How and why chiefs formalise land use in recent times: the politics of land dispossession through biofuels investments in Ghana

Introduction

The recent surge in large-scale land deals for agricultural investments has dominated political economy literature, media discussions and conference debates. Public attention to the land deals debate is driven by the analysis of potential outcomes and the intents of the actors involved. The actors often involved in the recent land deals are governments or government-backed companies, national and international companies, individuals etc. and their involvement is driven by commercial considerations (Food and Agriculture Organization [FAO] 2012). These actors are often labelled ‘land grabbers’ due to perceived exploitative acts in host regions and the potential negative impact on livelihoods and on land tenure (von Braun and Meinzen-Dick 2009; Bull 2010; Wisborg 2012). The term ‘land grabbing’ has thus been used in the land deals debate to denote ‘unfair practices’ often perpetrated by investors (FAO 2012).

Moreover, the land grabbing debate often portrays local resource users as homogenous groups who are equally denied land access as a result of large-scale land deals (Bull 2010; Noe 2013). This ‘land grabbing’ label at times comprises a state's complicity in land deals by creating enabling environments for land appropriation by corporate entities in seemingly legitimate ways, but to the detriment of local land users (Grajales 2013; Huggins 2013; see also Evers, Seagle, and Krijtenburg 2013). These assertions overlap other propositions that most large-scale land leases occur in state-owned land tenure regimes (Cotula 2012) and hence pinpoint ‘weak governance’ (Arezki, Deininger, and Selod 2011) or ‘coercive land governance’ (Huggins 2013) as decisive factors for land dispossession. The proposed recommendations therefore include the payment of due compensation to those persons affected, and the making of effective land contractual arrangements or suspension of large-scale land deals in order to prevent land use rights abuses in host countries (von Braun and Meinzen-Dick 2009; Cotula 2011; ActionAid International 2012). The debate clearly indicates ‘what ought to be’ the procedure and outcomes of land allocations but focuses one-sidedly on the role of investors and state institutions.

Notwithstanding the seemingly incontrovertible logic underlying such arguments, clarity is lacking in the debate, especially on predominantly customary land tenure regimes such as Ghana, where most recent land allocations have been sanctioned by chiefs who customarily hold land in trust for community members. This customary role of chiefs as trustees of land is upheld by Ghana's constitution (see Ubink 2008). The issue is complicated by the fact that formal records of the location and size of stool land¹ rarely exist, yet chiefs confidently claim to know their boundaries (see Berry 2001). Although studies have highlighted the active involvement of national political elites, civil servants, professional farmers etc. in land acquisitions in their home countries (Zoomers 2010; Hall 2011; Cotula 2012), the mediating role of local social institutions in land deals seldom features prominently. Fold and Gough (2008), for example, demonstrate the establishment of large-scale farms by multinational companies in Ghana and the consequent loss of smallholder farmers' access to customary land. The mediating role of customary institutions in land dispossession however does not feature prominently in their analysis. Inasmuch as customary institutions define resource access (Berry 1989, 2009; Lund 2011a), discussions of dispossession cannot be detached from the power of persons who interpret and administer custom (Berry 2001; Yaro 2012). Consequently, this paper does not limit itself to an evaluation of the role of investors or state institutions in land deals but extends the debate by assessing the power and the motivation of chiefs when they are involved in the making of the recent land deals.

The paper argues that, instead of a one-sided focus on investors' role in causing land dispossession, an understanding of how local social institutions and politics mediate investments in land will enrich analyses of processes of land dispossession. The entry point for this argument is set out around three central issues. The first section discusses the fluidity of property rights and citizenship concepts. The next section discusses the new wave of large-scale land deals in Ghana and the role of chiefs in the land deals. In the final section, I analyse the motivation of chiefs in land allocations for two biofuels investment projects in Ghana and their implications for land access by indigenes and migrants in the project areas. This is followed by a discussion and a conclusion.

Conceptualising land property rights and citizenship

A fundamental route to achieving property rights begins with claims and claim-making processes mediated by established rules in a polity (see Berry 1989; Sikor and Lund 2009). For instance, since land claims are partly defined by social identity, and social identity partly defined through rights to land (Sikor and Lund 2009), legitimacy of claims exists alongside the constitution of authority in a society (Lund 2011a). Environmental resource use and control therefore operate within the context of political authority. This constitutes the field of political ecology, which encapsulates how power and politics influence understanding of the environment and of environmental resources (Escobar 1999; Robbins 2004). Of particular relevance for this paper is the influence of post-structuralism on political ecology, especially social construction of ethnicity or identity (Neumann 2005). These constructions have necessitated an exploration into the symbolic meanings ascribed to environmental resources and how they correspondingly overlap conflicts regarding access to and control of material resources (Ibid.). Competing symbolic meanings ascribed to environmental resources denote attempts to legitimise (or undermine) social identity or social claims.

In order to investigate how competition for vital resources is organised and structured, it is therefore important to examine ‘how polities emerge, consolidate and recede through processes of legitimisation, inclusion, exclusion and violence’ (Sikor and Lund 2009, 2). Unlike traditional political theories that seek legitimacy from state institutions, state formation rather emerges from property rights and citizenship (see Sikor and Lund 2009; Lund 2011a). Lund (2011a, 71) expresses the inextricable relationship between property and citizenship by the definitions ‘what we have’ and ‘who we are’, respectively. A fundamental element of both terms is ‘recognition’ (Ibid.). Social identity thus determines whether or not one can legitimately access particular resources collectively owned by a group. Access to particular resources can usher people into citizenship status and vice versa. In this context, ‘citizenship’ does not merely denote nationality but also an identity of belonging to a group which can be achieved by fulfilling particular local conditions. Rules of succession or inheritance, for instance, defining potential heirs to property or office, may grant access to resources (Berry 1989) and consequently grant ‘citizenship’ rights. However, ‘strangers’ may be granted similar access by virtue of gaining permission or joining local resource-owning groups through marriage, fostering, capture or by living under the protection and authority of its leaders (Ibid., 42). The conditions of integration into ‘local citizenship’ hence mediate strangers' access to land (Chauveau and Colin 2010, 99). Such determinations are made by the institutions or actors recognised by subjects as legitimate, who in return recognise subjects as entitled actors (Lund 2011a).

By implication, entitlements to resources may be lost, gained or even bought and this is a clear indication of highly fluid situations when analysing property rights. ‘Citizens’ may lose their entitlements whereas ‘non-citizens’ may gain access to resources which are otherwise impossible. ‘Labelling’ and ‘self-labelling’ hence become important strategies employed in the legitimisation and delegitimation of resource claims. For Lund (2011a), the way people acquire and secure land rights may seem like a simple process. However, this process can often become complex when several competing normative orders may be brought to bear to legitimise specific claims, and several actors may at the same point in time compete for the authority to settle disputes and set norms by precedent or practice (Ibid., 72). In the context of institutional pluralism, the institution or institutional actor that is able to enforce and define collectively binding decisions on members of a society has ‘state quality’ (Lund 2011b, 887). With reference to Lund's analytical construct and the prominence of chieftaincy institutions in terms of authority over most land in Ghana, such state quality lies with chiefs. Chiefs' authority over land is almost unquestionable in Ghana, especially in instances when chiefs dabble in partisan politics or have strong party-political affiliations, even though the country's constitutions (1979 and 1992) prohibit this practice. Persons who acknowledge the authority of chiefs in return gain access to land resources controlled by chiefs. A break in this reciprocal relationship either undermines chiefs' authority or affects residents' access to land. In this study, the concept of ‘citizenship’ is used in the discussion of land dispossession during the implementation of two biofuel investment projects.

Jatropha biofuel investments and the new wave of land deals in Ghana

Ghana has a predominantly customary land tenure regime with about 80% of land in Ghana held by customary landowners, mainly families, clans and stools² (Kasanga and Kotey 2001; Ubink 2008). The remaining 20% of the land is held by private individuals or controlled by the state. Land ownership in Ghana is an embodiment of the rights of primordial groups; villages, stools, families and kinship groups (Aryeetey et al. 2007). The heads of such primordial groups, such as chiefs, village headmen and family heads, possess allodial rights over the groups' land, whereas members enjoy usufructuary rights, which confer on them a ‘certificate of group ownership’ (Ibid., 7). Families or individuals express their allodial title holdings by providing narratives of ancestry and historical events which are difficult to refute or question (Berry 2001, 2009; Lund 2008). The expression of allodial land titles by way of reference to customs based on myths about ancestors ‘descending from the sky’ to settle on the claimed land, for example, does not only solidify land claims (Berry 2001, 152) but also pre-empts any potential attempt to question the identity of the claimant (Berry 2009, 25). Narratives about land are thus systematically conveyed to descendants in order to protect lands for posterity and to ensure continuity of custom.

Indeed, primordial groups have always devised strategies to protect their allodial land titles. However, the changing customary land tenure system has created uncertainties as the title and responsibilities of land titleholders depend on the interpretation by chiefs and other traditional leaders who administer custom (Yaro 2012). The vague definitions of Ghana's constitution recognising chiefs as allodial landholders have created the leeway for chiefs to reinvent customs to arrogate to themselves the power to own and sell land, often to the detriment of weaker social groups (Ibid.). The role of chiefs in the reinvention of customs to the detriment of weaker social groups, especially migrant farmers, are evident in some post-independence governments' land reforms (see Boni 2005). For instance, Ghana's first socialist party, the Convention People's Party (CPP) (1950s–1966), introduced land reforms which were intended to break the authority of chiefs over land, bring economic relief to tenant farmers and eventually increase the state's dominion over land (Boni 2005; Aryeetey et al. 2007). Nonetheless, the overthrow of the CPP administration in 1966 led to a repeal of these reforms, which hitherto had weakened the chiefs' prerogatives over land or agricultural tributes (Boni 2005). Another attempt to usurp the authority of chiefs over land resurfaced subsequent to the Land Titling Registration Law of 1986, introduced by the PNDC³ government for reasons similar to those of the CPP government (Boni 2005; Aryeetey et al. 2007). In almost all the post-independence governments' land reforms, chiefs have always manoeuvred to reassert control over tenant or migrant farmers in subtle ways (Boni 2005). In spite of chiefs' resilience to maintain a continuous exercise of authority over land, large-scale land allocations for projects remained the sole responsibility of governments or state institutions (Larbi, Antwi, and Olomolaiye 2004; Ghana Lands Commission 2012).

During the last decade, many Ghanaian chiefs have allocated numerous large land areas to agricultural investors – mainly from Italy, Norway, Israel and Canada – for biofuels⁴ and other agricultural projects in Ghana. Most of these investments involve the cultivation of the jatropha plant for biofuel production as a result of the oil price increases in 2006–07 in Ghana. Noted for a high oil content of its seeds and its perceived economic viability in so-called marginal lands, jatropha was promoted by the government, chiefs and many energy policy institutes in Ghana (Ghana Energy Commission 2005; Technoserve 2007). Overwhelmed by the numerous jatropha investments in Ghana, a Ghanaian newspaper, Public Agenda, in November 2010 described Ghana as the ‘Jatropha Centre in Africa’. It is estimated that by 2009, 13 out of a total of 17 biofuel projects in Ghana focused on jatropha cultivation (Schoneveld, German, and Nutakor 2010). This is unprecedented in the history of agribusiness in Ghana as investors make direct land negotiations with chiefs, contrary to earlier decades when large-scale land deals for agricultural activities (usually rubber, cocoa and palm plantations) and mining concessions were solely negotiated and sanctioned by the state (Larbi, Antwi, and Olomolaiye 2004; Ghana Lands Commission 2012). For example, a total land area of 158,906 hectares was acquired by both the colonial and post-independence governments of Ghana combined (up to 2001) (Larbi, Antwi, and Olomolaiye 2004), compared with a land area of 1,075,000 hectares reportedly acquired by biofuels investors usually through direct negotiations with Ghanaian chiefs during the last decade alone (Schoneveld, German, and Nutakor 2010).

Although estimates of biofuels land deals are often contentious, it nonetheless suggests that the authority of chiefs in land allocations has increased in recent times relative to that of the state. The involvement of state institutions in land allocations now merely takes the form of confirmation and registration of the agreements reached between chiefs and investors through Ghana Lands Commission and local government officials. In response to these developments, the Lands Commission issued a report in February 2012 to contest chiefs' competence to sanction such large land allocations and consequently ordered that land areas above 400 hectares (1000 acres) must be approved by the National Lands Commission instead of the Regional Lands Commission as was the case before (Ghana Lands Commission 2012). This new regulation is not a novelty but rather seeks to ensure local participation in land negotiations, re-emphasise the importance of environmental impact assessment by appropriate state agencies and the involvement of many stakeholders for deliberations, especially in land deals that involve vast land areas. The need for this regulation confirms that the increased land allocation by chiefs is partly caused by state institutional weaknesses in land governance.

In a country where land, in addition to its economic significance, serves as a source of political authority and as a symbol of social identity, illuminating the new wave of land deals and their implications for land access in project areas is worthwhile.

The two cases and methodology

The paper compares the two biofuels projects Kimminic Estates Ltd (henceforth called Kimminic) (Case I) and ScanFarm Ghana Ltd (henceforth called ScanFarm) (Case II) in Southern Ghana. The paper compares chiefs' motivations for allocating land for the projects and the implications for land access in the two cases. The cases were selected for three reasons: firstly, both land deals were sanctioned by chiefs; secondly, both projects involve large-scale plantation models; thirdly, both project areas have a large number of migrants who mostly have temporary land use rights either on family and individual land areas or on stool land. The comparative analysis will show why similar strategies and motivations of chiefs in the respective land deals generated different incidences of land dispossession in the two cases.

The research involved a two-month preliminary fieldwork period (April–June 2012), followed by a six-month fieldwork period (August 2012–January 2013). The preliminary fieldwork involved key informant interviews, group interviews and a review of biofuels and land tenure literature. The preliminary fieldwork provided contextual information about social institutions mediating access to land by different social groups in the study areas. The major part of the fieldwork involved household surveys of predominantly farming households, though some household members occasionally engage in off-farm activities such as charcoal production and firewood collection. In each of the two cases, 40 households were purposively sampled making a total of 80 households. A total of 18 migrant households and 22 indigenous households were selected in each of the two cases. After a series of key informant interviews, I gained an impression of a higher proportion of indigenous residents compared with migrant residents and this formed the basis for the sampling of the households. Arguments in this paper are based on the analysis of symbolic meanings ascribed to the control over land and how citizenship identity mediates land claims during the project implementation.

Notions of land entitlement in the study areas

This section provides a background to the land dispossession debate by unveiling customs underpinning notions of entitlement to land by different social groups in the study areas. Interviews and informal discussions with residents of the project areas show the following diverse sources of entitlement to land. Individuals or families own land either by virtue of being first settlers, the first to cultivate cocoa or virgin forests in the project villages or as descendants of these settlers. This is an allodial land right, often called exclusive land rights, though it is customarily prohibited to sell land. This group of residents, called ‘indigenes’ (Kuromafo) by virtue of their allodial land rights, arrived in the villages between 1900 and the early 1950s and made claims to most available land using physical features such as rivers, river valleys and huge trees as land boundaries. Since migration into the villages continued, some indigenes strategically trace their ancestry to ‘time immemorial’ or label themselves as founders of the villages in order to distinguish themselves from later arrivals in terms of exclusive control over land. These categories of indigenous citizens are usually installed as Odikro. Land areas that are not claimed by the Kuromafo as well as vacated farmlands (called atuagya) constitute stool land which is vested in the village chief in trust for the subjects and the paramount chief (Omanhene). Residents who settled in the villages between the late 1950s and the 1970s thus could only access land through negotiation with the indigenes or chiefs. Although labelled migrants by chiefs and indigenous people, these migrants also have allodial land entitlements. These are migrants mainly from Northern Ghana who received land as gifts thanks to their honest service to the host indigenous families or chiefs. Descendants of these first-generation migrants also possess allodial land rights through inheritance. Based on the property rights concept, the first-generation migrants are indigenous citizens of the villages by virtue of their exclusive land entitlements. It is striking to note that stool land areas are interspersed with family land without formal demarcations. In times of disputes over land boundary, family members resort to narratives of family migration history though often subject to interpretation by the chiefs, who serve as administrators of custom.

Some residents have land entitlement based on marriage or friendship with indigenes or allodial landholders. Apart from those described above, the remaining sources of land entitlements are derived from market transactions. Examples include leasehold and share-cropping agreements. Allodial title holders (usually family heads) can rent out part of their land in exchange for money or something in kind. In share cropping, tenant farmers pay an agreed amount of money or crop yields to their landlords at the end of every farming season. Moreover, indigenes and migrants can use stool land by seeking permission from traditional councils through village chiefs. Whereas payment of initial token sums of money or Nsa Sika (money for drinks) and annual agricultural tributes (afehyiatuo) to chiefs are obligatory for migrants, indigenes are often exempt due to the ‘local citizenship’ identity of the latter. Stool land constitutes the most predominant source of land entitlement in the sampled households in both cases. Payment of annual tributes to chiefs in exchange for land is a symbolic acknowledgement of chiefs' authority over land, which in turn assures migrant farmers of legitimate and continuous access to land. Migrants (whether from Northern or Southern Ghana) who gained land use rights either from chiefs, or through share cropping and leaseholds, arrived in the study villages after the 1980s and thus have weaker land use rights compared with the descendants of first-generation migrants. As the subsequent sections will show, notions of entitlement are central in land claims and claim-making processes and cannot be ignored or downplayed in the dispossession debate.

Implications of chiefs' motivation to re-establish authority over land

Historically, chiefs' motivation to collect agricultural tributes from migrant farmers cultivating stool lands is often accompanied by increased land values, especially in Southern Ghana following the cocoa boom from the 1940s (Berry 2001; Boni 2005). In 1996, the Office of the Administrator of Stool Lands (OASL) was established under Ghana Lands Commission to ensure equitable distribution of revenues accrued from stool land for the benefit of all residents. Under this regulation, only migrants cultivating stool land are required to pay annual taxes called ground rents. Whilst ground rents are collected by OASL, the collection of revenues from charcoal producers is reserved exclusively for messengers of traditional councils. Disbursement of stool land revenues followed the constitutional formula: 55% for district assemblies,⁵ 25% for stools and 20% for traditional councils. Despite this regulation, chiefs demand agricultural tributes from migrants cultivating stool land as an acknowledgement of their continued authority over land.

Interviews with chiefs and the sampled households indicated that, during the past few years, migrants cultivating stool land rarely pay agricultural tributes to the stools. This unwillingness on the part of migrant farmers often arises from two dilemmas: firstly, most migrant famers perceive payment of ground rents as a replacement for annual agricultural tributes collected by chiefs; secondly, chiefs perceive evasion of tributes as threatening to their authority whereas migrants perceive simultaneous payment of ground rents and agricultural tributes as exploitation by chiefs. In response to this, the chiefs allocated large land areas to agricultural investors as a way of re-establishing authority over stool land. Two major issues are worth noting. First, the land deals in the two project areas involve the transfer of land use rights from previous land owners and land users to the investors, Kimminic and ScanFarm. Secondly, there are no formal demarcations between stool land and family or individual landholdings. The land allocations therefore affected the land areas of some residents of the project areas (see Table 1), and the boundaries of such land were difficult to trace. This led to the adoption of strategies by residents in order to regain land areas they lost to the Kimminic and ScanFarm projects. Reference to identity then became the strategy used by residents and chiefs either to legitimise or to undermine claims based on custom. For example, the presentation of complaints to chiefs and the investors, and of petitions to law courts and the Ghana Lands Commission that were employed by the allegedly affected residents to legitimise their land claims were based on reference to citizenship identity. Residents lacking citizenship identity can only protect their land through the goodwill of the investors. The next section shows how and why chiefs' motivation in the land allocations was a decisive factor for land dispossession as a result of the implementation of the two project areas.

Case I (Kimminic Project)

The Kimminic project involved a 40-year joint venture land deal with six traditional councils in the Brong-Ahafo region of Ghana for the cultivation of jatropha for biofuel production. The entire project involved a land area of 65,000 hectares. This case (Case I) focuses on the village of Bredi near one of the Kimminic project areas in the Nkoranza Traditional Council (henceforth referred to as NTC) (see Figure 1), covering land areas of 13,000 hectares. An annual profit-sharing allocation of 75% and 25% for Kimminic and the NTC respectively was agreed. Funding for the Kimminic project came from Canadian investors and Ghanaian residents in Canada. Since it was a joint venture, whereby the project village is a partner, the Ghanaian investors together with chiefs of the NTC advocated the protection of land areas cultivated by residents of the project village, especially the indigenous citizens.

Figure 1.

Case I – the Kimminic project (Cartography: Department of Geography and Resource Development, University of Ghana).

The NTC is one of the traditional authorities in Ghana endowed with vast land areas. However, successive generations of ‘indigenous citizens’ are thought to have lost track of the actual size of stool land areas, as non-permanent features (rivers and river valleys) are used to mark land boundaries (interviews, 2012). In the absence of formal boundaries, some migrants are staying on the outskirts of NTC stool land without the permission of chiefs. Moreover, migrant farmers from adjacent traditional councils, especially the Ejura Traditional Council (see Figure 1) in the Ashanti region, are accused of occupying stool land without paying tributes to NTC (interview with the chief of the NTC, 2012). This accusation does not mean that so-called migrants are ‘non-Ghanaians’ but rather that they do not meet ‘local citizenship’ criteria.

Referring to these observations, the NTC indicated that unclear land boundary demarcations and the gradual influx of migrant farmers into the areas pose potential threats to chiefs' authority over stool land. The NTC's responses to the migrants' alliance with some non-governmental organisations (NGOs) to contest the land allocation for the Kimminic project illustrate this scepticism. ‘We were shocked when migrants from Ejura Traditional Council cultivating our land sought to contest the land allocation for Kimminic’ (interview, 2012). Furthermore, the urge to prevent possible future land contestations was evident: ‘We opted for a joint venture instead of a lease in order to avoid land encroachment. Leases usually lead to occupation by encroachers [migrants] if intended projects are not successful. … But reversal of the land to the NTC is possible with the joint venture agreement’ (Ibid.). According to the NTC, previous experiences of land leases led to the occupation of their stool by migrants when the project was abandoned midway, but the NTC could not legally regain the land areas until the lease period expired (interviews, 2012).

In order to pre-empt the migrants' efforts to contest the land allocation, the NTC further sought to strategically challenge the migrant farmers: ‘All migrant farmers were asked to show receipts of payments of ground rent to show observance of custom guiding the use of stool land’ (interview with chief of NTC, 2012). The efficacy of this strategy implemented by the chiefs was evident during an interview with a migrant farmer: ‘Some of us [migrants] pay agricultural tributes every year. … But nowadays yields are poor. We were issued receipts for payment of ground rents many years ago but that is no longer done’ (interview, 2012). This response implies that migrants' inability to provide receipts of payments of both ground rents and agricultural tributes does not necessarily justify the noncompliance accusation made by chiefs. Nonetheless, the accusation prevented affected migrants from protesting over land dispossession. Subsequent interviews indicated that the migrants' identity further denied them the wherewithal to successfully contest the chiefs' land deal, which affected their farmland. According to an affected migrant farmer, ‘The village chief who is supposed to lead us [migrants] to the palace of the Omanhene [paramount chief] is an employee of Kimminic. He always postpones our proposed meetings but we cannot bypass him’ (interview, 2012).

Chiefs' motivation in the making of the land deal was further evident in their justification for the joint venture agreement. ‘We gave out the lands to the company so that the youth can secure jobs. Most occupants [migrants] of our land have not paid anything to us lately. … It is not a lease, the land areas are not sold out. The local people [indigenes] are also owners of the project’ (interview with a chief of the NTC, 2012). The land allocation for the Kimminic project was thought to be the ultimate measure to protect land for the indigenous residents and to create economic opportunities. With the approval of the land deal by Ghana Lands Commission, the NTC expressed the confidence to contest potential future land litigations as boundaries are demarcated and legally formalised.

Indeed, the NTC's preference for the joint venture agreement protected some land areas in the jatropha plantation. However, the NTC strategically placed the land use rights of some residents into the hands of Kimminic. Land access thereafter depended upon ‘local citizenship’ and the goodwill of Kimminic. It is striking to note that, since the NTC's preference for joint venture was in part intended to re-establish authority over stool land cultivated by perceived noncompliant migrants, the land areas of most indigenous people were not affected. Migrants who cultivated land areas owned by these indigenous landholders were equally not affected. Moreover, migrant charcoal producers and farmers who regularly paid tributes to the NTC or secretly paid bribes to village chiefs were offered new land areas during the project implementation. The fate of other migrant farmers depended on the goodwill of Kimminic. Indeed, Kimminic created reserved land areas for use by residents, including even migrants. The land areas of a few migrant farmers were however affected. The development of the jatropha plantation fields into rectangular shapes and the creation of reserved areas around the plantation (called fire belts) to prevent potential fire outbreaks affected such farmland.

The extent to which the joint venture agreement protected the land areas of most sampled households (as shown in Table 1) clearly shows that the motivation of chiefs in the land deal plays a decisive role in land dispossession.

Case II (ScanFarm Project)

ScanFarm Ghana Ltd (formerly called ScanFuel) is an affiliate of a Norwegian company, ScanFuel AS. The project initially involved a 50-year (13,000 hectares) lease agreement with the Agogo Traditional Council (henceforth referred to as ATC), established in 2008–09 for jatropha biofuel production. However, expectations of quick profit-making from jatropha by ScanFarm management were not forthcoming as the company claimed there was a limited market for the harvested jatropha nuts (interviews, 2012). ScanFuel therefore switched to maize production in 2010, prompting the change of name from ScanFuel to ScanFarm. In 2011, the lease period was reduced to 15 years. Case II focuses on the village of Nsonyameye near the ScanFarm project (see Figure 2).

Figure 2.

Case II – the ScanFarm project (Cartography: Department of Geography and Resource Development, University of Ghana).

I introduce the discussion of land dispossession in the ScanFarm project area by elucidating the historical backdrop of the current land politics in Agogo. Land ownership in the political history of Ghana was achieved through struggles over territorial hegemony by powerful chiefdoms or kingdoms. Gaining control over large land areas is achieved through displays of supremacy in warfare. Agogo town is adjacent to the towns of Kumawu and Kwaman (see Figure 2) in the Ashanti region of Ghana. The three towns are the traditional capitals of Agogo, Kumawu and Kwaman traditional councils respectively. The political history of the three towns is dominated by their collective efforts to annex the boundary of what elders of Kumawu and Agogo call the ‘oppressive’ and ‘powerful’ neighbouring ruler in the seventeenth century (interviews, 2012; see also Berry 2001, 172–176). After successful annexation, the first three chiefs (descendants of the same clan) of the three traditional councils located themselves at strategic places for defence purposes but made verbal agreements not to draw boundaries between their areas of jurisdiction owing to kinship ties (interviews, 2012). This oral tradition has existed for centuries for successive chiefs of the three traditional councils. In the absence of formal boundaries, an individual who settles or cultivates land in either of the traditional councils correspondingly acknowledges the authority of any of these chiefs. No formal boundaries existed beyond the settlements or farmlands in the three traditional councils though the chiefs refer to rivers and river valleys as superficial boundaries.

The increased demand for land during the last decade has triggered contentions over large-scale land allocations, which are undermining the custom uniting the three traditional councils. Chiefs of the three traditional councils have strategically extended their land rights to adjacent areas with extensive bush or grassland where there are no physical boundary features such as rivers and river valleys. The paramount chiefs of Agogo and Kumawu have allocated many land areas to investors over the last decade without any form of mutual consent. The Kumawu Traditional Council (KTC) rented out land areas for teak production near a village in the ATC in the early 2000s but no consent was sought from the other two (interviews, 2012). In 2008–09, the ATC also allocated a 13,000-hectare land area for ScanFarm's jatropha project without the consent of the other two. The ATC communicated this land deal to the Agogo residents as an arrangement which virtually covered marginal areas suitable for jatropha cultivation. ‘The lease covers only mfofoa [marginal land] … , farming will not be compromised’ (interview with the chief of the ATC, 2012). Meanwhile, since there are no formal demarcations between stool land and family land, the leased areas described as mfofoa included farmland and other productive land areas owned by families and individuals. The mfofoa, often perceived as marginal land, is usually fallow land which forms a significant part of agricultural production cycles and also serves as a source of firewood, grasses for livestock, fruits etc. for village residents (interviews, 2012).

The ScanFarm land deal was followed by another land allocation to Fulani herdsmen for cattle rearing, which reinforced the residents' scepticism about potential losses of their land areas to migrants (interviews, 2012; Daily Guide Newspaper 2011). It is important to note that the paramount chief who approved these lease agreements is a lawyer and a former minister of the ruling NDC government who has a thorough knowledge of property rights in Ghana.

Public scepticism about land dispossession further increased in 2010 when ScanFarm switched to maize production. This was because from 2010, ScanFarm relocated from predominantly grassland areas used for jatropha in Afirisere and Dukusen villages, towards forested areas near Nsonyameye and Baamaa villages (see Figure 2). The Municipal Chief Executive (MCE) of the ScanFarm project area who had family land in Nsonyameye village became equally sceptical about the land deal. As a result of public agitation and advocacy by NGOs (Bull 2010; ActionAid Ghana 2011), the MCE of the area initiated renegotiation of the lease agreement. In 2011, the lease period was reduced from 50 to 15 years and compensation payments for affected farmers increased from Ghana Cedis (GHS) 15 to 30 per acre per year. The revision of the lease was intended to reduce public discontent.

Nonetheless, the ATC land allocations posed threats to the KTC's authority over land. To prevent future loss of their portion of the jointly owned non-demarcated land, subsequent land allocations by the ATC have been contested by the KTC (interview with the registrar of the ATC, 2012). According to an elder of the KTC, the council has erected pillars on the bank of the River Afram (see Figure 2) north of the three traditional councils to prevent future land allocations by the ATC. The ATC's reactions to agitation by its residents and the KTC revealed their intentions in numerous land deals, as discussed in the sections to follow.

After a series of unsuccessful complaints to the ATC, community-based activist groups in Agogo made public demonstrations to contest numerous non-transparent land deals. The agitation reached a peak with a petition submitted to the King of Asante⁶ to oust the paramount chief of the ATC in order to protect their land and local citizenship identity. The petition states: ‘Recalling the Oath of Allegiance sworn before you, … , and the entire Asante Nation by [name withheld] to protect the lands our forebears fought for and left behind, and to protect and defend the citizens at all times, he has failed woefully and miserably to honour this Oath and thereby does not deserve to serve you and the people of Agogo.’⁷ In sharp contrast, the spokesperson of the paramount chief of ATC cautioned the feuding residents as follows: ‘The Kumawu Traditional Council has sanctioned numerous land deals including an area near Agogo without our consent. The Kuromanfo [indigenes] involved in the demonstrations are naive. … Land leases to investors will protect our land for posterity’ (interviews, 2012). Subsequent interviews with other elders of the ATC revealed similar accusations directed to the KTC regarding land allocations thought to have been sanctioned without the ATC's consent and hence the need to do the same to secure land for posterity. Based on the above illustrations, the ATC's motivation to sanction large-scale deals for agricultural investments, including the ScanFarm project, is a strategic way of formalising its boundaries.

The ATC further sought to display a highly altruistic disposition in the making of the land allocations by claiming that they could generate development opportunities from so-called marginal lands rather than continuing with the irregular agricultural tributes from alleged noncompliant migrant farmers. While elucidating on expected outcomes of the ScanFarm project, a chief of the ATC said: ‘We have vast land areas but we don't benefit from it. Most migrant farmers do not want to pay agricultural tributes. Only charcoal producers honour the payments. … There will be employment for our unemployed youth’ (interview, 2012). Indeed, during interviews, some migrants confirmed the ‘noncompliance’ accusation because of previous experiences of multiple payments of agricultural tributes to people who impersonated messengers of chiefs. Other migrants attributed the tribute evasion to unfair treatment in the collection of tributes. According to a migrant charcoal producer, ‘Chiefs threaten us so we have to obtain permits from them before we produce charcoal but indigenous charcoal producers evade the permits without punishment.’ Despite admissions of evasion of tributes by some migrants, subsequent interviews revealed the ATC's intension to undermine land use rights of migrants by tracing their genealogy. According to the spokesperson of the paramount of the ATC, ‘The migrants came here in the 1980s and asked for land from us [the ATC]. … Migrants do not own land here; we gave them land for free. They only pay Nsa Sika [money for drinks]’ (interview, 2012).

In addition to this, the ATC accused some village chiefs of dishonesty in the delivery of the agricultural tributes collected, while efforts to ensure effective collection of agricultural tributes have proven equally futile (interview with the registrar of the ATC, 2012). As elucidated earlier, although chiefs claim to evict only noncompliant migrants from stool land, generate development opportunities and formalise and secure land for the living and unborn citizens, the land areas of many residents, including indigenous residents, were affected. Meanwhile, the leased land areas had been swapped for royalties paid by ScanFarm. The next sections explain the outcome of land claims during the lease renegotiation phase.

A salient feature of the renegotiation phase was that allegedly affected residents were empowered to negotiate directly with ScanFarm without the involvement of the ATC. Indeed, this introduced transparency into the lease agreement. Disgruntled allodial landholders however deemed it an opportune time to strategically increase the size of their land areas. For the allodial landholders, the initial lease agreement reached solely between the ATC and ScanFarm was an affront to their land use rights and thus they made many competing and controversial land claims during the renegotiation phase. Owing to ScanFarm's earlier land preparations, which were undertaken without the consent of the residents, land owners within the leased area were identified and confirmed by neighbouring farmers and village chiefs during the renegotiation phase. There were however three shortfalls in the renegotiation process. Firstly, residents seldom record actual land sizes. Secondly, some land owners were not present in the village during the renegotiation phase. Thirdly, ScanFarm had paid royalties to the ATC for the leased areas. Residents resorted to the following strategies: firstly, they made land claims only after land areas had been prepared by ScanFarm; secondly, others magnified the size of land areas, having realised that features marking boundaries, such as teak tree stumps and hedge plants, had already been removed without their consent. ScanFarm's strategic responses to these controversial claims were critical in causing land dispossession, as elucidated below.

A resident's ability to invoke customs or provide narratives of family ancestry to claim local citizenship identity was important in the determination of the outcome of land claims. This ‘citizenship’ identity provided the allegedly affected residents an opportunity to successfully contest the ATC's land deals in law courts, or petition the office of the Asante King. This is evident in a proud statement by an indigenous resident over the unfettered land rights of indigenous people: ‘I am Kuromani [indigene] of this town, the Omanhene [paramount chief] can never take away my land. The … court ordered ScanFarm to pay compensation for encroaching on my land and ScanFarm had no other choice’ (interview with a successful land claimant, 2012). Successful land claimants either regained land areas, which were usually bigger than the affected land areas, or received compensation from ScanFarm. Migrant farmers who had accessed land from these successful claimants or landlords equally regained land.

Since earlier successful claimants often magnified the size of affected land areas, ScanFarm observed many competing claims and multiple compensation payments for farmlands which were very similar. In response, ScanFarm reduced the sizes of some claimed land areas. Furthermore, ScanFarm rejected several subsequent land claims, even some made by allodial landholders who were not present during the project implementation or the renegotiation phase. ScanFarm's equally strategic responses to perceived incessant claims were based on the assumption that royalties for the leased areas had been paid to the ATC and therefore claims perceived as contentious could not be addressed. According to a member of ScanFarm Management, ‘Every day people come here with new land claims but we cannot take the risk of responding to all claims. Royalties [undisclosed] for the leased land are paid to the chiefs' (interview, 2012).

Moreover, the development of ScanFarm's maize farms into rectangular shapes during land preparations affected portions of nearby farmland of residents, but no compensation whatsoever was provided. Tenant farmers whose landlords either accepted compensation payments or lost land claims were faced with land dispossession. Residents who cultivated part of the leased land without the permission of the ATC also faced land dispossession. These were mostly migrants who remained unrecognised by chiefs because of the noncompliance and migrant labels. The migrants lamented over how such labelling by chiefs affected their farmlands during the ScanFarm project. According to a 37-year-old migrant, ‘I was born in this village but because my parents are migrants from Northern Ghana, I have no land here. … ScanFarm has destroyed my crops and taken over my farmland’ (interview, 2012). Paradoxically, migrant charcoal producers who often paid tributes to the ATC were protected against land dispossession by village chiefs during the project.

The motivation of the ATC and the limited expression of goodwill on the part of ScanFarm obviously contributed to land dispossession. The following section discusses the main issues extracted from the two cases to examine which of the actors (the investors and chiefs) have a decisive role in causing land dispossession.

Why similar motivations of chiefs but different implications for land access?

In both cases, references to citizenship identity featured prominently in land claims. Both chiefs and residents invoked the citizenship concept as the canon to protect or undermine social identity in land claim-making processes. As elucidated above, land claims are not merely aimed at circumventing land dispossession but more importantly at securing citizenship identity. The petition to oust the paramount chief of the ATC over land allocations initiated by residents who labelled themselves ‘concerned citizens’, attests to efforts to pre-empt potential loss of local citizenship identity. Similarly, the NTC's motivations for undermining migrants' control over stool land and the preference for joint venture agreement illustrate quests to protect land rights in order to pre-empt any potential loss of local citizenship.

However, the concept of local citizenship remains highly fluid since customs that define land access and hence citizenship identity are interpreted and legitimated by chiefs. Although both the 1979 and 1992 constitutions of Ghana bar chiefs from dabbling in partisan politics, it has become a convention that, owing to the reverence for the chieftaincy institution, political parties strategically lure chiefs into their camps for vote-seeking purposes. This has made chiefs increasingly powerful especially as environmental resource brokers since Ghana was ushered into a democratic regime. This relative increase in the power of chiefs has had a telling effect on the definition of social identity and land access. Individuals – whether migrant or not – who can influence chiefs in turn gain access to resources controlled by chiefs and vice versa. Recognition (or lack of recognition) from chiefs thus has implications for land access.

Referring to the land grabbing debate, ScanFarm evicted many residents from the leased areas, denying them resources which once formed the backbone of their livelihoods. ScanFarm's contribution to land dispossession is clearly evident in their rejection of land claims even by allodial landholders in order to mitigate multiple compensation payments to other residents. This investment clearly deserves the label ‘land grabbing’. However, the study finds that the role played by the ATC (chiefs), who hold land in trust for residents, is predominant in this so-called ‘land grabbing’. The improved terms of the lease agreement during the renegotiation phase were achieved when the power of the chiefs was curtailed. This presupposes that incidences of land dispossession could have been reduced if proper negotiations and consultations were initiated by so-called trustees of land (chiefs) during the initial lease agreement with ScanFarm. Furthermore, this would have been easier given the legal training and the party-political affiliation of the paramount chief of ATC, who approved the ScanFarm land deal. The ATC's motivation for a lease agreement in order to re-establish authority over land eventually affected the land areas of many residents, including even some allodial landholders. Conversely, although similar motivations compelled the NTC to allocate land for the Kimminic project, the preference for a joint venture land deal significantly reduced incidences of land dispossession (refer to Table 1). Furthermore, the NTC's motivation protected the land areas cultivated by many migrants, who have weaker land use rights. The reverse was the case in the ScanFarm project area, where the land areas cultivated by many indigenous households was affected. It is noteworthy that the different incidences of land dispossession in both Cases I and II were largely determined by the different land contractual arrangements decided upon by the chiefs of the respective project areas.

Conclusion

The paper does not dispel evidence of land dispossession created by agricultural investors elsewhere in the world. Instead, it shows that analyses of motivations, strategies and of the power of actors defining land entitlements are equally important as the investors' strategies in the land deals debate. The empirical evidence suggest that, in instances where chiefs' motivations in the land deals corresponded with expressions of goodwill and quests to protect local citizenship identity, the land use rights of residents, whether migrants or indigenes, were protected. Indeed, both investors contributed to land dispossession in the respective project areas, and I argue that the process of land dispossession was largely initiated and reinforced by chiefs' motivation to re-establish authority over land. The different roles played by the chiefs of the NTC and ATC and the corresponding differences in incidence of land dispossession in the respective project areas illustrate the potential capability of chiefs to pre-empt land dispossession during large-scale land deals. This further shows that the causes of land dispossession are often connected to and depend on local sociopolitical dynamics, which leads to situations where similar projects generate different outcomes even in the same country.