INTRODUCTION
In Africa, conflict has continually weakened and reduced countries' capacity to achieve their development agendas. Those occurring at micro-level over specific resources prevent many countries from undergoing rapid social and economic change (Huggins Reference Huggins2003). Conflict over grazing lands in pastoral and agropastoral areas of the continent is one such issue. As some studies have shown, change in conflict technologies (from use of spears to automatic weapons) causes significant instability (Otim Reference Otim and Babiker2002). The increasing trend of disputes within and between pastoral and agropastoral groups due to resource scarcity, the resulting tensions between them, and the inadequacy of existing institutional frameworks for conflict management, indicate the need to search for an innovative approach to overcome the problem (Cousins Reference Cousins1996; Vanderlinden Reference Vanderlinden and McCarthy1999).
It is argued that conflict over resource use should be considered as part and parcel of resource management, the main reason being that effective community-based management institutions can improve access by different users when there is variation in resource conditions across space at a given time period (Haro & Doyo Reference Haro and Doyo2004). Grazing resources are either underused or completely unused in areas where there is high insecurity. Conflict thus causes environmental damage due to poor distribution of animals over large areas and herd concentration in a limited space. Though diverse ecological systems produce appropriate environments for pastoralists to manage risk through spatial dispersion of their herd (Ahmed et al. Reference Ahmed, Azeze, Babiker and Tsegaye2002), conflict impairs pastoral mobility and access to different grazing zones (Coppock Reference Coppock1994).
What would customary institutions do and how can the state assist them to manage conflict? In this connection, Turner (Reference Turner1999) argues that customary institutions have been seriously eroded over the past century while state-initiated programmes have failed. There are also different points of view on the role of customary institutions in conflict management in Africa.Footnote 1 For instance, Thebaud and Batterbury (Reference Thebaud and Batterbury2001) emphasise that frameworks for negotiation and resolution of conflict must be developed locally and enforced through state support. A related work indicates the failure of customary institutions to prevent or suppress violence between ethnic groups with distinct social and cultural norms (Cousins Reference Cousins1996). To overcome this problem, the state may play a key role in establishing grazing reserves and introducing land reform in pastoral areas (Scoones 1995a). Other scholars contend that the state has to support and strengthen customary institutions (Lane & Moorehead Reference Lane, Moorehead and Scoones1995; Opschoor Reference Opschoor, Mohamed, Dietz and Ahmed2001). A good example of inter-ethnic conflict management through state intervention comes from northern Kenya, where the state encourages traditional authorities to negotiate on use rights (Haro & Doyo Reference Haro and Doyo2004). Such an endeavour succeeds only when it ensures sustainable distribution of rights and power to suppress the basic driving forces of conflict (Vanderlinden Reference Vanderlinden and McCarthy1999). Another view in this regard is that the state needs to apply ‘procedural law’ rather than ‘substantive law’ in managing conflicts, as the former provides a general framework within which various sources of claims obtain legitimacy (Vedeld Reference Vedeld1998). In applying procedural law, claims for access and use of resources are often contested, negotiated and settled at different levels involving different individuals and groups directly or indirectly (Babiker Reference Babiker, Mohamed, Dietz and Ahmed2001).
Moreover, integration of informal and formal procedures to transform conflict by applying procedural law should address three important aspects. First, it is essential to encourage conflicting parties to rely on a traditional way of conflict resolution as far as it helps develop trust and credibility for the commitments that they make (CHF International 2006: 25–7). There are two reasons for reliance on this approach: (1) traditional mechanisms take account of the specific local social and ecological contexts in which different communities live; and (2) the immediate impact of conflict remains local, not national, in terms of asset losses and inducing instability (Huggins Reference Huggins2003). Second, one has to further examine why or how such institutions succeed or fail to resolve conflicts among multiple resource-users under changing resource conditions (Cousins Reference Cousins1996). The third aspect is enabling customary procedures to make distinctions and linkages between root causes, triggers and sustainers of conflict, since one of the basic problems in understanding conflict is overlooking such linkages and cross-influences. Empirical work shows that while raids activate conflict (Hendrickson et al. Reference Hendrickson, Mearns and Armon1998), expansion of farming supported by land distribution policy can sustain it (Gebre Reference Gebre2001; Shide Reference Shide2005). In a broader sense of the term, those dealing with environmental security analysis consider that the overall increase in ecological scarcity is the root cause of conflict (Huggins Reference Huggins2003).
To understand conflict better, there is a need to distinguish between (1) conflict of interests on how and who should use resources, which is a source of dispute, and (2) violent conflict (Hussein et al. Reference Hussein, Sumberg and Seddon1999). Conflict of interests involves competitive demand over scarce resources. Violence is one of the products of conflict of interests in which the two form a continuum and hence are not necessarily dichotomous. In situations where there is litigation (institutional means), conflict of interests may take a non-violent form (ibid.). The meaning of conflict thus extends from simple disputes, due to unfair distribution of benefits, to violent conflicts involving guns, killing and loss of property. In pastoral and agropastoral regions, the key reason why conflicts of interest evolve into violence is the gradual erosion of elders' authority and the state's failure to provide security (Kratli & Swift Reference Kratli and Swift1999). Hence, violent conflicts may need different institutional arrangements and political intervention from disputes. For eastern Ethiopia, little or no information is available on factors that have led to violent conflict and the challenge to customary authority (Opschoor Reference Opschoor, Mohamed, Dietz and Ahmed2001).
This review indicates a variety of contexts in which resource use conflicts occur. Whether conflicts take the form of dispute or are violent, they can take place between internally divided groups or between large distinct groups. They may also be linked to other external factors, or be connected to competition over resources in the event of environmental change. Such complex aspects need to be considered in conflict analysis. The nature of the conflict between the ethnic groups in Mieso encompasses some of these features. The fact that this conflict presently receives national attention, and the need to understand the process that has transformed it to this acute stage, were the central motives in undertaking this study. Mieso district, located at the boundary of Oromia and Somali regional states, provides an ideal case to analyse property rights conflict for two central reasons: (1) the existence of different land use systems immediately raises the question of rights to land resources; and (2) this situation has been reinforced by the introduction of ethnic-based federalism, which gives critical importance to issues of territorial control.
The primary focus of this paper is to explain the causes of the conflict and explore the conditions that support (or inhibit) customary institutions in managing it. It examines the conflict from different dimensions, taking into consideration the existing local institutional capacity and the challenge to transform it. The findings address how benefits from livestock raids, change in land use, and increasing scarcity of grazing resources on communal grazing land have affected customary institutions in managing conflict. It also reflects on the important role of raids in transforming distributional conflict over grazing resources into violence. Livestock raids are the principal cause of the failure of informal agreements to prevent conflict between pastoralists and agropastoralists. Finally, results suggest the need for state intervention to focus on devising mechanisms to control raids prior to delineation of the rights to the contested grazing land.
The second section of this paper introduces competing arguments to explain causes of conflict. The third indicates sources of data and provides general background to the groups in conflict. The fourth briefly describes the historical relationships between resource-users, laying a basis for the fifth, which examines the overall factors related to the conflict. The sixth relates the insights gained from the findings to existing literature in order to explore the property rights structures needed to minimise the incidence of violent conflict. The last part concludes and provides suggestions to improve the situation.
THEORETICAL CONCEPTS
The central objectives of this paper are to explain the causes of conflict and to understand the role and limitations of customary institutions in conflict management. In the current theoretical literature, there are three streams of thoughts that may help to explain causes of resource conflict. The first comes from the ‘property rights school’ that considers poorly defined and enforced property rights as sources of conflict (Alston et al. Reference Alston, Libecap, Mueller, Drobak and Nye1997; Barzel Reference Barzel1989; Bromley Reference Bromley1991; Demsetz Reference Demsetz1967). Many scholars generalise that institutions for the commons are likely to have their origins in ‘conflicting claims to resources rather than concern for sustainable resource use’ (McCay Reference McCay, Ostrom, Dietz, Dolsak, Stern, Stonich and Weber2002: 370). The second derives from environmental security groups, and associates resource conflict with resource scarcity but refrains from commenting on property rights problems (Hauge & Ellingsen Reference Hauge and Ellingsen1998; Homer-Dixon et al. Reference Homer-Dixon, Boutwell and Rathjens1993; Homer-Dixon Reference Homer-Dixon1994).
A number of empirical studies drawing on this view do not provide much support, as case studies reveal that resource scarcity does not necessarily have a linear relationship with conflict, as long as good ‘governance and market development’ improve resource productivity (Homer-Dixon Reference Homer-Dixon2001). In particular, Hauge and Ellingsen (Reference Hauge and Ellingsen1998) comment on the methodological weaknesses in controlling for other factors generating conflict. They show that when economic and political factors are not controlled, these have more predictive power than resource scarcity. A great deal of evidence discloses resource conflict in resource abundant countries and communities, leading to the understanding that the causes of conflict remain intricate (Lind & Sturman Reference Lind and Sturman2002).
The third conceptual lens comes from distributive bargaining theory, which suggests that conflict management depends on the power relations between actors bargaining for change in institutions that affect the distribution of benefits from the commons (Knight Reference Knight1992). A prerequisite to apply bargaining theory in conflict analysis is that some social actors are more powerful than others; the theory then investigates the effect of such differences (ibid.: 127). An actor's success in bargaining is directly related to the ability to produce strategic commitments (or threats) to the rules of the game that in turn affect the strategy of other actors.
Differences in resources (e.g. weaponry and political capital) may strengthen or weaken an actor's position and motivation in reproducing consistent behaviour (ibid.). Strong actors tend to show credible commitments to their strategic choice for two reasons. First, they are able to manipulate benefits by altering their own strategies if desirable. Second, the relative bargaining power embedded in resources available to strong actors may make the weak risk averse. It pushes them to refrain from challenging the strong actor unless the existing social networks, by providing a basis for cooperation, can create a space to combine forces and increase their countervailing power or bargaining position.
Therefore, differences in resources can bring about power asymmetry. This influences the outcome of distributive bargaining (Knight Reference Knight1992: 127). When strong actors constrain others to choose a particular equilibrium strategy, the weak will comply whether they like it or not. This is based on the implicit assumption that the weak lack alternatives. Change to such an established institution by the weak comes at a greater cost. ‘Change in informal rules of a society can be generated either through change in distributional outcomes of those rules or shift in relative bargaining power of the actors’ (ibid.: 145). Improved bargaining power of the weak actor will lead to a gradual shift in strategy, producing a new equilibrium. This equilibrium may bring a fairer allocation of benefit from the commons over which distributional conflict exists.
In this case, examining the link between property rights conflict, customary institutions and the state enables one to understand the role of different actors in managing conflict. It can also widen our knowledge on how change in political regimes and ideologies can shift the context of the conflict, and whether this would escalate or reduce its intensity. In using distributive bargaining theory as an analytical concept, the aim is to investigate how far change in the political system has influenced or is able to create space for bargaining and negotiation to manage conflict on common property resources.
DISTRICT BACKGROUND AND DATA SOURCES
Mieso district has a total population of 53,665 (CSA 2006) and an area of 2,573 square kilometres. The amount of rainfall varies from 332 to 1,385 m with an average of 790 m per annum. Its altitude ranges from 1,107 to 3,106 m above sea level (ILRI/MoARD 2005). Mieso, the district capital, forms the central market for different groups, as it is located on the railway and main highway connecting the eastern region to Addis Ababa. It is 300 m from Addis Ababa and provides a strategic livestock market for traders. With a population density of 20·5 p/km2, Mieso district is inhabited by two ethnic groups: the Oromo (Ala, Nole and Ittu clans) and the Somali (Issa and Hawiya clans). The Oromos are mainly agropastoral, whereas the Somalis are pastoral, entirely relying on livestock.
Three important features explain the socio-economic conditions of this district. First, different production systems (pastoral and agropastoral) coexist in the same district. Second, these two systems compete for land and the way it needs to be used. Notably, one group cultivates land, though on a smaller scale than sedentary farming systems, whereas the other opposes this practice to secure free communal grazing land. Third, there has long been resulting inter-ethnic conflict. While livestock is given priority, a decline in rangeland resources manifested through the reduction in livestock feed availability has presented a challenge to sustaining the livelihoods of both groups. The overall perception, however, is that access to grazing resources is constrained more by the threat of conflict than by environmental change and demographic pressure. Among agropastoralists, there is a general trend to expand crop cultivation such as sorghum, maize and beans, as a means to diversify livelihoods to ensure food security (ILRI/MoARD 2005).
Results presented in this article are based on data collected from Mieso district between November 2004 and May 2005, and during August/September 2006. These data were derived from a household survey and two consecutive focus group discussions. The survey focused on various themes: economic activities involving relations with neighbouring clans, pattern of competition for grazing land, reasons for conflict, number of animals raided, involvement in violent conflict, and participation in negotiation meetings. The focus group discussion (involving key informants) was organised parallel to the survey. Some issues addressed in the interview questions were recapped in discussion with key informants.
The discussion emphasised historical and current relations between Ittu and Issa, causes of violence, how negotiations are organised, cooperation of other clans with Ittu, tensions among Ittu clan members, and reasons why Issa negotiate or attack. The composition of the key informants varied in the two focus group rounds. In the first round, mainly elders were involved, and in the second, formal leaders of peasant associations, district administrative officials and village leaders were contacted. This was done to understand how far customary and formal procedures are integrated. The analysis follows a simple narration in order to have a clear account of historical and political factors that primarily determine the dynamics of the conflict.
RELATIONS BETWEEN MULTIPLE RESOURCE-USERS IN MIESO DISTRICT
In explaining resource use relations in the semi-arid pastoral areas, the mobile system of livestock production induced by the variable supply of environmental resources raises the need to explain the interactions between multiple resource-users. This section describes relationships between resource-users, to show how these have changed and conflict has developed. A number of studies have shown that recurring drought in eastern Ethiopia has led to a declining grazing resource base (Baars & Mussa Reference Baars and Mussa1999). This scarcity has gradually intensified competition, uncertainty and conflict among multiple users (Beyene et al. Reference Beyene, Tafesse and Lemessa2001; Shide Reference Shide2005).
The relationships between pastoral clans of different ethnic groups in eastern Ethiopia are becoming increasingly complex and dynamic in response to change in resource settings and land use (Gebre Reference Gebre2001: 227–67). The change in land use has implications for change in property rights, as some herders cultivate the land while others claim it solely for grazing. This is one important factor explaining the conflict between Ittu and Issa. In addition, changes in their relationships are partly explained by relations with neighbouring clans. For instance, Afar and Arsi Oromo have been long-standing common enemies for Karrayu, Ittu and Issa clans, as they fought for the best pastures (Gebre Reference Gebre2001). Karrayu enjoyed peaceful relations with Issa, Hawiya and Ittu in sharing grazing land. As elders recall, during the period prior to the 1940s the Issa territory served as a bridge for proliferation of firearms into areas that Ittu and Karrayu inhabit.
However, in response to recent developments, the Karrayu clan has resisted access for Ittu and Issa. These developments include the expansion of irrigated agriculture, the establishment of extensive conservation areas for a game park, the occurrence of two major droughts in the mid 1970s and 1980s, and population explosion (Abule et al. Reference Abule, Snyman and Smit2005). The restriction has become more severe for Issa than Ittu due to close kinship and marital relations between Ittu and Karrayu. This condition has aggravated the hostilities among the three (Gebre Reference Gebre2001). Consequently, the encroachment and competition for grazing resources that have affected primary users and their relationships with neighbouring clans had an impact on the customary grazing rights established by different clans. The resulting instability to the common property rights of Ittu and Issa provides an attractive environment for conflict analysis. The above evidence advances our understanding, in that the history and evolution of property rights conflict between Ittu and Issa have not taken place in an isolated setting, but should be viewed as an element within a larger system of competition and conflict over resources.Footnote 2 Nevertheless, to provide greater detail and keep the analysis focused, the empirical materials presented in the next section emphasise the conflict between Ittu and Issa.
THE NATURE OF THE CONFLICT
This section presents evidence from the field-based interviews and focus group discussions. The most relevant work carried out in other pastoral regions is noted in order to strengthen the findings. This paper points out that where both state and customary systems fail to secure property rights, it appears legitimate for each claimant to invest in conflict to secure access through violence. The conflict analysed here represents the gradual transformation of communal grazing land into contested land. The findings emphasise five important factors associated with this: (1) historical political factors; (2) history of change in land use; (3) inter-clan cooperation and mobilising resources invested in conflict; (4) breakdown of customary institutions; and (5) the economic incentive associated with livestock raids.
Historical political relationships
Understanding the dynamics of the conflict requires a historical review of events that took place during the three political regimes: imperial, socialist and post-socialist. To put it chronologically, during the imperial regime in the early 1970s, the commander of the armed forces in eastern Ethiopia encouraged the Afar and Oromo to attack the Issa, and the livestock raided were redistributed to the group members who conducted raids. Subsequently, Issa youngsters left the country and migrated to Somalia and Djibouti, but those Issa clan members who continued to reside on their land suffered from the drought and famine without assistance from the Ethiopian government. A number of Issa wells were deliberately poisoned, resulting in the loss of livestock and human deaths under the imperial regime (Khalif & Doornbos Reference Khalif and Doornbos2002). Since then, the Issa and the Oromo (to which the Ittu clan belongs) have become enemies.
Under the socialist regime (1974–91), Issa who had fled the country and migrated to Somalia in 1972 were organised under the leadership of an educated Issa clan member (Hamud Farrah), who escaped from the attacks of the Afar and Oromo. Hamud Farrah also cooperated with the Siad Barre government of Somalia to fight against Ethiopia.Footnote 3 In 1977, the young Issa migrants participated in the war launched by the Somali government, and rejoined their clan members to fight against the Oromo. They re-established themselves in Dewele and Mulu (small towns of the Issa Somali), while cooperating with those Somalis who were fighting for the liberation of Ogaden.
However, the objectives of the Issa and Somali fighters were different. Issa elders disclose that the Issa were taking retaliatory action, whereas the other Somalis were targeting the liberation of Ogaden, in order to form ‘Greater Somalia’. The Oromo have a different understanding that the intention of the Issa coincides with that of the other Somalis, claiming that the Ethiopian territory up to Awash Valley should be a part of Somalia. After the end of Ethio-Somali war, the Marxist government had an ambivalent position towards the Issa, sometimes providing support and at other times undertaking massive killings.Footnote 4 Since these events took place, neighbouring states such as Somalia and Djibouti invested in the informal supply of automatic weapons to the Issa, to fight against other ethnic groups including Oromo and Afar.
The new political and administrative structure introduced in the early 1990s (after removal of the Marxist regime) along ethnic lines has partly shifted the context.Footnote 5 This new system has brought its own challenges and opportunities (although it hardly worked in Mieso district due to its mixed ethnic composition).Footnote 6 As a result, the national government gave exclusive administrative responsibility to neither of the two ethnic groups. Instead, the two regional states, Somali and Oromia, continued to jointly administer the district. However, a dispute emerged between the two regional administrations over the control of Bordede town, which is a customs point. Controlling this town has huge economic benefit attached to the revenue generated from tax collection.Footnote 7 Though the claim over the strategic town took the form of an administrative dispute, the confrontation between the Issa and Ittu over resource access remained violent. There is a direct link between the two levels of conflict. In some cases, this regional-level dispute has exacerbated violent attacks between pastoral and agropastoral groups over the struggle in resource use. Thus, a wider regional economic interest has made the conflict more complicated. As this continued, the Somali region claimed that twenty-one peasant associations should be administered under its purview including Bordede.
To find a political solution to this complicated problem, the Ethiopian government conducted a referendum in November 2004, while the fieldwork for this study was going on. The results were disappointing for the Somali region, which lost twenty out of the twenty-one peasant associations contested. Consequently, the Issa undertook indiscriminate retaliation against non-Somalis, to punish other ethnic groups that favoured Mieso's administration under Oromia. Many were displaced, lost their lives or assets, and had to be supported by relief agencies. According to the report of the Mieso District Neighbouring Regions Affairs office, a violent Issa attack on 20 July 2005 caused looting of assets estimated at 2,489,141 Ethiopian Birr (US $252,550). Fifty-five people were killed and many others wounded, including the district administrator.
Finally, the conflict has intensified under the post-socialist regime. At present, frequent incidences of conflict and indiscriminate killings are common whenever the federal army is not in the district (Shide Reference Shide2005). Although it might be expected that the Ittu would have a stronger power position than the Issa, since the political party representing them (the Oromo People's Democratic Organisation, OPDO) is a member of the ruling coalition, the reality has turned out to be different. OPDO does not have an influential position in the prevailing political system, despite being co-opted and remaining loyal to the ruling party to assist the domination of the TPLF in the new political order (Joireman Reference Joireman1997).Footnote 8 A wide range of literature discloses the complexity of such a seemingly decentralised system of governance and the consequences associated with it, particularly in the Somali region, where political order is highly centralised to exercise control, despite the constitutional provisions for regional autonomy (Hagmann Reference Hagmann2005; Samatar Reference Samatar2005), and where ‘people who are elected or appointed to positions of power tend to exercise that power to benefit the interests of the group that installed them in their post rather than representing the interests of Somali Region as a whole (Devereux Reference Devereux2006: 115).
Change in land use
Historic information from elders indicates that the settlement of Ittu clan members in Mieso dates back to the early 1900s. They migrated from the western Hararghe highlands because the grassland in the district formed the basis for livestock production. During this period, livestock were their main source of livelihood and they were pastoralists. In the early 1930s, the imperial government of Ethiopia allocated land (about 500 ha) for two private investors to cultivate maize, sorghum and beans, as well as to produce livestock, by fencing large pasture-lands. These two activities competed with the pastoral communal grazing land. Investors employed armed guards to protect access to the enclosed land. Ittu and Issa collaborated in destroying the investors’ farms and raided their animals in order to discourage them and block further encroachment of highlanders into their grazing area. The collaboration was strong and successful.
However, cooperation in sharing grazing resources has disappeared through time, due first to a decline in the resource base and demographic change, and second to the impact of land reform subsequent to political and ideological change in 1974. The reform brought a dramatic change in land rights, as all land came under state ownership. As a result, those landlords who were allocated land during the imperial regime were dispossessed and forced to leave the area. This change in formal institutions resulted in the allocation of the enclosed land to Ittu clan members who were previously expropriated and marginalised. Since then, private land use for cultivation has become common. Many Ittu clan members started to till the land, though livestock remained the basic source of their livelihood. This became a source of conflict, as the Issa increasingly resisted expansion of cultivation. Field inquiry shows that Issa have never been involved in crop-farming activity. One mechanism of resistance to land use change was to organise attacks during the planting and harvesting seasons. This has been designed to secure extensive communal grazing land.
Since the mid 1990s, following the formation of regionalism along ethnic lines, attacks have become even more frequent.Footnote 9 At present, frequent attacks are resulting in reduced benefits for Ittu from communal grazing land. Many clan members fear that this may cause displacement from their settlement area in the long term. As the Issa resistance increased over time, by restricting access to communal grazing area to discourage crop farming, Ittu elders started to negotiate for access to the land they once enjoyed as common property with the Issa. Among those agropastoralists who combine cultivation and herding, the claim to presently contested grazing area is based on the claim that it was common property. To sum up, changes in land use by some users of common property, as an income diversification strategy, have become a threat to others who wish to sustain pastoralism as a livelihood, leading to distributional conflict that is transformed into violence.
The basic problem here is not the evolution of co-users of the grazing commons into ‘disputants’ by virtue of pursuing different production systems. Rather, it is the lack of a legal institutional framework through which claims for access can be negotiated, settled and sustained. In Ethiopia, the existing land tenure policy supports private use of land for cultivation as stated in the 2005 land use and administration proclamation (FDRE 2005).Footnote 10 At a local level, the allocation of land for cultivation favoured by this policy environment contributes to the shrinkage of the grazing area, putting the pastoral Issa at the losing end.
Furthermore, while state policies support crop production via the supply of agricultural technologies, extension services and training for agropastoralists, there is little or no legal backing to alleviate insecure property rights to the communal grazing land. As demographic change in agropastoral systems results in large tract of land coming under cultivation, tension arises when pastoralists practice mobility. This tension will intensify as agropastoralists increasingly hold land in private, while also claiming access to communal land for livestock production. This can be conceived as a struggle between diversification and specialisation. Interviews with Issa elders suggest that large areas of grazing land have been lost due to cultivation practiced by agropastoralists and farmers who have been allocated land by state functionaries.
Group identity, cooperation and resource mobilisation
Property rights literature indicates that the state is the ultimate supplier of security to fruits of one's investment. Where no one has invested but claims to be the right holder, as in the case of natural rangelands, who should give protection to claimants? Various forms of cooperation prevail in the case study district. In this section, emphasis is given to the influence of two major factors: (1) the link between group identity and cooperation as access strategy, and (2) the effect of mobilising resources in conflict prevention and the challenge faced in doing so.
Inter-clan cooperation
Historically, when conflict was started, the Issa divided the three clans of the Oromo in the district (Ittu, Nole and Ala) in resisting Ittu access to communal grazing land. The first step was to form an alliance with Ala and Nole clans to use the communal grazing land in harmony. These two clans, unlike the Ittu, had long been involved in crop growing. The second step was to resist the Ala while sharing the resource with the Nole alone. Unexpectedly, this strategy created an alliance between the three Oromo clans to resist the Issa. In this case, emphasis is given to conflict between the Ittu and Issa, which share a long geographical boundary, making their exposure to conflict higher. The other two clans are usually conditional co-operators.
The nature and extent of cooperation between clans affects the decisions of the parties in conflict. When both clans cooperate, the Ittu can use the contested resource without negotiation because of lower marginal costs of conflict per additional person cooperating. This condition undermines the intimidating character of the Issa. Thus, a credible threat from the Issa encourages inter-clan cooperation, which in turn improves access for the Ittu due to increased capacity for self-protection. Inter-clan cooperation increases security of rights to the contested grazing land, while other studies have indicated that such cooperation cannot be sustainable if there is variation in settlement patterns and geographical locations (Gebre Reference Gebre2001; Hagmann Reference Hagmann2005).
Under repeated interaction, inter-clan cooperation causes a different behavioural outcome (preventing conflict) only if the stronger actor is willing to negotiate over rights. At the same time, sustainability of inter-clan cooperation influences the perpetuation of such behaviour. As others have observed (Bogale & Korf Reference Bogale and Korf2005; Vanderlinden Reference Vanderlinden and McCarthy1999), cooperation is determined by the nature of the resources involved. The Ittu secure cooperation from the Ala and Nole whenever these face shortage of grazing resource in their area. When Issa resistance is reduced under persistent inter-clan cooperation, the Ittu can use the grazing area without being accompanied by the other cooperating clans. The Ittu, as a constrained group with limited chance of extended mobility, follow this path when the Issa migrate further from the contested land or when they believe the Issa to be unaware of it. The Issa observe such a move but keep quiet to plan for massive raids and attacks, which will force the Ittu to retreat and begin negotiation.
A recent study in eastern Ethiopia shows that poor agropastoralists cooperate with outsider pastoralists in permitting access to their grazing area in exchange for livestock and other benefits. They rent out access rights to common grazing land that improves their wellbeing. Where cooperation involves mutual gains, differences in production system do not necessarily lead to violent conflict (Bogale & Korf Reference Bogale and Korf2005). A more interesting argument comes from the work of Babiker (Reference Babiker, Mohamed, Dietz and Ahmed2001), which underlines the need to emphasise the interdependence and cooperation of pastoralism and agriculture, rather than considering them as discrete systems. The context examined here is different: actors pursuing different production systems compete to consolidate their physical power as a mechanism to constrain one another, because institutional arrangements favouring cooperation to stop attacks bring no additional gains for the strong actor (the Issa). This may tempt the weak to mobilise power resources.
Resource mobilisation
In the absence of the complete assignment of enforceable property rights, the capacity to mobilise resources invested in conflict provides an alternative means to secure access to grazing resources. Conflict, like any human activity, requires an investment. However, its reward may vary, depending on group competence and choice of strategy, and is interpreted on economic, moral and psychological grounds. Mobilising more resources in conflict is considered as a collective effort to reduce the cost of conflict. This becomes realistic only if resource mobilisation leads to decline in power asymmetry and ultimately motivates Issa and Ittu to shift from violence to compromise and negotiation. From the Ittu perspective, resource mobilisation is believed to deter the Issa, rather than giving rise to warfare in which one side wins and the other loses. Successful resource mobilisation increases efficient use of pasture when it creates such an incentive. To succeed in this, religious leaders and clan elders mobilise resources by allocating tasks to their community members. Accordingly, there are four groups of actors engaged in violent conflict through collective action, in which every member of a group participates in one way or another. For example, the council of village elders selects: (1) the herding group (five adults per village); (2) the youth and middle-aged group who penetrate into contested areas or resist raids; (3) the surveillance group who assess pre-mobility information; and finally (4) the village security group who guard the village from incidental attacks.
The Issa can mobilise more easily and possess more weapons than the Ittu for two reasons. First, the pastoral Issa moving freely across international borders have better access to modern weapons. Second, weak control of firearms proliferation has created a suitable environment for the Issa. Before the change in government in 1991, when customary rights were more enforceable, the Ittu owned more weapons than today. Many Ittu clan members were armed during the border war between Ethiopia and Somalia, and got support from the Ethiopian government to fight the Issa. The present government however disarms the Ittu for political reasons, such as a suspected link with opposition movements, including the Oromo Liberation Front (OLF). Ittu elders regarded such intervention as ‘unfair’, since nothing has been done to disarm the Issa. This has gradually created power asymmetry. Similar conditions in Niger reveal that wherever there is an unbalanced power relationship between pastoral and agropastoral groups, the weaker party's need is considered as illegitimate, while the dominant one becomes inflexible and avoids negotiations (Ngaido Reference Ngaido, McCarthy, Swallow, Kirk and Hazell1999).
Breaching of customary institutions
This section first describes the effectiveness of customary institutions when conflict occurs within an ethnic group, and then shows why such institutions have become ineffective. Both the Oromo and the Somali have their own customary institutions. Discussions with key Oromo informants show the existence of institutions embedded in the belief system called Abarsa and Kakku, which in turn emanates from the long-established constitution of the Oromo known as the Gada system. Both are used as tools in conflict management. Abarsa of elders and religious leaders is a statement or saying that God hears and acts. Abarsa and Kakku, literally translated as a curse and swear respectively, are the two most important institutions regulating conflict-triggering behaviour. Gada is a system of social organisation among the Oromo where laws governing access to resources and conflict resolution are established (Desalegn et al. Reference Desalegn, Seleshi, Regassa, Mukand, Ashim, van Koppen, Giordano and Butterworth2007).
Among the Somalis, one can find a well-structured conflict management system involving an inter-clan council (Guurti) consisting of elders who apply the Xeer as an institutional framework. It is through the Xeer that relations between clans are governed not only in conflict management but also in sharing pastoral grazing resources (Beyene Reference Beyene2008). For example, CHF International (2006: 13) states:
The Xeer is not formally codified but in reality trumps any laws created by the state. It is the set of rules, regulations, and values that form the foundation of Somali society. Xeer can also represent agreements between sub-clans that govern their relations and lay out rules for interaction, but not all sub-clans have Xeer with one another.
When it comes to conflict management between the two ethnic groups, it has been a tradition that the group that provoked conflict sacrifices bulls. Recently, however, the nature of the conflict in relation to socio-economic and political factors has changed, as a result of ethnic-based federalism.Footnote 11 As the basis of the conflict has expanded from resource access and use to territorial expansion in connection with the formation of regional states and ethnic-based federalism, the capacity of elders to manage the conflict has deteriorated. This shows the close link between change in political system and the declining capacity of elders to overcome conflict. A related work shows that the decentralisation of governance has introduced new forms of resource competition and struggle between pastoral and agropastoral groups, undermining ‘traditional reconciliatory mechanisms’ (Shide Reference Shide2005). Traditional norms are most effective in resolving resource-based conflicts occurring within clans of specific ethnic groups or at village level under conditions of social proximity and cultural homogeneity that make communication faster and sanctioning easier. As widely discussed in the literature, customary institutions with conflict management functions are deeply embedded in social and cultural norms (Watson Reference Watson2001; Sanginga et al. Reference Sanginga, Kamugisha and Martin2007; Unruh Reference Unruh2006). However, the mechanisms for negotiating agreements between ethnic groups are less successful when there are unequal power relations, and conflict of interest is defined along identity lines (such as ethnicity) rather than simple resource needs (Cousins Reference Cousins1996).
These weaknesses of the customary authorities in resolving conflicts have increased participation of state representatives, though their task is limited to facilitation of elders' meetings. They simply play a mediating role. In those meetings, elders from the two ethnic groups investigate the reasons for conflict and identify the group who initiated the conflict; they then determine the compensation for property and livestock lost during conflict, and make the parties agree to avoid further attacks. Payment of compensation is actually very unusual. Important questions including how grazing resources should be used, who has the right to graze where and under what conditions remain marginal. Most interviewees felt that such questions were unimportant, as the grazing area has been communally used.
Passive state involvement has undermined the ability of local efforts to produce positive outcomes in reducing the incidence of conflict. Negotiations usually lead to promises to refrain from further raids and killings. However, the effect is usually short-lived. The actual practice often turns out to be the other way round. An outbreak of livestock raid-related violence is observed shortly after mediation and negotiations. Such violation of commitment is one of the reasons for the recurrence of the conflict. For the Issa, the opportunity cost of not raiding is high, even though this creates insecurity for the Ittu. An alternative choice for the Ittu is to shift to confrontation as already described, rather than going for another round of negotiation. This experience augments the extensively documented evidence on the limited capacity of customary institutions in managing resource conflict between pastoral and agropastoral groups (Ocan Reference Ocan1994; Kratli & Swift Reference Kratli and Swift1999: 30–5; Vedeld Reference Vedeld1994).Footnote 12
The key lesson to be learnt from this analysis is that conflict resolution efforts concentrate more on compensation and punishments of the wrongdoer, rather than developing clear rules of resource access. The victim of the conflict attaches more value to the compensation to be made for the lives and livestock lost. But this has a temporary remedial effect. Some authors argue that such ‘palliative processes’, by attempting to promote interaction between conflicting parties, divert attention from the underlying structural causes of protracted conflict (Cousins Reference Cousins1996; Shide Reference Shide2005). One of these structural causes is the difference in power and unclear property rights arrangements for the different land use systems evolved over time.
Benefits from livestock raids
Livestock raids produce a threat to the weaker actor using communal grazing land. There is a direct link between fear of raids and avoidance of the grazing area. The gradual increase in the scale and severity of conflict is partly associated with ever-worsening livestock raiding. Results of the household survey show that over the last five years, the three Oromo clans have lost 446 animals, with an average of 11 animals per household. Raiding has been organised on a commercial basis, involving accumulation of wealth by one group at the cost of impoverishing the other. As there is retribution, this practice is expected to be reciprocal among different resource-users. However, the benefit that the Issa gain from raids is different from that expected. The robbed animals can be instantly sold at local markets or trekked for export to neighbouring Djibouti. Moreover, informal export arrangement ahead of raiding is another reason for engagement in raids, which shows the trans-boundary flow of benefits from raids. Meanwhile, such arrangements help the Issa to evade repossession efforts by the Ittu.
Raids stimulate conflict since part of the revenue generated from them is invested in conflict technology. Access to better conflict technology generates an incentive to perpetrate further raids. There is always a temptation to violate customary institutions. Hence, the higher the benefits from raids, the more frequent the violence will be. This poses a challenge to the transformation of conflict and emergence of customary arrangements improving access. Some speculate that when property rights are unassigned, power asymmetry compounded over time encourages actors to invest in coercive activities in order to gain an advantage over their opponents (Skaperdas Reference Skaperdas1992). Though livestock raiding is often high when the Ittu are grazing on the contested land, in exceptional cases it occurs at grazing places close to the villages. This warns the agropastoralists that more losses could occur if they are away for any time from their permanent settlement. While underlining the severity of raids and loss of lives, a religious leader and PA council member of Annanno peasant association in Mieso district states:
The fact that the Oromo agropastoralists have been disarmed due to the political allegation that they are supporters of Oromo Liberation Front has intensified the livestock raids, human life losses and general insecurity all of which led to chronic poverty in the area. For instance, our people cannot sleep in their house due to increasing fear. In the year 2001 alone, 40 household heads were slaughtered in their houses during the night. Mr Mohammed Abdela who is my neighbour and was chairman of Annanno PA in 2003 has lost 28 cattle due to raids. He is now relying purely on food aid and community support.
Interview, Annanno peasant association, 12.11.2004Issa's gains from raids are closely associated with their strategic location. They are ethnically affiliated with political power holders in Djibouti, and inhabit the main trade route connecting Addis Ababa to the major port. This route has become vital for the national economy since the outbreak of Ethio-Eritrean war in May 1998, and has in turn resulted in a shift in the Ethiopian state's position against taking any measures to control Issa engagement in local conflicts. There is also open cross-border movement and uncontrollable import of firearms. From the Ittu perspective, all these factors are important in creating an environment that perpetuates violence. Some interviewees suggested that the reason for the conflict goes beyond questions of grazing land and raids. They perceive these as short-term economic gains with the long-term goal of expelling agropastoralists from their permanent settlement. However, this paper emphasises the short-term interest, which is clearer, and largely sensible, given the opinion of district experts. In situations of on-going food insecurity, socio-political volatility and ecological variability, raiding between herding groups is practiced as a strategy to defend and acquire access to vital resources (Hendrickson et al. Reference Hendrickson, Mearns and Armon1998).
A quest for solutions
This section discusses state interventions in 2005 and afterwards to prevent the conflict between the two ethnic groups. Discussions with the various district administrators and elders indicate that there has been a search for further political measures to calm down the situation through the formation of a peace committee consisting of ten elders from both groups. The committee was set up through the facilitation of the federal government in an attempt to revitalise the limited role of customary leaders to manage conflict. This has been in effect since March 2006. It was a scheme with new ambition to end the violence. The elders swore on behalf of the group they represent in a customary fashion to perform their duties in conflict management that involve monitoring conflict-inducing actions and reporting to the district authority. Meeting people from various villages on market days is the major means by which the elders become aware of the symptoms predating any conflict. This enables elders to collect information to take immediate proactive measures before conflict gets out of control. Meanwhile, the district authorities have given a few of these elders training and a permanent office in Mieso town. Due to the short time interval between this important move and the fieldwork period, it was not possible to observe how effective this has been.
However, a comprehensive assessment of the role of customary institutions in pastoral regions of Ethiopia indicates the diverse experiences (success and failure) and variable judgement of stakeholders about state intervention in backing elders to resolve or manage conflicts (Unruh Reference Unruh2006: 229). Although customary institutions are legally recognised with respect to conflict resolution, the assessment indicates that a lack of capacity at a very local level limits the amount that they can do in practice (ibid.).
The perpetuation of conflict also serves as a means to generate funding for conflict prevention. The allocation of budgets to mitigate conflicts in the marginal areas of the country by federal and regional governments provides local actors with an opportunity to increase their earnings (Warsame Reference Warsame2005).Footnote 13 In addition, one study shows that government-appointed elders collect bribes from the guilty side by sharing the cash from the sale of livestock raided. Such corruption and distortion may make customary conflict management efforts unreliable, and lead to the emergence of rent-seeking behaviour (Gebre Reference Gebre2001). In short, a wide range of literature on eastern Ethiopia emphasises that the political instability arising from the introduction of ethnic-based federalism has increased uncertainty at the local level, as the federal government continues to manipulate elections and political elites. This political process has gradually put a limit to the customary procedures to manage the conflict, due to a deterioration of public trust in elected local officials, who often intervene in the traditional conflict management process (Hagmann Reference Hagmann2005; Lister Reference Lister2004; Markakis Reference Markakis1996).
DELINEATION OF RIGHTS AND MANAGING CONFLICT
The previous section described the causes of inter-ethnic conflict, the historical role of the state in triggering or managing conflict at different times, and the overall setting of the conflict. The causes of the conflict are multidimensional: poorly defined property rights, resource scarcity, politics (boundary disputes) and livestock raids, where one cause is linked to the other. All these can be categorised as property rights conflicts. The purpose in this section is to bring together the empirical experience in other pastoral areas, the case study evidence and the theoretical explanations on the causes of conflict. An attempt has been made to delineate conditions that would support customary institutions in managing conflict. An important question is then: ‘What type of property rights arrangement is required to ensure security of rights and to undermine conflict?’
There are some insights from empirical and theoretical literature in establishing secured rights for multiple resource-users who can potentially come into conflict over resources. Based on experience in other parts of Africa, some emphasise the establishment of formal institutions for negotiation over rights (Scoones 1995a); others suggest that it be left to the conflicting parties themselves, given that they are self-organised (Sylla Reference Sylla and Scoones1995). For example, Thebaud and Batterbury (Reference Thebaud and Batterbury2001) indicate the success of customary institutions in avoiding exclusive rights and undermining the incidence of conflict over pastoral resources by arranging access year round. Alternatively, the creation of new institutional structures supporting the co-existence of formal and informal institutions can be effective when conflict becomes acute and internal capacity is limited (Cousins Reference Cousins1996).
There is no obvious and easy answer to the question posed above. But there are nonetheless theoretical views that may provide some hints. Two of them in particular are pertinent to this case. The first suggests establishing ‘a clearly defined boundary’, which is often regarded as a key design principle for commons management. This is useful since rights granted along with an observable and unchanging boundary will reduce enforcement costs of those rights (Ostrom Reference Ostrom1990). The second view examines the conditions under which this principle can be put into practice. A well-defined boundary is effective when the flow of benefits is predictable and groups relying on a resource are stationary. However, when there are large variations in benefit flows and the group relying on the resource system is mobile, then resource boundaries should be fuzzy to accommodate variations in group needs and resource flows (Agrawal Reference Agrawal2001). The implication is that rights have to be defined in accordance with the resource conditions whenever claims are made. This view coincides with that of the new rangeland ecologists, who propose that territorial boundaries should be fuzzy in order to provide margins of manoeuvre (Behnke et al. Reference Behnke, Scoones and Kerven1993; Behnke Reference Behnke1994).
These lessons have an implication for the delineation of rights to the contested land. While the first view implies a need for the state to establish group rights, perhaps based on identity, the second concept implicitly carries the assertion that the state has to enhance the capacity of customary leaders (clan heads) to define rights based on resource-users' needs, i.e. flexibility, as suggested in most empirical work (Goodhue & McCarthy Reference Goodhue, McCarthy and McCarthy1999). If institutions for flexible access options are not in place, an excluded group suffering from increased vulnerability at a point in time will encroach on the territory of other groups with relatively better pasture. To induce flexibility, the findings suggest state or third party involvement to improve the relative bargaining position of the weak group, as persistence of power asymmetry puts a limit to attaining stability. This is also based on the work of Vanderlinden (Reference Vanderlinden and McCarthy1999) and distributive bargaining theory (Knight Reference Knight1992) that support the view that power asymmetry undermines the capacity to negotiate and the possibility to sustain negotiated agreements.
* * *
This paper examines causes of conflict and identifies factors supporting or inhibiting customary institutions in managing resource conflict. Evidence from the fieldwork shows the embeddedness of interdependent situational variables that are central to the evolution of conflict. Some of these are economic incentives associated with livestock raids (explained by power asymmetry), power asymmetry, changes in land use influenced by state intervention, breaching of customary norms, and historically based political factors that have negatively affected the relations between the two conflicting parties. A study carried out in the same district concludes that competition over new sources of revenue, manipulation of local elites and the politicisation of ethnicity have increased the intensity of conflict (Shide Reference Shide2005).
In traditional pastoral societies, livestock raiding was believed to play a crucial role in redistributing wealth provided that groups undertake retribution, which could create stability in the pastoral economic system. However, the findings of this study indicate the unidirectional flow of benefits from raids under conditions of power asymmetry. The findings also show that a rise in the level of raids prevents the weak party from using contested communal land. Insofar as the agropastoralists rely mainly on livestock as livelihood source (direct source of food, soil fertility improvement and traction power), raids can increase vulnerability to food insecurity and poverty either through asset losses or by restricting access options and causing localised range degradation. As a result, livestock production is constrained. For the strong party, flexibility between violence and negotiation secures economic gains. The incentive embedded in this practice limits the capacity of customary authorities to manage conflict. Therefore, in a mobile (agro)pastoral resource use system where distinct ethnic groups are characterised by power asymmetry, traditional authority has limited capacity to enforce property rights to grazing resources and manage conflict. This has resulted from a change in political governance system that has increased power asymmetry and reduced the bargaining power of the Ittu. Although it is theoretically appealing that the Ittu need to be empowered in order to increase their bargaining ability, such an action needs to be paralleled by persuading the Issa to negotiate over rights, undermining their long-term incentive to form cross-border coalitions, using their ethnic ties, which might trigger inter-state conflict or at least increase regional instability.
Can the state do anything? Insights from the findings show that delimiting a strict boundary between the two groups has little meaning. The state can play a central role in facilitating the definition of property rights, by building the capacity of clan elders through which local level negotiations will create a means for flexible access to the grazing resource. This action is desirable since different patches of range resources have different economic values for a set of users in different periods, depending on livestock species kept. This may not be realised without devising effective political measures. The first step should be to discourage livestock raids that frequently trigger conflict. These measures may break, or at least weaken, the link between disputes and violent forms of property rights conflict. The second measure needs to be a change in the role of the state from mediation to facilitation of rule enforcement and internal negotiation on rights. An important move could be to work in favour of changing public image towards locally elected state representatives to build trust and to increase transparency in their action. This is crucial in a democratisation process that involves decentralisation of power to regional governments, and where the court system and decision-making responsibility that support conflict management practices rest with the regions. Strengthening the offices within each region that deal with relations with its neighbours could advance the role of the state in managing property rights conflict.
The pastoral group claims access to grazing land on a historical and customary basis without considering state allocation of land for the agropastoralists, because state intervention in this respect has put pressure on the grazing commons. However, the establishment of internal negotiation on rights can affirm the recognition of customary rights of both ethnic groups. The success of this is determined by the extent to which state law provides a back-up and enables local state agencies to work closely and openly with customary leaders. Experiences from many countries indicate the success of such an approach in managing resource conflict among different groups (Haro & Doyo Reference Haro and Doyo2004; Hussein et al. Reference Hussein, Sumberg and Seddon1999; Vedeld Reference Vedeld1998).
