programmes; and side selling. In short, the cotton problem is framed as a complete breakdown in market coordination. Under these conditions, farmers are no longer following best agricultural practices, cotton yields and production have seriously declined, and the costs of cotton company extension services have greatly increased (RCI 2013).

32What is absent in this framing of the cotton problem is any discussion of Côte d’Ivoire’s inefficient cotton ginning industry and the lack of price competition. The fact that the price-setting mechanism benefits cotton companies at the expense of growers is omitted from the problem definition (Bassett 2014). This is because the Ministry of Agriculture views the disorder from the perspective of cotton companies. It diagnoses the multiple dysfunctions as stemming from the ‘suppression of agricultural extension zones since 2000’ (RCI 2016). To treat these multiple ills, the Ministry recommended a return to exclusive zoning. Exclusive zoning, the Ministry argued, would ‘rationalise the interventions of different actors’, improve the quality of extension services, and boost production so that ginning companies could run at full capacity (BNETD 2015, 50). [18] In summary, the Ministry of Agriculture viewed zoning as the panacea to the multiple ills afflicting the cotton economy.

33For cotton growers, the zoning pill is a hard one to swallow. Representatives of the Fédération des producteurs de coton en Côte d’Ivoire (Federation of Côte d’Ivoire Cotton Producers) stated that the ‘great majority’ of cotton growers are against cotton zoning. [19] But because they are not well organised, ‘the voice of producers was not heard’ in the reform process. [20] Producer representatives view zoning as an anti-liberal policy that stifles market competition and depresses farmer incomes. In contrast to cotton companies, they view the decline in cotton as stemming from the high costs of agro- inputs and low producer prices.

34FPC-CI representatives called the division of the cotton growing areas among cotton companies as the ‘Yalta du coton’. [21] They ironically noted that during the reform process, cotton companies were competing against each other, but not in the sphere of market. Rather, they vied to control more territory to include in their prospective zones.

35Farmers are not opposed to zoning per se. Rather, they object to being marginalised by the cotton company-centric zoning plan promoted by the Ministry of Agriculture (Ouattara 2014). An alternative zoning scheme would give producer organisations the right to organise their own zones and to invite bids from cotton companies to operate in ‘their zone’. But the divide-and-rule strategy of cotton companies has weakened farmer organisations and muted their calls for such alternative zoning strategies (Interviews with Intercoton 15 July 2016 and FPC-CI, 2 August 2016). In the end, the FPC-CI representatives stated that producers have ‘more to lose than to gain from zoning’. In particular, they lose the freedom to choose with whom they wish to work.

36The zoning process took place in five stages. The Minister of Agriculture first drafted a reform plan that was approved at a cabinet meeting on 22 March 2013. The second phase involved the promulgation of laws and decrees concerning the commercialisation of cotton and cashews and the regulatory bodies governing these sectors (CCA 2015). The third phase took place in September 2013 in Yamoussoukro at a workshop where cotton and cashew actors assembled to discuss and validate a document on the implementation of the sector reforms (RCI 2013). This document proposed, among many other changes, to establish exclusive zoning of the cotton growing areas.

37One of the recommendations that came out of this workshop was to establish a committee to implement the zoning plan. The Comité zoning was created on March 2014. One of its first activities was to engage the Bureau National d’Études Techniques et de Développement [22] to delimit the cotton zones around existing cotton gins. BNETD determined the area of each zone based on the capacity of the gin and the amount of cotton historically produced in that area. It used the subprefecture as the basic territorial unit in its analysis and mapping. BNETD planners attached and detached subprefectures from zones based on their proximity to a cotton gin.

38Three criteria informed the process of allocating zones to the six cotton companies operating in the region [23]: gin ownership; the company that operated in that zone during the two-year transition following the 1998 privatisation agreement; and the percentage of the cotton collected by a company in a particular area during the 2013/14 season (BNETD 2015).

39The fourth stage involved heated discussions primarily among cotton companies about the boundaries and attribution of cotton zones. The scene was reminiscent of the scramble for Africa in which cotton companies, most of them foreign owned, haggled over the boundaries of their respective zones of influence. The discussions began on 30 November 2015 in Yamoussoukro, where the sector actors met to discuss the BNETD report. The focus of the exchanges centred on BNETD’s mapping of the cotton zones.

40Some cotton companies claimed that certain subprefectures should be attached to their zone based on the criteria used to delimit each zone (see above). The Ivoire Coton company argued, for example, that the subprefecture of Sarhala should be attached to its zone based on its proximity to its Dianra gin. In its report, BNETD had placed Sarhala in the CIDT zone based on its proximity to the company’s Mankono gin. Over the protests of CIDT, the Comité zoning attached Sarhala to Ivoire Coton’s zone (RCI 2016).

41CIDT representatives protested the loss of Sarhala. ‘Ils ont pris notre terrtioire’ a spokesman said with reference to the encroachment of other cotton companies into its zone. [24] CIDT was particularly vocal about BNETD’s recommendation that four subprefectures (Niakara, Tortiya, Arikokaha, and Katiola) be attached to the Compagnie Ivoirien de Coton (COIC) zone. CIDT argued that these areas had been attributed to its company during the 1998 privatisation agreement. COIC countered that its extension agents were more active than CIDT’s in these areas. The Comité zoning considered COIC’s extension work to be more important than historical attribution and concurred with BNETD’s recommendation.

42The final stage of the zoning process took place on 28 December 2016, when President Alassane Ouattara signed the decree (re)establishing exclusive zoning in the cotton growing areas of Côte d’Ivoire. The new order became effective for the 2017/18 cotton campaign. [25] Figure 3 shows the (re)territorialisation of the cotton growing areas of Côte d’Ivoire by five cotton companies. CIDT exists in name only. The former state company was bought by the owners of COIC in 2017.

The cotton company zones of Côte d’Ivoire (2017)

The cotton company zones of Côte d’Ivoire (2017)

This map shows the re-territorialisation of the cotton growing areas of Côte d’Ivoire by five cotton companies. CIDT exists in name only. The former state company was bought by the owners of COIC in 2017.

43Cashew market reform. In contrast to the cotton sector reforms, which limit price competition through fixed prices and exclusive zoning, the cashew sector reform takes a market-oriented approach that allows producers to sell their crop to any licensed buyer. The reform focuses primarily on improving the quality as well as the quantity of nuts, increasing transparency in marketing, and enforcing floor prices. The statutes give the Conseil Coton et Anacarde (CCA) (Cotton and Cashew Council) the regulatory authority and resources to implement these changes. To do so, CCA has created regional delegations to oversee greater transparency in commercial transactions, from licensing traders to certifying the price and origins of nuts.

44From the producer’s perspective, the reform’s most important feature is its guaranteed floor price. Merchants must now pay producers at least the floor price for their crop. The purchase price can go higher, as it did in 2015, 2016, and 2017, when demand and producer prices were at all-time highs. To enforce its floor price policy, CCA has established an elaborate receipt system that traces the revenues paid and received from the farm gate to the port. It has also created a stabilisation fund to guarantee producers the floor price when market conditions drive the price below that level. [26] A stable and fair floor price is welcomed by producers and local buyers who bear the highest risks of fluctuating prices.

45The contrast between the cashew and cotton markets could not be more striking. Although price-setting mechanisms are designed to establish minimal prices, the price for cotton does not vary throughout the year, nor by region. The cashew price, on the other hand, is a minimum price. Cashew prices vary throughout the year and by region. [27]

46Second, cashew producers can sell to whomever they wish, while cotton growers are obligated to sell their crop to the cotton company controlling their zone. The presence of many cashew buyers has strongly influenced producer prices. Market reports over the past few years link increasing prices with the heightened world demand and to the large number of buyers operating in Côte d’Ivoire (N’Kalo 2016).

47Third, the free but not laissez-faire market actually works for cashew growers whose net incomes are superior to those of cotton growers. The following case study of cashew and cotton growers in the subprefecture of Katiali (Département de M’Bengué) indicates some of the local-level effects of the contrasting market structures on farmer incomes and livelihoods.

Agricultural diversification and rural livelihoods in Katiali

48Katiali farmers have historically grown cotton as their main cash crop. Beginning in the 1990s, they began to diversify their production system by adding cashews to the mix. In 2015, we conducted a farming systems survey with the heads of forty households who were randomly selected and stratified by income and ethnicity. [28] The survey recorded the area in different crops, the use of agro-inputs (fertiliser, pesticides) for each crop, and net incomes from crop sales for 2014. We also asked farmers which cash crops they considered to be the most and least profitable and whether or not their livelihoods had improved, or not, as a result of agricultural diversification.

49Table 1 summarises the crop area based on the 2015 farm survey. Cotton and cashews dominate the farming system, accounting for 63 percent of the total cultivated area. Cotton was grown by 82 percent of the households in the sample, cashews by 97 percent. The average cotton area per household came to 5.6 ha with a range of two to twenty hectares. The average cashew area amounted to 6.5 ha, with a range of one to eighteen hectares. From a historical perspective (see Figure 1), the relationship appears to be one of cash crop diversification rather than an outright substitution of cotton by cashews.

50When asked to name the most profitable cash crop they cultivated, 85 percent of the respondents answered ‘cashews’. The reasons they gave for their response point to more than just income. Farmers mentioned the less onerous work of cashew cultivation, the lower costs of production, as well as the high price of cashews that year. When asked what they considered to be the least profitable cash crop, 80 percent of respondents stated ‘cotton’. Farmers emphasised the high costs of inputs, low market prices, and recurring debt to explain their dissatisfaction with cotton.

51Table 2 summarises the economic returns for a hectare of cotton and cashews. [29] It shows that net incomes from cashews were similar to those from cotton for the 2011–2013 period. Cotton incomes would have been lower if the state had not subsidised fertiliser prices by 25 percent in 2011/12 and by 20 percent in 2012/13 (ACE 2014). [30] Moreover, cashew grower incomes were obtained without the drudgery and high risks (e.g., debt) associated with cotton. The average net income for a hectare of cashews was significantly higher during the 2014–16 period. This is when the CCA enforced floor prices and the cashew market became more competitive. Cashew incomes were double that of cotton during this second period. Cashew prices reached historic highs in 2017, ranging from 700–850 CFAF/kg (N’Kalo 22 March 2017). If we take 700 CFAF/kg as the average producer price for cashews that year, net cashew incomes were 3.5 times higher than net cotton incomes per hectare.

52Why do farmers continue to grow cotton under these clearly unfavourable conditions? Until recently, the only way for most farmers to obtain commercial fertilisers was to grow cotton. It is well known that farmers divert fertiliser obtained for their cotton fields to their food crops fields. [31] This dependency on cotton for securing fertiliser to boost food crop yields is now changing. The 2015 farm survey revealed that more than half (55 percent) of Katiali farmers are now buying fertilisers with cash from their cashew earnings.

53Our last question elicited farmers’ perceptions on whether agricultural diversification had affected their livelihoods for the better, for the worse, or not at all. An overwhelming majority (88 percent) of household heads reported that their livelihoods had improved thanks to cashew growing. The responses range from general statements to specific instances in which the money earned from cashew cultivation contributed to improved household welfare. Nearly half of the households stated that cashew earnings solved a number of their problems, such as house construction, buying oxen and herbicides, employing off-farm labour, paying off debts, and greater food security.

54In summary, cashew cultivation has given farmers more resources to address a host of rural household problems that cotton alone cannot solve. One can only imagine how women would use their cashew earnings if more of them had access to land and labour to plant orchards. Given the close relationship between women’s status and household welfare, particularly children’s well-being, it is certain that the benefits of crop diversification would be even greater if resources were more equitably distributed within households.

Conclusion

55The 2013 reforms of the cotton and cashew sectors were ostensibly designed to increase the agricultural yields and incomes of rural producers through improved market coordination. Yet, the paths of these reforms lead in quite different directions. The main coordination challenge facing the cotton sector is the timely provision of agro-inputs on credit to a large number of growers. Since the colonial era, the solution to this market coordination problem has been to tie cotton growers to companies that provide them with agro-input loans on the condition that farmers sell their crop to the company that furnished the loan. The vertically integrated ‘CFDT system’ that emerged in the post- independence period was constructed to solve this coordination problem. It was also designed to eliminate competition in seed cotton markets in which Jula traders consistently offered higher prices than French merchants.

56The 2013 reform of the cotton economy draws upon these deep political economic roots. The reform began with the partial privatisation of CIDT in 1998 and ended in 2016 with the establishment of cotton zoning. The end result has been the transformation of the cotton economy from a monopsonistic to an oligopsonistic market structure that ties cotton growers to the company controlling their zone. Like the price setting mechanism, exclusive zoning means that small farmers are subsidising inefficient and over-extended cotton ginning companies.

57Cotton growers had proposed an alternative zoning plan that pivoted on producer organisations. In this producer-centred scheme, farmer unions would consider bids from cotton companies to operate in ‘their zone’. Farmer organisations like URECOS-CI were sufficiently strong in the 1990s to construct such an alternative cotton economy. But their organisation was weakened by the 2002–2010 political crisis and by cotton companies who adopted divide and rule tactics, such as working with informal as well as formal producer organisations. When the 2013 reforms were being discussed, farmers were too weak to advance their alternative zoning plans (Interviews at Intercoton, 25 May 2015 and FPC-CI, 2 August 2016).

58The cashew sector is far more competitive. Its dynamism can be partly explained by the less demanding agricultural requirements of the cashew tree. Cashews are less input and capital intensive than cotton. As a result, cash-poor farmers are not dependent on agribusiness and the state to coordinate the provision of inputs. And, in contrast to the cotton sector, the state has actively promoted competition by licensing large numbers of traders. Most importantly, the state has not structured the cashew economy in a way that forces farmers to subsidise an inefficient and inexperienced processing sector. [32]

59The 2013 reforms of the cotton and cashew sector take farmers down very different roads. The contrasting market structures, production systems, and prices present farmers with contrasting opportunities for improving their livelihoods. The cashew boom does not mean the end of cotton growing. Constrained access to land to plant cashew trees and relatively easy access to fertilisers to grow cotton mean that some farmers will continue to invest in cotton. The larger point of the cotton and cashew case studies suggests that a more prosperous rural Côte d’Ivoire is more likely to emerge under conditions in which the state acts to increase farmer power and options in agricultural markets that are too often organised against their interests.