Land distribution and acquisition practices in Ghana's cocoa frontier: The impact of a state-regulated marketing system

Land Use Policy 28 (2011) 378–387

Contents lists available at ScienceDirect

Land Use Policy journal homepage: www.elsevier.com/locate/landusepol

Land distribution and acquisition practices in Ghana’s cocoa frontier: The impact of a state-regulated marketing system Michael Helt Knudsen ∗ , Niels Fold Department of Geography and Geology, University of Copenhagen, Østervoldgade 10, 1350 Copenhagen K, Denmark

a r t i c l e

i n f o

Article history: Received 15 July 2009 Received in revised form 6 June 2010 Accepted 31 July 2010 Keywords: Africa Ghana Cocoa Small-scale farmers Land acquisition State regulation

a b s t r a c t Substantial differences in the size of landholdings among cocoa farmers in the Western Region – the last cocoa “frontier” in Ghana – are primarily a result of inheritance practices and the purchase of vast tracts of land by migrants in the initial period of the cocoa boom. Individual accumulation of land over the last decade has mainly taken place via inheritance (among indigenous farmers) without takeovers of land and dispossession of small-scale farmers outside the extended family. Land accumulation among migrant farmers is rare beyond the initial acquisition. Large-scale farmers transfer surplus from their higher volume of cocoa production into investments in non-farm activities and construction of new residential houses—and not in land acquisition based on market transactions. State regulation of the cocoa sector has spurred increased efficiency among private cocoa purchasing companies and thereby reduced the marginalization of farmers with small landholdings by preserving their access to a vital source of income. The unique character of the Ghanaian purchasing system is a major factor behind the relatively stable proportion in the access and control of land for cocoa between extended families. © 2010 Elsevier Ltd. All rights reserved.

Introduction In most studies of Ghana’s cocoa sector, no particular efforts are made to distinguish between different types of “smallholders” according to the size of their landholdings. In extreme cases, the cocoa sector is considered to be constituted by a large number of homogenous peasants with almost negligible landholdings: “Ghana in West Africa produces 16% of the world’s cocoa. . ., which is grown by 265,000 small-scale farmers owning between 1 and 2 hectares of land. . .” (Doherty and Tranchell, 2005: p. 167). Another observer of Ghana’s cocoa sector claims that individual landholdings range from 1 to 20 hectares, averaging about 5 hectares (Nyanteng, 1995). Other case study based figures, however, refer to holdings ranging from 6 to 138 hectares (Awanyo, 1998). Even though these analyses demonstrate striking differences in landholdings, none of them reflect on the causes or implications of such differences. The aim of this paper is to contribute to a deeper understanding of the structures and explanations behind the landholding patterns in the Ghanaian cocoa industry. We examine the extent to which landholdings differ in size among cocoa smallholders in Ghana’s

∗ Corresponding author. Present address: Sneppevej 2, 3. th., 2400 Copenhagen NV, Denmark. Tel.: +45 29 91 42 38; fax: +45 35 32 25 01. E-mail addresses: [email protected] (M.H. Knudsen), [email protected] (N. Fold). 0264-8377/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.landusepol.2010.07.004

Western Region (the last ‘cocoa frontier’) and how these differences have been generated: Are they a result of private purchase of land or some type of customary land transfer? Further, are these observable patterns of a stable nature or are they undergoing rapid change—and if so, what are the reasons? We document that the size of landholdings varies considerably among farmers and that land accumulation processes differ between indigenous and migrant groups. This unequal access to land means that revenues from cocoa sales differ substantially among farmers. Therefore the ability to allocate resources to agricultural and other investments also varies among farmers. However, because of a number of land rights issues (including “traditional” forms of inheritance and contracts between landowners, tenants and labourers) and relative land scarcity, agricultural investments to expand landholdings are relatively infrequent. Instead, surpluses are channelled into other activities, which results in increased income for large-scale farmers. Despite these variations in access to land, size of income and opportunities for accumulation, the structure of landholdings between families seems to be relatively stable over time. We argue that the regulatory system serves to maintain the existing structure of landholdings in Ghana’s cocoa frontier. In particular, the organizational set-up of the purchasing segment after liberalization is identified as an important element in this process. The organization of the cocoa sector in Ghana is quite unique in comparison with the other West African cocoa sectors, where the earlier marketing boards have been dismantled completely. In Ghana, the state is still

M.H. Knudsen, N. Fold / Land Use Policy 28 (2011) 378–387

in command of quality control and the marketing of exports, while the local purchasing system has been liberalized (Fold, 2001, 2004; Ruf, 2009). The argument leads us to reflect on whether the uniqueness of the Ghanaian system goes beyond organizational aspects of the cocoa sector per se and has wider significance in a broader perspective on rural development strategy. We start the paper by reviewing the theoretical contributions on cocoa smallholder land rights in Ghana and their incorporation in the wider national political economy. After introducing the study area and data, we then illuminate the different ways in which land in the cocoa frontier is acquired and expanded by largescale indigenous and migrant farmers. Subsequently, we discuss the transfer of surpluses from large-scale farmers’ cocoa production to investments in non-farm activities. In the penultimate section, we explain the state regulatory mechanisms in the purchasing segment, assess their impact on land use and land acquisition practices, and portray farmers’ attitudes towards the organizational set-up. We conclude by summarizing the lessons for rural policies that aim to reduce tendencies towards marginalization of smallholders and enhance regional development in the context of non-market-based accumulation of land for export crops. Smallholders and land rights in Ghana’s cocoa frontier Various arrangements of land rights, share-cropping agreements and obligations among cocoa landowners, tenants and workers have been of crucial importance for the westward movement of the cocoa frontier from its cradle in the Akwapim area (Hill, 1963) through successive expansions in Ashanti and Brong Ahafo to, most recently, the Western Region. The latter is now by far the most important region in terms of production volume and is considered to be the final cocoa frontier in Ghana. As in earlier frontier areas, expansion of cocoa land in the Western Region has relied on migrant farmers from well-established cocoa areas in combination with hired labour, typically from the North of Ghana.1 The hired labour may take different forms, as labourers are hired daily, annually or for specific tasks, but a common feature is that all decisions concerning cultivation practices and sales rest with the landowner. Migrant farmers and labourers have been crucial in the clearing of virgin forests, planting of seedlings and production of food crops (for instance, cassava, cocoyam and plantain) that serve as shade for the cocoa trees during the first years after planting. In some instances, particularly in the initial years of an area’s transformation, migrant farmers have been able to buy land from indigenous groups and thus become landowners in their own right. Other migrants have entered into land-share contracts with local landowners and have later acquired usufruct rights to those parts of the land they have cleared and cultivated. The content of these land arrangements is highly complex and the many variations across space and time are probably due to different local traditions and shifting balances in negotiating strength between indigenous landowners and migrants according to the local availability of land and labour (Boni, 2005). At the most basic level, contracts can be divided into two main types (Takane, 2002). In the abusa contract, the migrant (locally referred to as the ‘caretaker’) works on established cocoa land and is responsible for spraying, weeding and harvesting. The landowner usually lives nearby and monitors the migrant’s work. Landowners also decide

1 The northerners have traditionally undertaken cyclical migration to the cocoa areas as hired labourers in the peak period of the main crop harvest season (October–February) and return to work on their own land with soil preparation, planting, weeding and harvesting before going back to the cocoa areas the following season (Belas and Menezes, 1970).

379

when the cocoa should be sold, to which company and the form of payment (cash or cheque). The caretaker usually receives a third of the net revenue from the cocoa sales. The contractual relationship is not seasonal but can continue for many years until either the landowner or the caretaker wants to end it. In the abunu contract, the tenants establish the cocoa farm by clearing the land and taking full responsibility for all farm tasks (planting of seedlings, weeding, spraying and harvesting). Decisions concerning simple processing (fermentation and drying) and sale of cocoa beans also rest with the tenant. During the first years before the cocoa trees mature and become harvestable, the tenant receives no rewards apart from the right to produce and consume food crops on the cocoa land. When the cocoa trees can be harvested, the net revenue from sales is split between the landowner and the tenant according to the particular agreement (Boni, 2005). Usually the usufruct right to half of the cultivated land is transferred to the tenant when the farm is considered complete. However, land rights are not stable and cannot be taken for granted. Awanyo (1998) compares investment strategies among indigenous landowners and migrants who have acquired usufruct rights through abunu. While some of the indigenous cocoa farmers shift parts of their production away from cocoa in periods of low prices, all migrant tenants continue and even expand cocoa production in the same period. The expansion of the cocoa farm (ideally covering all the negotiated farmland), in addition to obtaining surveyed and signed plans of their land, is part of the migrants’ efforts to solidify land rights and protect themselves against land predators from the landowning group (see also Takane, 2000). These exclusionary strategies of migrants supplement their inclusionary strategies, which are practices that serve to please the indigenous landowning groups by giving timely and regular tributes and ceremonial gifts and demonstrating respect and humility in encounters in day-to-day life. In practice, the abunu contract is not necessarily clearly agreed upon at the outset of the landowner–tenant relationship, but according to Takane (2002), the ambiguity in the contractual outcome constitutes both uncertainty and flexibility for both parties. For instance, the landowner is able to check the ability of the tenant, while the tenant can postpone negotiations to a later phase if conditions are expected to change in his favour. Another “functional” benefit of the land-share contracts is that they can provide an incentive to increase production and a means to cushion production and market risks for both landowner and tenant. Finally, the interwoven system of different labour and land contracts is considered to fit ideally into a life cycle: the system provides landless tenants with an opportunity to become landholding farmers over an extended period of time, but without the need to provide capital at the start of the process—only labour input is needed, including the labour necessary to grow food crops. On the other hand, landowners can hire labour and expand cocoa farmland in different ways (various labour contracts, abusa, abunu) that suit their particular age, needs and accumulation strategies (see Takane, 2002). According to Crook (2001), the “legalization” by the state of the traditional system of labour and land contracts (with all their local modifications) has played a major role in the relatively conflict-free trajectory of frontier expansion in Ghana’s cocoa-growing areas. The system was established during British colonial indirect rule, under which the “customary” law was incorporated into a unified common law system through the institution of Native Courts. The courts applied native law and procedures in the areas of land, family, debt, religious customs and petty crime, in addition to colonial regulations and local taxations. The legal ideology of “communal landownership” was accepted by the British colonial government and served the interests of the chieftaincy and their lawyer allies. It became part of state law after independence and served to absorb

380

M.H. Knudsen, N. Fold / Land Use Policy 28 (2011) 378–387

the influx of foreign migrants within the context of land use and production relationships set by the local indigenous communities. These patterns of control over land have not resolved problems related to the uncertainty and insecurity of land tenure in Ghana—in fact, some observers claim that the system created a great deal of harmful litigation between stools, which in turn reduced the potential production (Leiter and Harding, 2004). In addition to these conflicts over land between stools (what he terms ‘chiefly disputes’) and the continued struggle over land between indigenous and migrant farmers, Boni (2008) also stresses the inter-generational conflicts between the old landowning heads of households and young men who wishes to establish themselves as cocoa farmers. In a situation with land scarcity, the youngsters increasingly question the legitimate right of chiefs and elders to allocate land to migrants but they have not been able to successfully pressure for institutional changes to secure their access to land for cocoa farming. Furthermore, conflicts occur within extended families over distribution of land after divorce or death of the household head (Boni, 2008). These conflicts and uncertainties have not been politicized and manipulated by the state, as in the case of Côte d’Ivoire, which now faces massive political problems concerning the size and economic importance of its foreign population in the cocoa-growing areas. In Côte d’Ivoire, a flexible and pluralistic approach to land issues allowed local regulatory orders to construct new labour and land use relations that resulted in a rapid cocoa boom. However, the massive flow of foreign migrants from the Sahelian countries and the allocation of land “to the tillers” took place at the expense of local practices for the transfer of usufruct and ownership rights, and the resulting agrarian structure was never fully accepted by indigenous groups. Economic dynamism and rapid frontier development were gained in Côte d’Ivoire, but latent political problems were created at the same time. In Ghana, the conflicts have remained localized rather than escalating into national politics. The state has not interfered in local struggles over land and power, thereby deflecting rural protest away from itself, contributing to political stability on the national level and the consolidation of state power.2 As convincingly phrased by Berry (2009) “Since 1992. . .Ghana has presented the anomalous picture of an ostensible stable democratically elected national regime presiding over local battlefields where tensions over land and chieftaincy erupt, periodically, into open combat.” (p. 39). Thus, local variations in conflicts over principles of agreement between indigenous farmers and migrants/workers have been embedded into broader processes of territorial and economic expansion of cocoa production in Ghana. Yet the difference between the way indigenous and migrant farmers obtain land and the different degrees of security of landholdings imply that the process of land accumulation differs between the two groups. It is an open question, however, whether surplus is used differently or the regulatory system has different impacts on the two groups. In the following analysis we aim to examine these questions by

2 Also comparing Côte d’Ivoire and Ghana, Woods (2004) relates changes in the conditions of the cocoa sector with changes in the political regime. Due to its economic importance, the cocoa sector in both countries has had a disproportionate role in the determination of national development strategies. Ruling elites in both countries have tried to gain control over the all-important cocoa surplus and allocate it according to their investment and consumption priorities. This has resulted in highly politicized conflicts between the elites holding state power and those in opposition. During booms, resources have been allocated without seriously threatening regime stability, whereas in bust periods, allocations have been contested among elites that are often structured according to ethnic or regional cleavages, reflecting intensified competition over a shrinking resource.

distinguishing between the two groups and therefore deliberately select villages in which each of the groups dominates. We also focus on what we define as large-scale farmers within the villages in order to track the path of land accumulation and different uses of surplus. This stratification will serve to qualify the perception of a somewhat homogeneous cocoa sector and illuminate different trajectories of landholding expansion and use of surplus. Data and study area The paper is based on a baseline survey of 360 households located in four settlements in Juaboso District, Western Region (see Fig. 1). The production of cocoa in the district took off in the late 1950s and 1960s, but started to boom in the late 1970s; today Juaboso is one of the largest cocoa-producing districts in the country. As the indigenous population (Sefwis) of the district was mostly small-scale subsistence farmers, the move to cocoa production by migrants resulted in drastic changes in local settlement patterns. In some cases migrants settled in the outskirts of existing (indigenous) settlements but often they established completely new settlements to accommodate the high number of migrants. Migrants settled principally together with other migrants in areas where they had access to land, thus resulting in a clear distinction between settlements dominated by either indigenous or migrant groups. The selected settlements represent the two main types of settlement in the cocoa frontier, namely those dominated by indigenous households (Bodi and Amoaya) and those dominated by migrant households (Bonsu Nkwanta and Kefass). Bodi and Amoaya are old settlements (established around 1890 and 1850, respectively), while Bonsu Nkwanta and Kefass are relatively new settlements, established about 40 years ago. In terms of estimated population size (2006),3 Bodi (about 10,000) and Bonsu Nkwanta (about 8000–9000) are considerably bigger than Amoaya (about 4000) and Kefass (about 2000–2500). According to our survey data, Bonsu Nkwanta and Kefass are overwhelmingly populated by migrants (90–95% of households), while Bodi is primarily made up of indigenous households (more than 95%). Amoaya is somewhat unique in this context in that about 25% of households are migrants. The households included in the survey were selected using random sampling stratified by location of residence in the settlement. This selection ensures that a broad spectrum, ranging from consolidated households in the historical core of the settlement to newly established households in the periphery of the settlement, is included in the survey. All the selected households were interviewed. If the head of household was not available for an interview on the planned day, a new appointment was made and the interview conducted another day. As Table 1 reveals, nearly all surveyed households in the two indigenous settlements have access to land for farming (283 households of the 360 households surveyed own land). Most of the households in these settlements belong to one of the local matrilineages, this being a prerequisite for securing farming rights and access to land. In contrast, there is significantly lower access to land in the two migrant settlements. Access to land in these settlements has been hampered by the difficulties encountered by late migrants who wish to purchase land from indigenous stool chiefs, as well as limited land resources in general. Landholdings vary between 2 and 127 acres. However, landholdings are mainly between 4 and 12 acres and there is a minor concentration of households with landholdings of about 20–30 acres. Hence, the households surveyed reflect the same differences in the size of

3 Estimate by the authors (based on interviews with key informants and observations made in the study settlements).

M.H. Knudsen, N. Fold / Land Use Policy 28 (2011) 378–387

381

Fig. 1. Map of study area.

landholdings as observed in other studies of cocoa farming areas (see above). In the next sections we are mainly concerned with farmers who own sizeable acreages of land. We have therefore identified

and selected farmers with large landholdings (denoted “large-scale farmers” in the following) by selecting the top 10% of landholdings in each settlement (included in the baseline survey) for an in-depth study of large-scale cocoa farmers. The top 10% of landholdings will

Table 1 Characteristics of indigenous and migrant settlements.

Households with access to land Average land size (acres) Households involved in cocoa farming Households involved in cocoa farming as Landowner Tenant Labourer Sample size Source: Survey data.

Amoaya

Bodi

Kefass

95% 14.89 93%

100% 21.33 99%

67% 14.34 69%

86% 12% 2% 60

98% 2% 0% 130

81% 13% 5% 55

Bonsu Nkwanta 51% 11.78 50% 79% 18% 3% 115

382

M.H. Knudsen, N. Fold / Land Use Policy 28 (2011) 378–387

Table 2 Selection criteria: large-scale farmers. Settlement

Selection criteria

Number of HH

Amoaya Bodi Kefass Bonsu Nkwanta

Farmers who own 30 acres or more of farmland Farmers who own 40 acres or more of farmland Farmers who own 30 acres or more of farmland Farmers who own 20 acres or more of farmland

5 13 6 8

Total

necessarily vary according to the landholdings in each settlement, meaning that the thresholds for large-scale landholdings vary from 20 to 40 acres (Table 2). Using a semi-structured interview guide, a follow-up exercise was conducted with this group of large-scale farmers, who were chosen in order to learn more about land acquisition processes and the use of surplus from cocoa production. In total, 32 farmers were interviewed, including 18 from the indigenous settlements (all of them indigenous farmers) and 14 from the migrant settlements (all of them migrant farmers).

Land acquisition in indigenous and migrant settlements In Juaboso District, the indigenous Sefwi population is defined as consisting of members of any of the resident local matrilineages (Awanyo, 1998). They can be described as the holders of “stools”. The stools are defined by Boni (2005) as: “symbolic artefacts charged with religious and political significance, managed by stool-holders, mostly chiefs, chosen from among the members of the matrilineage” (2005: p. 127). The land belonging to the stool, or in practice to the matrilineage as a group, is termed stool land.4 An individual can be granted farming rights to land, but cannot claim ownership, as the land belongs to the whole community (matrilineage), those living as well as the ancestors and those as yet unborn (Awanyo, 1998). The three main types of land acquisition among the households surveyed are (1) purchase, (2) inheritance and (3) share-land contracts (abunu). All three types are evident in the study settlements, but particular types of land acquisition dominate in certain periods in indigenous and migrant settlements. (1) Purchase of land takes place between a migrant and a stool chief or between two farmers. A price is negotiated and the land is handed over, sometimes after performing customary rituals (see Boni, 2005 for details). This transaction is not purely market-based and is often embedded in conflicts at a later stage. Typically, the migrant farmers see the land as their property, while this is contested by indigenous stool chiefs. Complicating the matter is the fact that a large number of migrant farmers have no documentation or certificate for their landholdings, making court decisions problematic (Alhassan and Manuh, 2005). (2) Inheritance of land covers a number of practices whereby land is handed over between family members. Inheritance is the typical way for indigenous farmers to acquire land. In Juaboso District, inheritance of land includes the passing down of land to family members (matrilineally, patrimonially or in combination) if the farmer dies, gifts to family members and the allocation of communal land to farmers by stool chiefs.

4 Juabeso District is divided into four stool lands (Amoaya, Bodi, Boizan and Benchimaa). The study settlements are covered by the Boizan Stool land (Bonsu Nkwanta and Kefass), Amoaya Stool land (Amoaya and Kefass) and the Bodi Stool land (Bodi).

32

(3) The share-land contract is locally referred to as abunu, literally meaning “that which is divided into two” (Boni, 2005). In practice, the share-land contract is carried out in various ways, depending on negotiation between the parties. Since the mid1990s, it has become common practice for tenants in Juaboso District to pay a substantial amount of money before entering into a share-land contract. In special cases, the share-land contract can be negotiated so that the tenant receives less than half of the land if he/she is unable to raise the required initial payment. In the indigenous settlements, land is traditionally inherited matrilineally, and the large-scale farmer is obliged to pass the land on in this way (Awanyo, 1998). This system is to some extent still practised in the Juaboso district. Cases of “pure” matrilineal inheritance primarily take place in Amoaya, where the system is practised alongside or in combination with a patrimonial system in which children inherit from their parents. In Bodi, no households recorded matrilineal inheritance as the sole system being practised. Instead, the common arrangement is a complex system whereby parts of the land, typically two thirds, are given to the children and wife, while one third is handed over to the matrilineal family.5 The reason for the partial continuation of the matrilineal system is that the large-scale farmers obtained their land matrilineally before the system changed and they want to pay some of it back to the family. At the same time, the matrilineal inheritance functions as a form of insurance for the wife and children who are left behind. If the wife and children later run into trouble, the matrilineal family is obliged to help them. It is rare for indigenous large-scale farmers to sell their land, and the sale must usually be accepted by the whole (matrilineal) family, as well as the chief. Some indigenous largescale farmers maintain their right to sell their land without the agreement of the family, although agreement of the chief is always required. None of the indigenous large-scale farmers have any documentation or certificate for their landholdings, but all claim that it is well known in the settlement which farmer has the right to cultivate which plots of land. The 18 indigenous large-scale cocoa farmers own plots ranging between 2 and 52 acres. The vast majority of land is inherited (about 94% of the plots) and the inheritance of land has taken place during the period from 1959 to 2003, particularly from 1985 onwards.6 Large-scale farmers’ landholdings in the indigenous settlements are made up of several different plots due to the particularities of the inheritance system (Fig. 2). Typically, an indigenous large-scale farmer inherits one smaller plot, which is cleared and

5 In 1985 the Ghanaian government established a new law of succession (the Interstate Succession Law, 1985), which stipulates that, when a person dies intestate, the spouse is entitled to 3/16 of the personal property of the deceased, while the children are entitled to 9/16. Hence, the share to the matrilineal side is less than that of the nuclear family (Takane, 2002). Despite the law, gifts and inheritance in Juaboso District exist in a variety of forms, both within and outside the law. 6 Six percent of the plots of land have been purchased, mainly from other farmers in town in need of cash to repay loans; one large-scale smallholder bought the land from his in-laws.

M.H. Knudsen, N. Fold / Land Use Policy 28 (2011) 378–387

Fig. 2. Indigenous settlements: sequence of plots acquired by large-scale farmers.

cultivated with cocoa over a period of around 3–6 years. Subsequently, another plot of family land is allocated, and the process goes on until there is no more land to cultivate or until the farmer thinks he has sufficient land. The acquisition of land among large-scale migrant farmers shows a different picture. In migrant settlements, most of the largescale farmers have bought the land from the stool chiefs and claim the right to sell this land to anybody without the latter’s agreement. The stool chiefs are only involved when the new owner of the land wishes to draw up a plan of his new land. The majority of large-scale migrant farmers obtain all or the main part of their land at one time and do not accumulate land as a result of a multi-stranded strategy in which several ways of acquiring land are combined.7 In only three cases has land been acquired in more than one way (Fig. 3). HH 237 and HH 205 are similar cases of an initial purchase of a plot of land from an indigenous chief, followed by engagement in an abunu contract. In the case of HH 237, the migrant initially bought a plot of land of 25 acres in 1982 and later entered into an abunu contract with a close acquaintance, a Kefass farmer, paying only a bottle of spirits for it. Large-scale migrant farmers typically pass down land to their children when they retire or pass away, or hand it over when the children are old enough to start farming it on their own. Usually it is the eldest son who inherits the land and then shares the harvest with his siblings. The largest landholdings were all purchased except for HH 260, which the eldest son inherited from his deceased father in 1992. This particular farm was established in 1979 when the father was able to obtain 53 acres of virgin forest land from the chief for a small amount of money. The other cases of inheritance among the large-scale farmers are also transfers of land from deceased fathers who were pioneer migrants in the area. The main reason why inheritance of land is so infrequent in the migrant communities is most likely the relatively short history of cocoa farming within them. The large-scale farmers are still the same migrants who arrived in the early cocoa boom and acquired large plots of virgin forest land around 20–30 years ago. As time passes, inheritance will probably become a more typical way of acquiring land in the migrant settlements.

383

is spent. As described above, accumulated income is not used to invest in the expansion of landholdings. Land augmentation via inheritance or abunu relationships seems to be the prevailing practice, as suitable land for cocoa farming is becoming increasingly scarce. In migrant settlements most farmers regard the price of land as exorbitant and new land is not allocated to them by indigenous stool chiefs. In indigenous settlements increasing population has put pressure on available land for farming, and land purchase from fellow villagers is certainly not part of local custom. A few small- and large-scale farmers have already purchased or are planning to purchase land outside the district, most often in the Ashanti Region. These plots of land are old cocoa farms now covered with secondary forest and ready for rehabilitation. The majority of farmers, however, seek new economic opportunities within the district, typically in non-farm activities. Diversification of income and livelihoods in rural Africa away from agriculture is well documented (see, for instance, Bryceson, 1996, 2002; Ellis, 2000; Rigg, 2006; Knudsen, 2007). Diversification of incomes on the household or individual level can help reduce vulnerability to shocks and stabilize incomes, which may otherwise vary widely on a seasonal basis. Hence, diversification can be used to generate alternative sources of income outside the agricultural season. Income received from non-farm activities is generally small but constant, and it can play a vital role for the survival of poor rural households—whatever the reason is for their poverty. However, diversification of income into non-farm activities is not just a survival strategy for the poor: engagement in non-farm activities cuts across wealth categories (Barret et al., 2001). Indeed, large-scale farmers in Juaboso District have diversified into non-farm activities,8 as opportunities for diversification within agriculture are very limited. Almost all the farmland (about 97–98%) of large-scale farmers is presently cultivated with cocoa or reserved for future cocoa production; the remaining land is used for subsistence production of food crops. Cultivation practices and routines on large farms (including the application of fertilizers and pesticides) are considered to be a scaling-up of practices implemented on smaller farms. Accordingly, most of the large-scale farmers who were interviewed do not consider themselves to be more “efficient” than other cocoa farmers. A few large-scale farmers even complained that they are unable to maintain their farms well enough to produce a satisfactory yield. This is typically the case in the indigenous settlements, where farmers inherit the land and start by clearing and cultivating the four corners of their plot to demarcate it. Subsequently, they clear and cultivate the remaining land as soon as possible in order to secure it from claims by other family members. Large areas of land therefore have to be maintained, and large expenses are incurred due to the application of fertilizer and pesticides and the hiring of labour. These farmers struggle with diseases (especially black pod) and weed (especially mistletoe), resulting in low yields and deterioration of the cocoa farm. They estimate that the same size of production could be obtained on half the acreage if the farm had been well maintained, but the need to protect the land from family members and other farmers is considered to be so important that all the farmland has to be “cultivated”.

The use of surplus from large-scale cocoa farming A crucial question concerning rural development in the Western Region is where and how the surplus from cocoa production

7

The 14 large-scale migrant farmers own plots ranging between 2 and 127 acres.

8 Non-farm activities are defined in accordance with the definition provided by Satterthwaite and Tacoli (2003), as: “. . .all activities outside the agricultural sector. This excludes wage or exchange labour on other farms (sometimes classified as ‘offfarm’) but includes services and manufacturing related to the transformation and processing of agricultural produce, as well as non-related services and manufacturing activities.” (2003: p. 20). It includes all forms of work taking place in a variety of locations.

384

M.H. Knudsen, N. Fold / Land Use Policy 28 (2011) 378–387

Fig. 3. Migrant settlements: types of land acquisition by large-scale farmers.

Hence, diversification into other sectors offers higher benefits for large-scale farmers who are able to engage in relatively capital-intensive investments such as trade in food products from urban areas, sale of agro-chemicals and building materials, and provision of transportation. The percentage of the 32 large-scale farmer households involved in non-farm activities more than doubled from about 40% to almost 90% during the period 1996–2006, indicating that nearly all large-scale farmer households are now involved in local economic activities outside agriculture. The types of non-farm activities (trade and transport) established by indigenous and migrant large-scale farmers are more or less similar. However, even though cocoa farming is still the most important element in the income of large-scale farmers, constituting on average about three-quarters of total income in 2006, the importance varies from 65% in Bonsu Nkwanta to about 80% in Bodi. This variation reflects the different opportunities in migrant and indigenous settlements for circulation and use of surplus outside agricultural production. The rapid increase in population has sparked a comprehensive process of urbanization, and several settlements have now passed the population threshold of 5000 needed to be classified as urban areas. Bonsu Nkwanta in particular has developed into a regional market for consumer products and services while at the same time attracting entrepreneurial migrants with the sole purpose of engaging in non-farm activities (Knudsen, 2010). The increasing population pressure and reorientation of investments into a variety of non-farm activities has physically changed the settlement: Small shops and businesses now dominate the main street, and the settlements are rapidly expanding at the expense of nearby cocoa farms, which are being cleared to make space for the construction of simple houses for new migrants. The high influx of migrants, highly mobile on both a seasonal and a weekly basis, has increased mobility within the district, as well as between the district and other rural and urban areas. As road infrastructure slowly improves in the district, investments in transportation become attractive. A number of large-scale smallholders now invest in taxis, minibuses and small vans (“trotros”). However, a notable difference in the use of surplus is visible in the bigger indigenous settlements such as Bodi, where surpluses from cocoa production are transferred into construction of residential housing on the hilly outskirts. These new houses are large concrete buildings with open space in-between them. Migrant farmers seldom invest in the construction of houses in their frontier settlements but prefer to channel the surplus back to their home towns in other regions. Here, surpluses are typically used to cater to family needs and invest in agricultural and non-farm economic activities, as well as to construct houses. This practice

among migrants is criticized by both the indigenous population and the district administration, who accuse the migrant farmers of merely extracting resources and utilizing public services without returning any investment to the district. State regulation of the purchasing segment in the cocoa sector We now address the question posed in the introduction, namely, how state regulation (at sector level) of prices and purchasing practices within commercial transactions between farmers and licensed buying companies (LBCs) affects land acquisition practices. We focus on these two issues as they are the main parameters in the partly liberalized Ghanaian cocoa purchasing system and have decisive consequences for both large-scale as well as small-scale farmers. These issues are also relevant for wider reflections on the applicability of similar regulatory instruments in policies that aim to support the inclusion of small-scale farmers in cash crop production for the world market. We would argue that the Ghanaian cocoa system keeps small-scale farmers in cocoa production while at the same time preserving the agrarian structure in the cocoa areas. First a brief outline of the two parameters in the present system. The liberalization of the cocoa sector in Ghana was initiated in the early 1990s as part of successive structural adjustment programmes (SAPs). The dismantling of the state marketing board for cocoa (Cocobod) was only partial, however, as two crucial divisions of the board have retained sole control within their sphere of activities. Quality control and export marketing are still controlled by Cocobod, whereas all purchasing of cocoa beans up-country and their transport to the ports have been liberalized. These activities (purchasing, storage at local depots and transport to ports) are handled by local and international LBCs, including the former monopsony, the Produce Buying Company (PBC) of Cocobod. PBC is obliged to act as “the buyer of last resort”, i.e. securing all cocoa farmers an outlet for their production. The competition between the LBCs means that profit is maximized via expansion in both volume and speed of the circulation of cocoa (and money) between farmers and the Cocoa Marketing Corporation (CMC). Hence, all companies are eager to buy as much cocoa in the shortest period of time as possible, subject to the quality being accepted by Cocobod’s Quality Control Division (QCD). Hence, the system does not discriminate against either indigenous or migrant communities—the LBCs buy wherever there is cocoa of a certain volume and the PBC serves the somehow marginalized farmers. The system is based on another important heritage from the pre-SAP period, namely the state-controlled determination

M.H. Knudsen, N. Fold / Land Use Policy 28 (2011) 378–387 Table 3 Small- and large-scale farmers in indigenous and migrant settlements.

Small scale <20 acres Large-scale ≥20 acres Sample size

Amoaya

Bodi

Kefass

Bonsu Nkwanta

77% 23% 57

53% 47% 130

76% 24% 37

82% 18% 59

Source: Survey data.

of purchasing prices. A pan-seasonal and pan-territorial price is fixed by Cocobod at the beginning of the major crop period (October–March) and usually lasts for the whole cocoa year, i.e. also covering the harvest during the minor crop period (May–July), although revised prices are occasionally announced in the early phase of the latter period. During the period of export expansion, the purchasing price in the local currency (cedi) increased, albeit not at a steady and predictable rate. The price is determined on the basis of the estimated revenue for the particular cocoa year, made possible due to the considerable volume of forward sales by the CMC controlled by Cocobod. These forward sales are possible because Cocobod has won a global reputation over the years as a very reliable supplier of high-quality cocoa beans. All cocoa purchased by the LBCs must be resold to CMC, which acts as the sole seller of Ghana’s cocoa on the world market. Even though the stated intention of the liberalization of the cocoa sector was to remove the export monopsony, none of the LBCs have obtained a license to export so far. The forward sales are used as collateral on the international financial market, and a substantial loan is provided each year by a consortium of international banks (Fold, 2004). The following paragraphs are based on survey data (i.e. the 283 land owning households). Farmers are divided into two groups according to a threshold value of 20 acres. This somewhat arbitrary value is justified by the fact that a farm of more than about 20 acres is considered a sizeable holding among most farmers (smalland large-scale) and key informants who were interviewed. Concurrently, all farmers with landholdings of less than 20 acres are termed small-scale farmers, although they are a diverse group of farmers with landholdings ranging from 2 to 19 acres. The same diversity applies to the group of large-scale farmers with holdings of 20 acres and above (see Table 3). However, the survey data do not indicate that specific findings can be gained from a more detailed categorisation according to landholding size. Furthermore, the aim is to paint a broad picture of the practices and perceptions of the two categories of farmers. The vast majority of the farmers in the four settlements are very positive towards the liberalized purchasing system and stress the importance of the denser network of cocoa-buying sheds. Juaboso is a high-producing cocoa district, and most of the LBCs have erected cocoa-buying sheds, not only in the larger settlements, but also in the small hamlets located in the farming areas. By establishing sheds in the farming area, each LBC hopes to secure a greater share of the district’s cocoa harvest, thereby significantly reducing transport times and costs for the farmers. They now have several potential buyers, which enable them to leave a LBC if they suspect that the purchasing clerk is cheating them or payment is being delayed. Some of the large-scale farmers adopt a more diverse selling strategy in which they sell to several different LBCs during the cocoa season, depending on which LBC offers payment in cash on the spot. Other farmers, especially small-scale farmers, concentrate their sales on one LBC in order to improve their chances of obtaining loans and receiving advance payment from that particular purchasing clerk. These loans are offered against the security of the expected harvest. Depending on the purchasing clerk and his relationship with the cocoa farmer, the interest rate varies widely

385

Table 4 Preferred form of payment for small- and large-scale cocoa farmers. Settlement

Form of payment Cash

Amoaya Small scale Large scale Bodi Small scale Large scale Kefass Small scale Large scale Bonsu Nkwanta Small scale Large scale

Sample size

Cheque

95% 58%

5% 42%

44 13

96% 82%

3% 18%

69 61

100% 62%

38%

28 9

89% 56%

11% 45%

48 11

from 0 to 100%. Loans from purchasing clerks usually have to be paid back by the end of November, when farmers are expected to have sufficient money after selling their initial harvest of cocoa beans. Apart for payment in cash, some LBCs also offer different kinds of benefits for the farmers, such as fertilizer, pesticides or machetes at subsidized prices, as well as a seasonal bonus on cocoa produce. Only the large-scale farmers with a substantial harvest and a significant stable secondary income are able to store the cocoa beans and dispose of them at certain intervals, e.g. once a month. This practice reduces the costs of transporting cocoa from the farm to the LBC—and some are even able to attract LBCs to locate sheds on their farms due to their large harvests. However, the vast majority of the farmers, particularly the small-scale farmers, are in constant need of income and sell their harvests for cash as soon as they have one or a few bags of cocoa. About 40% of the large-scale farmers (except in Bodi, where the share is much lower) receive payment by cheque (the so-called ‘Akuafo cheque’) which Cocobod introduced a few decades ago in order to stimulate the establishment of rural credit institutions and avoid fraudulent pricing practices by purchasing clerks. The system is not popular among farmers, as delays occur when they wish to cash their cheque in one of the few rural bank branches. Payment by Akuafo cheque is, however, often a prerequisite for obtaining a bank loan and a significant proportion of the large-scale farmers exploit this opportunity (see Table 4) although the indigenous large-scale farmers settled in Bodi have a higher preference for cash than the others. Whereas the liberalized purchasing system is very popular, most farmers are dissatisfied with the fixed pricing system—at least according to their initial responses. Many are tempted by the idea of a broker system in which they expect to have good opportunities to negotiate and obtain a higher price for a bag of cocoa. However, many farmers are also conscious of the fluctuation of prices in a completely liberalized system and – on second thought – maintain that the fixed-price system is more secure and stable for the cocoa farmers. Actually, a majority of the farmers stress the stable prices and the easy access to markets as the main reasons why they still consider cocoa farming an attractive activity. Most farmers perceive that the purchasing power of one bag of cocoa has drastically declined over the last decade, but maintain that they still earn the same income today as 10 years ago: the drop in the purchasing power of cocoa has been compensated by higher yields achieved through the application of fertilizer, state funded mass spraying programmes and new plantings of higher yielding cocoa trees. Summing up, there are a number of reasons why the system acts to cushion land alienation and land accumulation processes in the cocoa frontier. Firstly, the system ensures small-scale farmers’ access to markets because there are more buyers in rural areas

386

M.H. Knudsen, N. Fold / Land Use Policy 28 (2011) 378–387

than previously, which reduces the dependence and asymmetrical power relationship between one buyer and many sellers. It has also become much easier to hand over the cocoa beans, as the LBCs have located their purchasing sheds within a much denser network than the one that previously existed. Secondly, due to the competition among LBCs, small-scale farmers have improved access to cash and other benefits, including loans, provided by the LBCs. Finally, the system protects farmers against the fluctuation of world market prices although this effect seems to be much less appreciated by both large- and small-scale farmers irrespective of whether they are indigenous or migrants.

Conclusion Data from Juaboso District indicate that nearly all land in the indigenous settlements has been inherited, while the vast majority of land in the migrant settlements has been purchased. The communal land controlled by the indigenous stools guaranteed indigenous farmers land for farming, while new migrants had increasingly difficulties accessing land. The early migrants (up to the early 1990s) were able to purchase vast areas of land, while more recent migrants have most often entered into share-land contracts (abunu) to acquire land. The large-scale migrant farmers who subsequently expanded their holdings have also entered into share-land contracts, while large-scale indigenous farmers mostly expand landholdings through various forms of inheritance. This paper shows that landholdings among “smallholders” in the Ghanaian cocoa frontier vary widely due to differences in initial endowments (inheritance or purchase) and the existence of two distinctive expansionary pathways for land accumulation (inheritance or abunu). Accordingly, landholdings are not expanded via market-like transactions between individual sellers and buyers of land, and the scope of dispossession caused by impoverished smallholders with no alternative to selling their means of production seems to be very limited. This structural segmentation is further enhanced by the internal mechanisms of the state regulatory system maintained in the Ghanaian cocoa sector with the extensive network of LBC purchasing sheds and the pan-seasonal and panterritorial price that secures a stable income from cocoa farming for small-scale farmers. However, there are major socio-economic differences among households in the frontier areas, as large-scale farmers seem to earn a higher income simply due to their greater levels of production. Large-scale farmers have easier access to credit from banks and the surplus from cocoa production enables them to invest in high-return, non-farm activities that secure an additional income outside the cocoa season. Hence, the dynamics of social differentiation seem to be operating “outside” the realm of agricultural production as the proportion of landholdings between extended families is relatively stable. At the same time, regional development is being stimulated, as a substantial part of the agricultural surplus is “captured” locally, being channelled into activities that diversify the local economy, as is particularly visible in towns that have some sort of relative advantage, e.g. as transportation or commercial nodes. The question in relation to broader policy issues is whether these findings from the Ghanaian cocoa sector can be transferred to other agricultural sectors in Africa. There are strong povertyreduction arguments behind sustaining a landholding structure in which small-scale farmers are able to maintain both their land and a relatively stable income without seriously reducing production and land productivity. This study indicates that an export monopoly, combined with a system of fixed producer prices, has the potential ability to promote and support a privately coordinated purchasing

segment that ensures efficiency and improves access to markets for all producers, irrespective of socio-economic status. There are, however, a number of important preconditions. First, the fix-priced system rests on the premises that the quality control mechanisms have the ability to ensure a consistent quality of exportable surplus, and that the volume of production can be appropriately estimated. Secondly, the lion’s share of exports needs to be sold via forward sales or some form of contractual arrangements that reveal the approximate unit prices. Finally, the volume of production should be substantial (measured on a world scale) or the quality unique in order to attract private buying companies to set-up purchasing operations. Hence, the monopsony that organizes exports has to be internationally recognized as a consistent and efficient supplier of the particular agricultural commodity in terms of volume, quality and delivery time. Obviously, it will take a number of years with a continuously good track record to establish such a reputation, but there are no theoretical arguments against the possibility and potential of revived and lean marketing boards—even though there are numerous examples of inefficient and corrupt marketing boards in African economic history, including Ghana’s. We would argue, however, that there is nothing intrinsically wrong with this kind of organizational construct. Similar institutions have been incredibly efficient and beneficial for farmers in industrialized countries over the years—it is all about setting up proper procedures for management practices and the adequate monitoring of operations. However, the real barrier to a revival of a leaner and efficient system modelled on the Ghanaian cocoa board is likely to come from the WTO (and the Bretton Wood institutions), which demonstratively opposes ideas linked to the operations of state-owned enterprises.

References Alhassan, O., Manuh, T., 2005. Land Registration in Eastern and Western Regions, Ghana. Research Report 5, IIED. Institute of African Studies, University of Ghana, Legon, Accra. Awanyo, L., 1998. Culture, markets, and agricultural production: a comparative study of the investment patterns of migrant and citizen cocoa farmers in the western region of Ghana. Professional Geographer 50 (4), 516–530. Barret, C.B., Reardon, T., Webb, P., 2001. Nonfarm income diversification and household livelihood strategies in rural Africa: concepts, dynamics, and policy implications. Food Policy 26, 315–331. Belas, R.E., Menezes, C.F., 1970. Migrant labour and agricultural output in Ghana. Oxford Economic Papers, New Series 22 (1), 109–127. Berry, S., 2009. Property, authority and citizenship: land claims, politics and the dynamics of social division in West Africa. Development and Change 40 (1), 23–45. Boni, S., 2005. Clearing the Ghanaian Forest: Theories and Practices of Acquisition, Transfer and Utilization of Farming Titles in the Sefwi-Akan Area. Institute of African Studies, University of Ghana, Legon, Accra. Boni, S., 2008. Traditional ambiguities and authoritarian interpretations in Sefwi land disputes. In: Ubink, J.M., Amanor, K.S. (Eds.), Contesting Land and Custom in Ghana. State, Chief and the Citizen. Leiden University Press, Leiden, pp. 81–111. Bryceson, D.F., 1996. Deagrarianization and rural employment in sub-Saharan Africa: a sectoral perspective. World Development 24 (1), 97–111. Bryceson, D.F., 2002. The scramble in Africa: reorienting rural livelihoods. World Development 30 (5), 725–739. Crook, R.C., 2001. Cocoa booms, the legislation of land relations and politics in Cote D’Ivoire and Ghana. IDS Bulletin 32 (1), 35–45. Doherty, B., Tranchell, S., 2005. New thinking in international trade? A case study of the day chocolate company. Sustainable Development 13 (3), 166–176. Ellis, F., 2000. Rural Livelihoods and Diversity in Developing Countries. Oxford University Press, Oxford. Fold, N., 2001. Restructuring of the European chocolate industry and its impact on cocoa production in West Africa. Journal of Economic Geography 1 (3), 405–420. Fold, N., 2004. Spilling the beans on a tough nut: liberalization and local supply system changes in Ghana’s cocoa and shea chains. In: Hughes, A., Reimer, S. (Eds.), Geographies of Commodity Chains. Routledge, London, pp. 63–80. Hill, P., 1963 [1997]. The Migrant Cocoa-farmers of Southern Ghana: A Study in Rural Capitalism. LIT James Curry with the International African Institute, Hamburg. Knudsen, M.H., 2007. Making a living in the cocoa frontier, Western Ghana: diversifying Incomes in a cocoa economy. Danish Journal of Geography 107 (2), 29–43. Knudsen, M.H., 2010. Small town development in the Ghanaian cocoa frontier. In: Agergaard, J., Fold, N., Gough, K.V. (Eds.), Rural–Urban Dynamics: Livelihoods,

M.H. Knudsen, N. Fold / Land Use Policy 28 (2011) 378–387 Mobility and Markets in African and Asian Frontiers. Routledge, London, pp. 23–40. Leiter, J., Harding, S., 2004. Trinidad, Brazil, and Ghana: three melting moments in the history of cocoa. Journal of Rural Studies 20, 113–130. Nyanteng, V.K, 1995. Prospects for Ghana’s cocoa industry in the 21st century. In: Ruf, F., Siswoputranto, P.S. (Eds.), Cocoa Cycles: The Economies of Cocoa Supply. Woodhead Publishing Ltd., Cambridge, England, pp. 125–150. Rigg, J., 2006. Land, farming, livelihoods, and poverty: rethinking the links in the rural South. World Development 34 (1), 180–202. Ruf, F., 2009. Libéralisation, cycles politiques et cycles du cacao: le décalage historique Côte-d’Ivoire—Ghana. Cahiers Agricultures 18 (4), 343–349.

387

Satterthwaite, D., Tacoli, C., 2003. The urban part of rural development: the role of small and intermediate urban centres in rural and regional development and poverty reduction. Working Paper 9, Rural-Urban Interactions and Livelihood Strategies Series, IIED, London. Takane, T., 2000. Incentives embedded in institutions: the case of share contracts in Ghanaian cocoa production. The Developing Economies 38 (3), 374–397. Takane, T., 2002. The Cocoa Farmers of Southern Ghana: Incentives, Institutions, and Change in Rural West Africa, I.D.E. Occasional Papers Series No. 37. Japan External Trade Organization, Chiba, Japan. Woods, D., 2004. Predatory elites, rents and cocoa: a comparative analysis of Ghana and Ivory Coast. Commonwealth & Comparative Politics 42 (2), 224–241.