Institutions and Pastoralist Conflicts in Africa: A Conceptual Framework

Pastoralist conflicts are important global development outcomes, especially in Africa. Analysing relevant literature on this phenomenon, we identify “institutions” as a key but fragmented theme. This blurs a composite understanding of how institutions affect these conflicts and their management. Hence, this article proposes a conceptual framework that brings harmony to this discourse by analysing 172 relevant publications. The framework was then tested using evidence from interviews and policy documents collected on a typical case in Agogo, Ghana. The findings show that pastoralist conflicts in Africa are shaped from three main dimensions: institutional change, institutional pluralism, and institutional meanings. Thus, state-level institutional changes create different institutions at the community level, and stakeholders using these institutions place different evaluations on them based on obtained outcomes. These dynamics contribute to conflict management dilemmas. Hence, the study recommends that intervention efforts examine whether new institutions contradict existing ones and to resolve them before implementation.


2015)
. In other instances, it is a contest between pastoralists over grazeland. Episodes of these conflicts have been reported between Bedouin herders and Jewish farmers in Israel (Tubi & Feitelson, 2016), amongst the Ghaddi of Himachal Pradesh in India (Agrawal & Saberwal, 2004), amongst Somali and Oromo pastoralists in Ethiopia (Beyene, 2017), and between smallholder crop farmers and livestock farmers in Southern Africa (Chakeredza et al., 2007).
Our analysis of the trends in global publications on the subject (see Figure 1) indicates that the thematic focus of these analyses has been evolving. Moreover, whilst cases in China and India have received considerable attention, African cases have attracted the highest publication density (including Kenya, Ethiopia, Tanzania, Uganda, Sudan, and Nigeria). Therefore, zooming into the African literature (Figure 2), we notice a general thematic evolution from themes such as "natural resources," "competition," and "land use" (2006-2008 era) to "wild-life," "institutions," and "pastoral conflicts" (2009-2011 era) and into the current era of "conservation," "information," and "climate change" (post 2012 era). Generally, "institutions" form a central theme in the evolving body of literature on pastoralist conflicts. This is noticeable in our visualisation map and is unsurprising because if we consider institutions as the rules by which access to community resources such as pasture are regulated, it is obvious how central they are to these conflicts. Hence, we notice that whether from the perspectives of old institutionalism, new institutionalism, or post institutionalism (Sakketa, 2018), many scholars refer to the role of institutions in their analysis of pastoralist conflicts.
Generally, "institutions" form a central theme in the evolving body of literature on pastoralist conflicts Yet, these institutional analyses remain fragmented. Consequently, a comprehensive understanding of how institutions and their dynamics influence these conflicts, and their management, is lacking in the scholarly discussions on the subject. This article intends to contribute harmony to these debates. Specifically, we seek to answer the question: how does the institutional environment in Africa impact the phenomenon of pastoralist conflicts on the continent? We propose a conceptual framework useful for understanding how colonial and postcolonial state transitions and politics impact the nature and function of institutions and which in turn shape the progression and management of these conflicts. As the analyses would show, these dynamics can be understood from a three-dimensional framework comprising institutional change, institutional pluralism, and institutional meanings, and a typical case in Ghana will be used to empirically illustrate this framework.
This objective to conceptualise a common institutional framework for pastoralist conflicts in Africa is necessary because many of the African cases have a significant cross-border dimension (Gonin & Gautier, 2016;Idowu, 2017), and as our discussion of the literature would show, there is a shared institutional environment across the continent. Yet, institution making and interpretation have become spaces of fierce contention and confusion within academic and policy debates. We notice that the institutional analyses around pastoralist conflicts have been case-specific and particularised. Although this is useful insight, the cross-border dimension of these conflicts begs for a general framework useful for consistency in these analyses across cases. This is what we contribute from this study. Besides, our test case study of the Agogo pastoralist conflicts in Ghana shows that to understand why some peace efforts targeted at these conflicts are not successful, the complex institutional context needs resolution, and our proposed framework helps to achieve this.
The rest of this article progresses by explaining the research methodology and examining the literature on pastoralist conflicts to reveal three major dimensions of the institutional discourse. These dimensions are then organised and proposed as a conceptual framework, which is then applied to analyse collected data on the Agogo conflict in Ghana. The conclusion justifies the application of the conceptual framework and makes policy recommendations based on our findings.

Research Methodology
We adopted a qualitative research strategy to analyse evidence from both secondary and primary data. The secondary sources mainly included 172 journal articles, which formed the basis for the conceptual framework. These articles are Scopus-indexed publications obtained from the search for "pastoralist conflicts in Africa." Analysing them followed a sequence of two steps. First, using VOSviewer 2015 visualisation software ( van Eck & Waltman, 2018), we conducted a trend analysis of the bibliographic data. We followed this up with a structured and in-depth content analyses of the most relevant data aided by the NVivo Version 12 qualitative data analysis software (NVivo version 12, 2019). This was done by refining the sample to 33 most relevant publications (spanning the period 1993-2020) using the NVivo text search query for the term "institutions" in the abstracts and keywords. The full texts of these articles were then subjected to an in-depth thematic cluster and content analysis to identity the three institutional dimensions. Hence, the literatures that were analysed served both as a review of existing knowledge on the subject and secondary data for developing the conceptual framework. This is similar to how others have applied the VOSviewer (Caputo et al., 2018) and NVivo tools (Johnston, 2006).
The primary data were drawn from semi-structured interviews and policy documents obtained during fieldwork in Ghana from March to April 2017. The interviews and policy documents were focused on the challenges faced by the government of Ghana in addressing farmer-herder conflicts in Agogo, which is the country's most topical (agro)pastoralist conflict. Two major conflict management policies, the Fulani Cattle Evacuation policy and a military-police operation dubbed Operation Cow-Leg, were examined. Respondents were purposively chosen through snowball sampling, based on their direct experience in the management of the conflict. They involved 21 interviews across four categories of respondents namely; 10 farmers and four herders/cattle owners in Agogo and Kumasi, three chiefs (as custodians of land), two police officers and the District/Municipal police commander in Agogo, and a representative of the Ministry of Interior (in charge of national security) in Accra. The data from these sources were thematically analysed with the NVivo Version 12 qualitative data analysis software based on the earlier determined conceptual framework.

Findings
Our comparison of the global publication trend (see Figure 1) with the Africa publication trend (see - Figure 2) indicates a similar thematic evolution except that the term "institutions" gains more prominence in the African discourse. The term is co-current with the terms "information," "control," "order," and "pasture" as well as "farmers" and "herders." Our NVivo cluster analysis identifies "institutions" as the 30th most prominent theme (after correcting for irrelevant terms like "article"). In the general sample of 172 publications, the term was associated with other terms such as "policy," "rights," "property," and "state" (see Figure 3). In the refined sample of 33 publications, the term is associated with the terms "policy," "community," "government," "traditions," and "rangelands" (see Figure 4).
Two main insights are gained from these macro analyses. First, the institutional discourse around pastoralist conflicts in Africa has been framed around the interaction between state-led policies and community-led traditions about the use of grazelands, especially in relation to the engagements between (agro)pastoralists. Second, these interactions have been framed in terms of access to and use of information, creating order and controlling the use of pasture. With these hints from the macro analysis, we proceed to analyse in depth the most relevant publications on the theme "institutions." Using this in-depth analysis, we develop a conceptual framework that summarises the dimensions by which institutions have been implicated in pastoralist conflicts in Africa.   Shettima and Tar (2008) consider pastoralist-related tensions to be deeply rooted in the political economy of Africa and therefore regard the development of these conflicts in the region as a manifestation of what Kaplan (1994) terms "the coming anarchy." However, these interactions have not always been conflictual. For instance, in reference to interactions between farmers and herders or agro-pastoralist and pastoralists, the relations between them were historically largely cordial. The former offered farm produce in exchange for animal products from the latter (Moritz, 2010), mirroring the "Cain and Abel economics" that is sometimes found between these actors in the Sahel (van den Brink et al., 1995). Although conflicts inevitably arose from these transactions, they were resolved at community levels mostly by traditional leaders (Benjaminsen & Ba, 2009;Tonah, 2003). An important factor to note, however, is that political developments over the years in Africa have resulted in some institutional transformations and outcomes that have impacted the prognosis of these pastoralist conflicts. As we have pointed out, and our data indeed show, institutions are not new in the discourse around pastoralist conflicts. However, the institutional dilemmas facing the management of pastoralist conflicts are common across the African cases and can be synthesised into a three-dimensional framework that is helpful for cohesive policymaking and scholarly enquiry into the phenomenon. We discuss in the next sections this three-dimensional framework, namely, institutional change, institutional plurality, and institutional meanings. Such thematic analyses are relevant in any critical institutional analyses (Cleaver & de Koning, 2015), such as analyses conducted in this article.

Transitioning From Indigenous to Colonial to Contemporary Institutions
The first dilemma concerns the waves of "institutional change" that have swept across the continent. The first wave concerns the (attempted) replacement of indigenous African institutions with colonial ones. In a study of conflicts between farmers and cattle herders in the Great Lakes region, Markakis (2004) contends that the colonial land tenure system paid little attention to livestock production and thus did not significantly protect the herder community by designating lands to them. According to Markakis, post-independent African governments have not reviewed land policies to accommodate herding, making herder communities landless people.
There is also a second wave of institutional change in relation to the state-sponsored decentralisation in the origin and host communities of (agro)pastoralists. Generally, Africa hosts a very large proportion of decentralisation programmes being experimented in the world, and this has contributed significantly to conflicts sometimes by ossifying ethnic identities through formalisation of land titles as seen in South Sudan (De Simone, 2015). Such a change alters the architecture and motivations that govern communal conflicts between pastoralists themselves and between them and farmers. For instance, in Niger, pastoral communities contain very useful customary but informal modes of conflict management that help to manage everyday conflict. Yet, some scholars have identified that decentralisation schemes induce "the reworking of the relationships among hereditary customary authorities, elected government leaders, and appointed oficials" (Turner et al., 2012, p. 755). This relevance of customary conflict management institutions is also noted in Sudan's dar-system, which was historically effective at rationing the use of landed resources between farmers and herders until it was abolished by the government in 1970 without a replacement, thereby creating a lacuna and setting the platform for heightened tensions between farming and herding communities (Abdul-Jalil as cited in Olsson & Siba, 2013).
Our analyses identify three key outcomes of institutional change, which are, changes in state function, changes in land use, and institutional innovation. For instance, Basupi et al. (2017) note in Botswana that there has been an increase in the centralisation of land resource management and a growing influx of technocratic approaches that trump local communities' experiences. Meanwhile in Ethiopia, others have noted changes in the property rights regime through a trend of land enclosures from communal rights to individual rights (Beyene & Korf, 2008;Kamara et al., 2004). This ownership regime may in turn be linked to the increasing modernist development interventions by the state which has led to improved diversification of land use (Hundie & Padmanabhan, 2011) and the rise in crop agriculture at the expense of pastoralism in Mali (Benjaminsen & Ba, 2009). These changes do not seem universal though. Galaty (2016) asserts that there are exceptions to this trend in the semi-arid pastoral regions of Kenya where there seems to be a reinvigoration of communal property tenure.
Our analyses identify three key outcomes of institutional change, which are, changes in state function, changes in land use, and institutional innovation Other changes are related to technocratic intrusion and border fluidity. For instance, in Ethiopia, traditional institutions are weakened because of outside actors' intrusion to distort the application of the traditional system of common resources governance (Unruh, 2005). Another important change in Kenya is that there have been profound fragmentation and contraction of the pastoral commons due to increasing contestation of borderlands (Greiner, 2016). As McPeak and Little (2018) identify in the Borana-Guji area in Ethiopia, the creation of new administrative borders is the main identified source of conflict. These cross-border conflicts were also the initial objects of multilateral institutions such as conflict early warning system (CEWARN) that are used to monitor the outbreak of cross-border pastoralist-related violence (Hassan, 2013;Kasaija, 2013). The pull and push forces described above have led to some innovations leading to changes in the type of institutions used to govern (agro)pastoralists interactions. For instance, in the Masai region of Kenya, there was the introduction of common pool resources (CPR) system when state authorities realised that herd mobility is a key feature of pastoralist resource management in the face of changing environmental circumstances. Before such a system, mobility of pastoralists had been considered a problem. However, the introduction of the CPR system meant that the state was beginning to legislate over common rangelands in a way that accommodates mobility of pastoralists (Toulmin et al., 2004).

Institutional Plurality Same Conflict But Varied Rules and Sources of Authority
The second dilemma concerns the plurality of structural institutions governing socioeconomic interactions in Africa. For instance, in relation to land, Eck (2014, p. 441) argues that it is "the institutional structure of the legal system [that] is central to understanding which countries are prone to experience communal land conflict" and that land conflicts are particularly rife in legally plural West African countries where there exists a heterogeneity of adjudication institutions to which disputants seek redress. This argument can be extended to pastoralist conflicts which are practically contests over land-based resources such as water, fodder, and shelter. Therefore, in the contexts where conflict protagonists are used to and prefer customary institutions for the management of everyday conflict, the advent of neoliberally decentralised institutions (e.g., law courts) could increase the avenues for forum shopping. For instance, in Ethiopia, there are various but usually incongruent means for conflict actors to renegotiate institutional rules to lay claims to land (Lenaerts et al., 2014). Where different rules govern access to the same resources, forum shopping leads to conflicts about legitimacy. Beyene and Korf (2008) also identify the role that decentralisation could play in either exacerbating or mitigating the problem of institutional pluralism. As they observe in the Somali region of Ethiopia, the regional state has handed over definition of communal versus private user rights to clan rule with ambivalent results: Elite capture has encouraged some clan elites to drive forward a politics of enclosure, which excludes asset-poor households from benefit streams, encourages unsustainable land use practices (such as charcoal production as windfall gain) and disturbs customary reciprocity patterns among and between clans potentially encouraging violent disputes over resource access (p. 17).
Such developments have led to the redefinition of the relationship between the state and traditional authorities, which affects the nature of human interactions (Beyene & Korf, 2008, p. 17). In the Borana rangelands of Ethiopia, certain national policies already yield conflicts of authority between traditional and formal institutions, thereby burning bridges for cooperation leading to associated conflicts (Kamara et al., 2004). The availability of these plural institutions is also observed in how "different types of conflict are taken to different conflict resolution institutions" in Kenya (McPeak & Little, 2018, p. 119).
From the above scenarios, it may be tempting to think of plural institutions as only a bane to peaceful cooperation between (agro)pastoralists. Yet, pluralism does not only involve competing but sometimes complementary interaction: between state or government institutions on one hand and customary or traditional institutions on the other hand. Also, there is this complementary interaction between so-called informal institutions and formal institutions. Galaty (2016), for instance, has argued, with evidence from Kenya, "that property is best seen both as a formal state-of-law and an informal state-of-practice, of holding, using and claiming values, most importantly in land" (pp. 724-725). We consider this a suggestion that institutional pluralism should be considered a normal feature of the pastoral resource management practice. Cleaver (2001) also provides a case of positive institutional pluralism in Tanzania, where cross-cultural institutional borrowing is important in the relationship between formal and informal actors in the management of the commons.
Hence, most scholars' proposals for improving institutional governance calls for accommodating more rather than less institutional pluralism. Marcussen (1994), for instance, argues that the retreat of the state to make room for civil society should not be complete and suggests a type of coexistence between these two forms of institutions. Also, some have argued using evidence from the Sahel that any innovation in institutions aimed at solving pastoralist conflicts is doomed to fail if they do not incorporate pastoralist perspectives into state-funded institutional provisions (Bonfoh et al., 2016). Although we agree to accommodating plural institutions in (agro)pastoral resource governance, our case study would show that where this accommodation involves contradicting rules, conflicts persist, and their management is challenged.
There are other salient dimensions of plurality. For instance, we identify a link between national and international institutions. In this mix, there is a complex relationship between what the state institutions provide and what the international institutions like the CEWARN have to offer amongst states in the Great Lakes region (Kasaija, 2013). Another dimension of pluralism occurs in the context of armed conflicts. As Unruh (2005) points out, in the context of armed conflict, effective rule-making is weakened because few institutions can endure such stresses of armed conflict. As would be indicated in our case analysis, pastoralist conflict situations lead to the birth of new rule interpretations that complicate already laid down arrangements for determining rights and control over the use of the commons.

Institutional Meanings Distributional Outcomes Influence Actors' Evaluation of Institutions
A third dilemma concerns the plurality of meanings attached to institutions for conflict management, which is largely a by-product of institutional change and institutional pluralism. Here also, Eck's (2014) study on land tenure in West African countries concludes that "when there are competing jurisdictions with overlapping mandates, the legitimacy of the system is undermined and the resultant incentive structure for people to settle disputes extrajudicially through the use of vigilante measures is enhanced" (p. 451). Hence, institutions regulating the commons use in (agro)pastoralist communities could have outcomes that give some actors agency to alter or maintain the national and local policies to their benefit. An example is observed by Benjaminsen and Ba (2009) in the inland Niger delta of Mali where jowros (managers of pasture) have gradually lost power to the previously underprivileged rimaybé (low caste farmers). The social and economic outcomes that stakeholders receive from applying institutions are what represent the meanings of those institutions. A discussion of some of the literature in the next paragraph links some of these outcomes to different areas such as information sharing, (lack of) access to property, entrenchment or denial of rights and ownership, changes or introduction of power asymmetries, the onset of conflicts, and the achievement of economic benefits and livelihoods.
The social and economic outcomes that stakeholders receive from applying institutions are what represent the meanings of those institutions.
For instance, in Botswana, mobile pastoralists are concerned with their ability to provide for their herds and seasonal movements served as a means of controlling such grazing lands (Basupi et al., 2017). This also has implications for human rights since mobile populations under international human rights laws are allowed to have access to resources and services (Bonfoh et al., 2016). At the national and local levels, institutions mean the ability to negotiate customary rights to property (Beyene, 2009;Hundie & Padmanabhan, 2011), endorsement of farmers' private rights to croplands (Kamara et al., 2004) or the security of collective rights (Toulmin et al., 2004). Another dimension is access, such as access to markets for (agro)pastoral products (Reda, 2015) and access to information and training for pastoralists to improve their bargaining position (Toulmin et al., 2004). Other implications of institutions are determined by the ability to maintain a diversified livelihood (Amede et al., 2020), especially for the women and the poor in society (Hundie & Padmanabhan, 2011;Reda, 2015).

A Conceptual Framework
Informed by the discussions above, we propose a conceptual framework that highlights a relationship between three dimensions of the institutional discourse that we trace in our data. This framework draws attention to the cumulative, multidimensional, and multilayered nature of the institutional environment within which pastoralist conflicts occur in Africa.
In Figure 5, institutional change sits at the core of the institutional environment within which pastoralist conflicts occur. We argue that there exists a relationship between the three institutional dimensions that we have highlighted in the previous section. Thus, changes to institutions in Africa, especially at the macro level (national or state), trigger changes in the local institutions closest to the setting within which pastoralist conflicts occur. As our case analysis would illustrate, this can be exemplified by the rise of colonial (and later modern state) institutions and their eventual competition with existing indigenous ones. It may also be illustrated by the collapse of the state because of civil war leading to the influx of unchecked small arms and light weapons, such as already referenced in Ethiopia (Unruh, 2005). Such macro-level changes lead to meso-level changes in the type of structural institutions that (co)exist within a region or community. For instance, the persistence of traditional governance may mean that traditional arbitration systems coexist with Western-style court systems as would be illustrated in our case study. Eventually, this plurality leads to outcomes that have different social and economic implications for the different members of the community in conflict. As our case would show, these perceived outcomes lead actors within the conflict setting to either support or reject the status quo institutional framework used to manage the conflict.
Our conceptual framework posits that institutional change produces plurality, which in turn engenders varied meanings. As indicated in our diagram, this process may be reversible where changing meanings could create incentives for the sponsoring of more parallel institutions, and over, time the institutional trade-offs may gather colossal momentum to induce a macro institutional change. Although our case does not illustrate such a process, we contend that this macro institutional change can become possible in a context where community actors unite to accept local or informal rules at the expense of state level or formal ones. Hence, proponents of the latter would have to modify them to keep their relevance. Nevertheless, in this reversed process, we expect the change to be less gradual than vice versa. In the next section, we use our case in Agogo, Ghana, to illustrate our proposed conceptual framework and the relationship between the various institutional dimensions within which pastoralist conflicts progress and are managed.
Our conceptual framework posits that institutional change produces plurality, which in turn engenders varied meanings.

The Institutional Context of the Agogo Farmer-Herder Conflicts in Ghana
Although a peaceful democratic state in West-Africa, Ghana has had its own spread of pastoralist conflicts dotted across the country. Amongst these conflicts in Ghana, a prominent case is the vastly destructive and frequent relapse of clashes between so-called Fulani herdsmen and crop farmers in Agogo, in the Ashanti Akyem North District of the Ashanti region since the 1980s. The district has an enabling climate of good vegetation, making it very suitable and increasing the activities of both crop cultivation and livestock rearing and consequently increasing occasions of conflict (Kuusaana & Bukari, 2015). From 2001 to 2016, the Agogo conflicts claimed over 40 human lives and led to the killing of 600 cattle and destruction of extensive farmlands and crops (Bukari, 2017, p. 143). Since the turn of the millennia, successive governments in Ghana have sought to resolve the conflict through many approaches, including amongst others, police and military intervention, mediation, and ejection of herders from the community (Paalo, 2020).
For about a decade, two of the most applied approaches to addressing the conflicts in Agogo and its environs have been efforts at ejecting the Fulani through the Agogo Fulani Cattle Evacuation policy (Agogo Fulani Cattle Evacuation Plan Committee Report, 2012) and a military-police operation-Operation Cow-Leg (Baido, 2014;Paalo, 2020). The former is a policy by the government to move the herdsmen out of the Agogo community, and the latter is the police-military crackdown intervention any time violence ensues. However, both approaches ultimately target the eviction of Fulani herdsmen from Ghana (Interview, police informant, Kumasi, April 1, 2017). The Herder Expulsion policy was enacted by the regional Security Council in 2010 in reaction to numerous demonstrations and agitations by citizens of the host community against the herders (Agogo Fulani Evacuation Plan Committee Report, 2012; Interview Krontihene, March 26, 2017). Its major recommendation was to expel Fulani herders from the Agogo traditional area, and this recommendation was upheld by the Court on January 20, 2012 (High Court in Kumasi, 2012). Nevertheless, the conflict keeps recurring despite various conflict resolution and peacebuilding attempts. The recurrence of the Agogo (agro)pastoralist conflict is explained by its complex institutional context. This complex context is occasioned by temporal changes to institutions, their plural existence within the community, and the contrasting meanings that they yield for stakeholders.
The recurrence of the Agogo (agro)pastoralist conflict is explained by its complex institutional context.
Institutional change: In terms of institutional change, the policies of both colonial and postcolonial governments have greatly altered the functioning of traditional institutions such as chieftaincy, changing the dynamics of the conflict and peace interventions altogether. In an interview at his palace in Agogo, the Krontihene-customarily the spokesperson for chiefs in the paramountcy-indicated that because the colonial government was rather interested in cocoa and minerals (e.g., gold and diamond) than livestock, the land administration did not provide for cattle rearing and herding. The then colonial government's non-commitment to the livestock industry is corroborated by archival records (see Caruth, 1932), and interviews with some herders and cattle owners who lamented that although the chiefs usually grant the herder community land to settle, the Lands Commission and the district assembly usually decline to validate such lands as legitimately acquired. They argue that this is because the state has favoured crop farming to livestock. This scenario is further complicated by the understanding within the Agogo traditional area and Ghana as a whole, that herdsmen (usually, the Fulani) in Ghana are considered foreigners because most migrate from neighbouring Burkina Faso, Niger, Mali, and Guinea (Baido, 2014;Tonah, 2003). The question of autochthony is relevant in the conflict discourse and is complicated by the cross-border pastoralist exploits. As an officer of Ghana's Interior Ministry dealing with this conflict acknowledges, despite the attempt to create cross-border legislation to make it easy for movements across West-Africa, the "foreigner-indigene" aspect of the conflict remains pronounced. This is because member states fail to synchronise the Economic Community of West African States (ECOWAS) protocols with local norms and laws (Interview, head of research, Ministry of the Interior, Accra, March 31, 2017).
Besides, the decentralisation policies of the post-independence Ghana governments further place local authority under political representatives at the district assemblies, regarding issues such as settling farmer-herder disputes, which was hitherto the responsibility of chiefs and other traditional leaders. To assert the significance of their precolonial and early post-independence position, the chiefs strongly advocated for and are usually part of the state-led peace interventions, especially peace negotiations between farmers and herders. Given their vast control over land and continuous relevance amidst the modern governance structure, the chiefs allocate lands for projects and economic activities, whilst the Lands Commission issues official documents to validate the ownership or transaction over the land (Interview, Krontihene, Agogo, March 26, 2017). Yet, as already mentioned and would be further discussed in the next paragraph, the lack of cooperation between these two authority points has created conflicts about legitimate ownership of these resources.
Institutional plurality: The morphing of the institutions in charge of land allocation and management led to the rise of plural institutions at the regional and community levels, and this impedes peacebuilding processes and outcomes. For instance, all the respondents indicated that although the Agogo conflicts were historically usually customarily mediated, this role has been replaced by the local state courts. Yet, chiefs and traditional representatives serving as customary interpreters of land ownership rules form part of the liberal court procedures and outcomes. Also, land management, which in the Agogo area resides in customary ownership, has become increasingly formalised, and the latter claims superior authority in transferring or legalising land ownership for economic activities (Interviews Agogo youth and farmers representatives, Agogo, March 28, 2017). Such institutional hybridity intended as complementary has rather created plurality of institutions, leading to confusion and limiting the peacebuilding potential of both state and non-state actors. This confusion is complicated by the operations of partisan politics, whereby politicians campaign based on the dynamics of the conflicts for votes, but cannot sustainably address the conflict as promised, due to the deeply conflictual and confounding mix of institutions and actors (Interview, police informant, Kumasi, April 1, 2017). Similarly, the hybrid institutions cause political ecologies, in which "identity conflicts emerge, because of the formation of alliances within and across the traditional and state institutions, based on individual or group interest in the conflict" (Interview, Divisional police commander, Konongo-Agogo, March 25, 2017).
Institutional meanings: The plurality of institutional arenas, in turn, leads to plurality of meaning-making not only amongst conflict parties but also amongst policymakers. This is due mainly to the mixed (and sometimes) conflicting policy interventions the successive governments of Ghana have used to manage the conflict. These conflicting policies have been characterised by the deep-seated ambivalence in the successive government's intervention to expel the herders. On one hand, the government promotes expulsion whilst at the same time providing lands to herders. As the Fulani representatives explain, "initially the high court ruled that the land is a forest reservation. The government intervened and settled the matter . . . to prevent conflict the government allocated a different place for us" (Interview, cattle owner and representative of Fulani in Ghana, March 29, 2017). A major example of this ambivalence is that although both the operation cow leg and Fulani evacuation plan intend to eject the Fulani herders from their host communities and ultimately out of the country, the government of Ghana set up the national cattle ranching policy in 2017 to settle herders in large tracts of ranches in strategic locations within the country (Interview, Divisional Police Commander, Konongo-Agogo, March 25, 2017). According to a police informant and some cattle owners (in separate interviews), such contradiction occurs because state officials are stuck between observing ECOWAS protocols of free movement and establishment of regional citizens whilst also yielding to the host communities' pressure to expel herders and preserve local land rights.
Furthermore, the Fulani view the policies as hostile and unacceptable, and they make both cultural and legal arguments to support this. They consider the expulsion policies as a contradiction of West African subregional policies on the free movement of persons (ECOWAS protocol, 1979). They also interpret it as a loss of their access or right to the use of resources that they sometimes believe were rightfully acquired. For instance, Alhaji Gurusa, a prominent representative of the Fulani community in Ghana, and cattle owner, asserts in an interview that "the chief gave me the land because I am a Ghanaian . . . they (the Agogo leaders) say the Fulani herdsmen are not southerners, among others, but some local people do not even agree with such arguments, others agree" (Interview, cattle owner and representative of Fulani in Ghana, March 29, 2017). The citizenship claim here suggests what he considers to be at stake in this matter; that Ghanaians can have access to land but non-Ghanaians cannot. On the contrary, the traditional leaders of the host community say, "we do not have indigenous Fulani people here. They are aliens in Ghana" (Interview, Krontihene, March 26, 2017). The farmers then blame state actors for conniving with cattle owners and herders not to carry through the ejection plan.
Although these herders and cattle owners may not be indigenous to the Agogo community, they argue their citizenship through Chapter 3 of the current constitution, which provides various routes to citizenship (Constitution of Ghana, 1992). At the courts where these matters have been contested, the formal institutions weigh heavier than community considerations of heredity and ancestry. We, therefore, see a tension between the formal statutes for defining citizenship clashing with the traditional and customary interpretations of citizenship. For officers at the Ministry of Interior, this web of (mis)interpretations has hampered the successful implementation of the operation cow leg and herder expulsion policies and most of the mechanisms to address the conflict. Individuals within each group (policymakers, farmers, herders, women, men, etc.) think differently about who benefits from the status quo or existing institutions (Interview, head of research, Ministry of Interior, Accra March 31, 2017). Indeed, as the earlier interview narratives indicate, these meanings are usually neither fixed against nor for any group. At some point, farmers feel herders benefit from the status quo, and the reverse is true where herders feel the policies are meant to deprive them of their resources. Others have also pointed to politicians as the beneficiaries. These meanings are fluid and depend on what outcome is being considered by a stakeholder in a particular circumstance. Nevertheless, these different interpretations arise because of the alternative but contradictory rules that apply.
In summary, the analysis of this case largely conforms to our conceptual framework, but it offers even more insight. For instance, it points to a closer and direct connection between institutional plurality and institutional meaning. We notice that even though the plurality of institutions leads to plural meanings, actors rationalise their interpretations by drawing in more references to other parallel institutions. In this case, for instance, it was not just between customary and state courts, but also actors referred to ECO-WAS protocols and the national constitution. The lesson then here is that the institutional reference point informing why conflict parties take certain actions may be farther than their immediate environments.

Conclusions and Recommendation
This article demonstrates the temporal persistence but fragmentation of "institutions" as a key theme in pastoralist conflict studies in Africa. Hence, the objective was to offer a conceptual framework to harmonise the understanding of how institutions impact these conflicts and to demonstrate its empirical validity using a typical case in Ghana. We identified three main dimensions through which institutions impact such conflicts, namely, institutional change, institutional pluralism, and institutional meanings. We argue that these themes are connected in a logical flow. First, there is institutional change, which largely occurs by the replacement or reshuffle of institutions at the national or subregional level (macro) as the state transitions from native governance to colonial governance, to democratic governance. These transitions and the institutional remnants of each governance regime lead to the coexistence of multiple institutional forms available for actors at the community level to use in governing access to resources such as pasture. Then, as actors engage with these institutions, the distributional outcomes lead to different meanings actors attach to these institutions in terms of whether they are useful or not, fair or unfair, or legitimate or illegitimate. These meanings lead to either cooperation or resistance to the macro-and meso-level rules and policies that are installed for regulating access to these common resources.
From the research point of view, we demonstrate with our test case study how our framework can be helpful for case-based analyses of pastoralist conflicts in Africa. As we show in the typical (agro)pastoralist conflict case in Agogo, Ghana, these institutional dimensions individually and more cumulatively could inhibit sustainable peace. In other cases, perhaps they may engender cooperation. The case indeed validates our framework that the macro-level shift from customary to technocratic governance ushered in a history of institutional pluralities, making it difficult for state agents or community members to fully agree on a policy to address the conflicts. We consider that our framework may be applicable to cases beyond Africa, but we admit the limitations of such an application considering that we use Africa-focused data in arriving at this framework. In that sense, similar application of our methodology or framework on cases outside the African continent may yield other helpful insights.
From the policy point of view, we propose that the prognosis and management of pastoralist conflicts in Africa are enhanced by an application of this three-dimensional framework. For instance, subregional bodies in Africa are taking steps to stem (agro)pastoralist conflicts by making policies to regulate cross-border (agro)pastoralist interactions. In this regard, one main area of dilemma highlighted by our case is the question of whether it is state-centric definitions of citizenship and associated rights or rather customary definitions of citizenship and rights at the community or village level that matter in determining access to resources. Our case shows that the state-level definition of citizenship seeks more accommodation, but this sometimes loses touch with the community approach, which is more restrictive. Hence, these two frames need to be better connected and made complementary.
Finally, we acknowledge that given the different nodes of meanings and negotiations surrounding policies for peace, it may be difficult to address all the changes, pluralities, and largely conflictual meanings attached to such policies. Yet, a small starting point is useful. Policy proposals to address pastoralist conflicts should examine whether new institutions conflict with existing ones and to resolve them before implementation. When these considerations hold a central position in discourses on pastoralist conflicts, the likelihood of policy success is enhanced.
Policy proposals to address pastoralist conflicts should examine whether new institutions conflict with existing ones and to resolve them before implementation

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research and/or authorship of this article: The fieldwork to collect data on the Agogo case was partly sponsored by the Patrick Duncan graduate research fund at the University of Oxford.