Towards an Arid Eden? Boundary-making, governance and benefit- sharing and the political ecology of the new commons of Kunene Region, Northern Namibia

Over the last two decades many sub-Saharan African countries have devolved rights and obligations in rural natural resource management from state to local communities in an effort to foster social-ecological sustainability and economic development at the same time. Often these governmental projects were launched in settings in which traditional commons, informed by both the demands of traditional subsistence-orientated agrarian systems and the tenure policies of colonial and postcolonial states were well established, and in which power struggles between rivaling traditional authorities, between seniors and juniors, and between state agents and local communities were pertinent. These moves were also embedded in (partially contradictory) discourses on decentralization, political participation, economic empowerment, and neo-liberally inspired commoditization of natural resources. In the process of devolvement rights and obligations were handed over to communities which were formalized in the process: formal membership, social and spatial boundaries, elected leadership, established models of governance, and accountability both to the wider community and to state bureaucracy. New commons were established around specified resources: pastures, water, forests, game. In the process these resources were (partially) commoditized: game owned by the community could be sold as trophies for hunting, lands could be rented out to private investors, and water had to be paid for. This contribution is intended to shed light on the process of establishing new commons – in the local context named conservancies – of game management in north-western Namibia. Game on communal lands had been state-owned and state-controlled in the colonial 1 The introduction of the Special Issue has a longer passage on the origins and the use of the concept


Introduction
During the past two decades natural resource management in the arid savannah of northern Namibia has been reshaped profoundly. It was not privatization or extended state control that set the tune for these changes, but determined efforts to reorganize communal natural resource management, and aspirations that this process would facilitate both increased levels of sustainability and the valorization and commoditization of natural resources. The argument was simple: once natural resources were valorized people would take care to manage them rationally to ensure continuous returns; valorization, commoditization and an opening up to emergent markets would ensure increased local incomes and a diversification of livelihoods. Two measures were necessary to achieve this: local institutions needed to be efficiently organized in order to ensure governance and surveillance of natural resources and a gearing between private sector investors and communities needed to be organized in order to facilitate marketization and income generation.
When Namibia became independent in 1990 there was an immediate need to reorganize conservation in the communal areas of northern Namibia: the former homeland Kaokoland had no legislation in place, 2 poaching by locals and probably more often by officials and South African politicians was rampant (Reardon 1986;Owen-Smith 2010), and resource degradation highly visible, (Owen-Smith 2010). Poverty, exclusion from markets, and the rule of traditional authorities were major challenges (Bollig 2006). Community-Based Natural Resource Management (CBNRM) became the umbrella term for a number of initiatives which were meant to establish and/or strengthen communal institutions capable of managing natural resources in a sustainable, economically rational and socially equitable manner (Bollig and Menestrey Schwieger 2014;Schnegg and Bollig 2016). These projects and programmes established new commons which were grafted onto and blended with earlier forms of communal management, e.g. with the predominant role of traditional authorities in the context of natural resource management and with heavy-handed state control of game. These initiatives were led by different ministries (Ministry for Environment and Tourism for game, Ministry for Agriculture for water, forests, and pasture) and were situated in the broader context of decentralization and community-based conservation affecting much of the Global South in the 1990s in the wake of the seminal Rio 1992 Earth Summit. CBNRM was designed as an effort to co-manage natural resources between the state and local communities and at the same time to valorize natural resources adequately and make them accessible to larger markets, thereby improving the income situation of local farmers. In many ways environmental and social challenges were to be solved through market-based solutions (Sullivan et al. 2016, 14).
In north-western Namibia the establishment of these new commons was facilitated either by ministerial extension staff (in the case of water) or by NGO staff and personnel of the Directorate of Nature Conservation and later the Ministry of Environment and Tourism (in the case of game). Both had to negotiate their programs with established institutions of resource management and actors legitimized through these institutions (e.g. traditional authorities, staff of line ministries). These pre-independence (that is, in Namibia, pre-1990) forms of communal resource management had on the one hand been deeply informed by the necessities of a mobile pastoral livelihood, entitlements to places and resources mediated through descent and ancestral beliefs and well-established pastoral patron-client networks (Bollig 2006). On the other hand pre-independence forms of resource tenure were shaped by colonial modes of surveillance and control (Bollig 1998;Van Wolputte 2007;Friedman 2011). South African dominance in northern Namibia had lasted until 1989, and all institutions of resource governance were deeply impacted by colonial legacies. Traditional authorities had been particularly tied in with the colonial administration: they had received salaries, enjoyed police support, and were granted territories.
Where exactly the aspirations and needs of a colonial administration intersected with local traditions differed from resource to resource: while the administration installed a grid of drilled boreholes, technically managed by an administrative office and in their day-to-day management controlled by local neighbourhoods, the administration of pastures remained in the hands of local neighbourhood councils and traditional authorities (Bollig 2006). Game management was usurped completely by the state, rendering any local use of the resource by locals illegal from the 1920s onwards, while permitting limited use by white officers through the pot-licence system (Bollig and Olwage 2016). Even traditional authorities who otherwise enjoyed many privileges were not allowed to hunt.
Establishing community-based natural resource management by instituting and reframing common pool resources had different implications 3 : In the case of game -a resource that had been owned by the state -partial entitlements were to be given back to the community. On an annual basis the responsible ministry gave a scientifically determined and locally negotiated quota to the community for the conservancy committee to decide what should happen with it. This quota, which is seen as a kind of reward for the efforts made and costs borne by a community in conserving wildlife, has to be considered as the governmentally determined commoditized share of game. The fact that part of the new commons (the quota) must be negotiated with the state bureaucracy and that the communities do not control resources autonomously is of particular interest to this article. With the gazetting of the conservancy a community also receives the right to tender land for rent through private investors. 4 Again, this has nothing to do with property rights to land -strictly speaking such property rights stay with the state -but tradeable use rights in parcels of land are given to a community on the provision that the community has established concise and governmentally endorsed governance structures and a management plan (Jones and Weaver 2009). Hence, if I speak of new commons in this context, these are commons in a specific sense. The communities gain limited management and transfer rights over game and land. Ownership rights in both instances remain with the state, and the rights devolved to communities have to be negotiated annually (in the case of game quotas) or at less frequent intervals (in the case of land rentals). The community can gain income and profit from these newly gained rights mainly when selling them (or renting them) to private sector investors and businesses. Of course, they may hunt game themselves and may profit in times of drought from grazing stored away in core conservation zones, but the major profit to be gained lies in the transferal of rights into cash income and jobs. The natural resources "captured" under this regime are moved from a state-owned phase into a community-owned phase, are then commoditized, and finally become privately owned by e.g. a trophy hunting company. To make this clear with an example: before a kudu ends up on a quota list it is state-owned; it then enters the quota phase and all decisions pertaining to it are made by a conservancy committee. The rights to the kudu are then sold on to a commercial hunter who in turn sells the animal to a client who typically comes from abroad.
The institutional structures and the mandate of the new commons are clearly delimited by law and are based on the assumption that substantive rights remain with the state. The clear legal restrictions on these new commons do not preclude that people perceive that more substantive rights to land were gained when a conservancy was gazetted, or that the institutions they devise to govern the communal resource -game -actually have a much broader mandate than that.
This paper focuses on conservancies in Namibia's north-west (Kunene Region) as prominent examples of an emergent practice of community-based natural resource management. 5 The contribution focuses on the social, economic, and cultural dynamics shaping the emergence of this new commons. It looks specifically into the construction and use of the game quotas. 6 After briefly describing the methodological approach, I will start off with a historical description of the development of the pastoral commons in northern Namibia, with the intention to show that commons have never been static, but have co-evolved within a political and bio-technical environment. I will then delineate the conservancy program and discuss economic and social dynamics linked to it, and highlight the hybrid character of institutions arising from co-management.

Methods
Data for this contribution result from a twenty-year engagement with communities in north-western Namibia. An initial phase of fieldwork was conducted for twenty-five months in the mid-1990s with a focus on strategies of risk management in a pastoral setting (Bollig 2006). More fieldwork followed in subsequent years for a number of months or weeks (the last short research period was in October 2015), resulting in a total of more than thirty-six months of fieldwork in the region. While most fieldwork was conducted with pastoral Himba communities in the Epupa area (nowadays the Epupa conservancy), from the late 1990s I also conducted research with other communities in the Kunene Region. Research dealt with the political ecology and environmental history of pastoralism in north-western Namibia in general, and more specifically with risk management, environmental change, and social exchange. Since 2002 research has been focused on new forms of commons management in general, and on the conservancies in particular. In 2012, 2014, and 2015 a series of qualitative interviews were conducted in ten conservancies with a specific focus on the development of social institutions within the new framework of conservancies. A survey was carried out in 2012 to determine to what extent conservancies contributed to local livelihoods and from what sources these incomes originated. In this survey I tried to procure data on changing perceptions of land tenure, emergent internal boundaries, income diversification and altered forms of mobility, At the same time I was involved in the LINGS (Local Institutions in Globalized Societies) 7 program, which was conducting research on emergent forms of rural water management in the area, after the state had handed boreholes back to communities and had urged them to establish communal water point associations (see contribution Linke 2015, 2016). Archival research in the Namibia National Archives in Windhoek accompanied ethnographic research. Generally the theoretical outlook on the emergence of new forms of communal resource management in north-western Namibia has been informed by approaches of political ecology and its interest in the articulation between local economies and larger markets and concomitant effects on valorization and commoditization. I have been keenly interested in discovering the ways in which global linkages (flows of discourses, institutional blueprints, and funds, markets) and local power dynamics have contributed to the specific shape of emergent forms of new commons.
3. The pastoral commons in a historical perspective: colonial state, local authority, and commons management In stark contrast to many East African pastoralists who define access to pastures along ethnic boundaries, pre-colonial land tenure among the Himba and Herero of north-western Namibia was different: pastoral households "owned" specific places that had reliable water, from which they organized grazing in the adjoining hills. A small number of households tied by bonds of kinship and patron-client relations managed dry-season and wet-season pastures together (Bollig 2013, 319). The heads of these place-owning households were addressed as oveni vehi, "owners of the earth/land". Ownership referred to their right to grant or to deny access to pastures and wells in a given area. The oveni vehi were also acting as patrons, lending livestock to poorer and often stockless households and attracting them as clients. In contrast to these rather well-defined pastoral commons, game was an open-access resource. 8 Oral traditions are quite explicit on the relevance of hunting in the pre-colonial set-up, but they lack any hint that game was owned by specific communities.
During colonial times, the system of pastoral resource management was reshaped according to the ideas of a colonial administration (Bollig 1998;Rizzo 2012). The South African administration established chiefs early on, and gradually expanded the number of traditional authorities, from three in 1923 to thirtysix in 1990 (Bollig 2006;Friedman 2011). From the 1930s onwards these chiefs nominated councillors who supported them, and established a tribal administration which specifically organized land tenure. In contrast to colonial authorities in East Africa, for example, the colonial administration did take into account local sentiments about legitimacy and chieftaincy: habitually the administration nominated a chief in consensus with the community, but in a few cases the administration also installed a chief according to its own considerations. This occasionally happened in situations where a number of incumbents were competing for the chief's throne. While at face value the colonial administration did leave the management of pastures and wells to local traditional authorities, in fact it impacted the local tenure system severely: e.g. mobility was contained, and far-reaching migrations were inhibited, and at times pastoral households even had to seek official permission when they wanted to shift their homesteads (Bollig 1998). Local hunting was prohibited altogether, all game being officially owned by the state. Poaching became one of the major offences for which local people were regularly prosecuted, 9 while limited hunting by white officials, visiting scientists and road working gangs was permitted.
The extensive borehole-drilling program of the South African administration in the 1960s and 1970s made a great number of new water points accessible (Bollig 2013). 10 Between the 1920s and the late 1990s the number of cattle herded in the Kaokoveld rose from c. 35,000 to about 200,000; i.e. while the livestock population grew more than sixfold, the pastoral population increased from c. 6000 people to about 25,000 people (i.e. about fourfold). The yearround accessibility of water on outlying pastures led to a complete reversal of the mobility pattern. Now livestock could be moved to distant pastures for eight to nine months during the dry season, whereas they had formerly been herded on these pastures only during the three to four months of the rainy season. The boreholes were fully maintained by the administration, and diesel was supplied free of charge to those boreholes fitted with engines. 11 Generally the 1950s and 1960s were a period in which paradigms of agricultural modernization overruled conservation agendas. Administrators of the semi-arid Kaokoland saw the future of the region in beef production. Measures for "vermin control" were 9 NAN SWAA 2513 File 552/1 Monthly Reports 1938Reports -1952Reports , 5/1939 instituted, and the hunting of e.g. jackals, and at times also of other predators, was actively supported. Other game species (e.g. Black-Faced Impala) were relocated to Etosha National Park. The 1970s and 1980s were also characterized by heavy poaching, the much bemoaned demise of elephant herds and tremendous decrease in game throughout the area (Owen-Smith 2010; Bollig and Olwage 2016).

Conservancies: development, governance, economic dynamics
In 1990 Namibia gained independence from South Africa and a new administrative setup was established (Friedman 2011;Wallace 2014). Like in many other sub-Saharan countries several decentralization reforms stipulated the devolvement of rights and obligations in natural resource management to rural communities in the second half of the 1990s. Since 1996 rural communities in Namibia could apply to the Namibian Ministry of Environment and Tourism (MET) for conservancy status in order to further their claims to game and other natural resources. While under the previous administration game had been "owned" and "protected" 12 by the state, now use-rights in game were to be devolved to local communities (Jones and Murphree 2001;Jones and Weaver 2009). This move was motivated by the aim to further rural development and rural incomes, to provide the conditions for private sector investment, to institute participatory planning, and to guarantee the conservation of game at the same time. It also intended to create more equal conditions between commercial (white) farmers and African inhabitants of communal areas: white farmers had already been given ownership rights to the game on their farms in the 1960s, and a sizeable game-hunting tourism industry subsequently developed on commercial farms. 13 The approach to delegating rights in game to communal farmers was based on the idea that once rural communities profited directly from game, they would be eager to protect it as a valuable common-pool resource. The Promulgation of Nature Conservation Amendment Act, 1996, defined conservancies and stipulated the way in which they were to be structured. Local communities were encouraged to establish corporate community-based organizations with a formalized membership, a well-defined territory of "jurisdiction", representative forms of internal leadership, and detailed management plans including the demarcation of core conservation areas in which no farming activities were allowed. In return, the MET delegated rights of game management to conservancies (Republic of Namibia 1996;Jones 1999;Owen-Smith 2010, 540). Once a conservancy is formally registered it may engage in contracts with the private sector (trophy-hunting companies, photo safari companies), lease out parts of the conservancy's land to tourism companies, and sell parts of the quota allotted to the conservancy to commercial hunters. The monetary gains of a conservancy were to be invested first of all in its upkeep (e.g. salaries for staff members, financing of meetings, transport). The remainder could be distributed to conservancy members or spent on communal projects.

The rapid expansion of conservancies
The number of conservancies in the Kunene Region has increased rapidly, particularly in recent years. In 2011/12 alone some fourteen new conservancies were gazetted in the region (see Map 1). The expansion of conservancies in the northern Kunene Region follows a clear pattern. In an initial phase, conservancies in the game-rich areas along the desert rim, along the boundary of Etosha Park and in areas linking both game-rich zones were gazzetted as conservancies. Then in two further consecutive phases, conservancies with less game and in more densely populated areas in the central Kaokoveld were also gazzetted. By 2015 almost the entire Kaokoveld was covered by conservancies (see Map 1). These figures suggest that after an initial push motivated by the activities of environmentalist NGOs (notably IRDNC in the Namibian context) and donors, the land areas most interesting for conservation had been gazzetted. Later on, actors in other communities, notably younger and educated males, realized that the move towards conservancy status conferred a number of potential benefits on individuals (e.g. employment, education, linkages) and the community (e.g. income diversification, clear-cut and legitimized boundaries) (see also Jones and Murphree 2001). 14 This resulted in a ripple effect, where the efforts of one community motivated activities in adjoining communities to apply for conservancy status. Certainly, the conservancies gazzetted more recently also saw potential benefits accruing from tourism and game quotas. The major motivation however seems to have been different! Repeatedly, the benefits of clear community boundaries were raised first of all. It was argued that such boundaries could enable a community in future to refuse outsiders seeking to access grazing land in their territory. Rather than securing exclusive rights over land the promulgation of boundaries was thought of as a way to protect access for the future. Although legally the state did not cede land ownership rights to local communities, but only devolved specific management and transfer rights to them, the delimitation of territorial boundaries fostered the idea held by local people that they had in fact wrenched land rights from the government (Bollig 2013 on Kunene; Hohmann 2004 on perceived land rights of San-speaking communities in connection with the establishment of a conservancy). While in normal rainfall years these boundaries would not be linked to rights of settlement and access to pasture and water, in drought they were made use of: In 2015 in the middle of a major drought, Himba homesteads which were settling in Herero-dominated areas in the southern Kaokoveld, some of them having migrated to those places many years ago, were summarily ordered to leave and to return to their original places of settlement. Most Himba had been denied membership of the conservancies in whose territory they settled; hence, they could be dubbed outsiders. Repeatedly it was argued that they did not adhere properly to the zonation rules conservancies had instituted -I will come back to this later. They were also rumoured to be implicated in a number of poaching cases.
In a number of cases the establishment of a conservancy was linked to the breaking away of an area from one chieftaincy and the establishment of a new chieftaincy; i.e. it was linked more to local divisive chieftaincy politics than to considerations of development or conservation. The Ozondundu conservancy's area (see Map 1, conservancy 26) had previously been under the chief residing in Ombombo, the main village in Okangundumba conservancy (see Map 1, conservancy 24). Only when local actors succeeded in negotiating the boundaries of Ozondundu conservancy as separate from Okangundumba conservancy could a new chief establish himself, making the newly established conservancy the area of his chieftaincy. In many instances boundary-making reconfirmed earlier chieftaincy boundaries and the administrative ward system of the second-tier administration (Sullivan et al. 2016, 11). In many ways boundary-making has led to a new type of territorialization. On the one hand these newly gazetted territorial entities conformed to the ideas and strategies of traditional and newly established leaders alike; in their view, bounded territories precluded unwanted immigration, (re-)legitimized and (re-)territorialized traditional leadership, reconfirmed communal ownership of pastures and other natural resources, and also opened venues for investment from the outside. On the other hand conservationists and administrative staff advocated spatial entities with clear-cut boundaries in order to facilitate their programs. The rapid emergence of conservancies is also a consequence of the quest of traditional authorities for influence, and of competition between rivalling traditional authorities for followers and income.
Boundaries could only be established in consultation with traditional authorities. Both NGOs and local administration were keen to involve those authorities, and the procedural rules of the Ministry of Environment and Tourism stipulated that it was mandatory to submit proof of the consent of the traditional authority together with the application for conservancy status. Hence, traditional authorities actively took part in this re-territorialization, reifying their chieftaincies' boundaries or creating new chieftaincies altogether.
Conservancies in the Kunene Region range in size from several hundred to a few thousand hectares (see Figure 1). Conservancies gazetted in 2012 were notably smaller and less populated than those gazetted earlier. This may be a hint that with the more recent conservancies, conservation motives became less salient: smaller game-management areas are certainly less attractive for conservation, while they may be better for reaching consensus on management and expenditure, and may better reflect local political structures and processes. Figure 2 shows the variation in population numbers across conservancies. While the smallest conservancy has only around 100 members, the largest has about 3500 members. Eighteen conservancies have less than 1500 members, eleven have between 2000 and 3500 members. Membership in a conservancy is easy to obtain. Most conservancies stipulate that any adult can become member who has lived more than five years in an area. A few conservancies allow for members who do not presently reside within the conservancy but are politically aligned with the traditional authority residing there (here membership comes very close to affiliation to a chief) or who have resided in the conservancy area for a long period in the past and whose absence is rated as only temporary.

Governing the new commons
Conservancies are governed by elected committees. The law does not exactly stipulate how such committees should look. The Nature Conservation Amendment Act of 1996 simply says that such a committee should be representative of the community residing in the area to which the application relates and that such a committee has the "ability to manage funds and has an appropriate method for the equitable distribution, to members of the community, of benefits derived from the consumptive and non-consumptive use of game in such area". (Republic of Namibia 1996). In Namibia each conservancy committee consists of a chairperson, a secretary, and a treasurer, along with their respective deputies. Committee members are elected at fixed intervals (between yearly and three-yearly) and, at least in theory, report to an annual general assembly. General assemblies are indeed organized in the greater proportion of Kunene's conservancies on an annual basis. Annual auditing procedures are currently being implemented for a larger number of conservancies by an NGO. When electing committees, most conservancies based the election process on a model of equal representation of villages within the conservancy; i.e. according to population size, each village was allowed to send a specific number of members to the committee. In this way all conservancy committees in the Kunene Region display similar structures. Committees usually consist of ten to fifteen people, and individuals are (re-)elected to committee positions at fixed intervals, usually every three years. In ten conservancies I enquired about the age, gender, educational status, and employment status of current committee members (Table 1). The majority of committee members (73.5%) are male, but women make up roughly a quarter of all committee members. The majority of committee members (70.7%) are younger than 40 years, and only 4% are older than 50 years: committees are not constituted by seniors but rather by "senior youth" (according to the local definition of youth). Half of all committee members fall within the age bracket of 31-40 years, and 25.1% are even younger than 30 years; i.e. more than three quarters of committee members were below 40 years. A number of the 25 committee members younger than 30 years are women. Committee members are fairly well educated. Roughly 40% had some secondary school education, many of them having finished grade 12, i.e. having successfully completed secondary school. Less than a third do not have any school education. In contrast to the fairly good educational status, the employment situation is miserable: almost 70% of all committee members did not have a job (some, however, had a job in the past). About a third were employed, and some 8.5% combined employment outside the conservancy with local engagement. The typical committee member is therefore male, between 20 and 40 years of age, has finished secondary school, and is without salaried work. All committee members interviewed sought in one way or the other to increase their herds. They were active herders. At the same time they sought to gain access to benefits from development programs and/or formal employment. From a local point of view committees are the arena for young and educated males to engage in their projects and visions of development and diversification of the local economy. The writing and reading skills of most committee members are necessary (or perhaps just make it easier) to communicate with NGOs, donors, extension workers of ministerial offices, trophy hunters, and tourism operators -and occasionally also with lawyers, who are increasingly made use of to settle conflicts. Many committee members are typical gate-keepers for their communities, capable of translating the ideas of donors, ministerial staff, and NGO workers to other community members, knowledgeable about the motivations of extension workers of different organizations (who are often their age-mates and former school-mates), and mobile enough to attend to a great number of meetings and workshops. They act as brokers between pastoral communities and national organizations. They have various sources of bargaining power that they use to advance their personal (and their community's) interests through the conservancy program. In stark contrast to other social fields (e.g. tenure rights, property rights in livestock) it is in this context that education and networking skills are of superior relevance than wealth in livestock. Apparently conservancy committees establish a new arena of local politics in management, and in a number of conservancies can channel substantial benefit flows. They decide how revenues accruing from contracts with the private sector or from other income-generating activities connected to the conservancy are distributed.
Do committee members profit financially from being engaged in the conservancy? Officially they are not allowed to receive any direct salaries. They may however receive allowances for attending meetings and undertaking trips e.g. to Windhoek for the conservancy. 15 Committee members enjoy other privileges however: they are well-informed about activities planned within their conservancy, and may directly profit from these. There are also a number of alleged fraud cases in which money has disappeared from conservancy accounts. It is likely that committee members are culpable in such cases, although quite often alleged fraud cases were in fact connected to poor book keeping. However, I found very little evidence for systematic elite capture, as De Vette et al. 2012 attest for conservancies in the Omusati region. First of all there is no clear evidence for the presence of an elite. 16 Wealthy and powerful people are tied in within kinship networks, and often compete with each other for influence. Instead of systematic enrichment, I found that influence is used to allocate benefits to relatives and friends in the form of small salaries.
Traditional authorities usually cannot be elected to conservancy committees. This stipulation was an intentional move by the government to prevent the continued prevalence of traditional authorities in matters pertaining to resource governance (Hohmann 2003). The previous paragraph has shown, however, how closely TAs are involved in the making and management of conservancies. TAs were instrumental in the initiation of conservancies some seeking legitimacy to their chieftaincy claims, some attempting to secure territory. In the day to day management of conservancies they matter too. Whenever, for example, regulations pertaining to core conservation zones are not adhered to and herders have to be convinced to move away from a protected area the TAs are coopted by the conservancy committee. TAs are also involved when land use and boundary issues are at stake or when private investors seek permission to operate in the area. A leasehold to a lodge, for example, can only be facilitated with the consent of the traditional authority, despite a conservancy committee's affirmative stand. Of course, this is not a one-way street: TAs also benefit directly from conservancies. In some cases a kind of fixed honorarium is paid directly to the chief. In other instances chiefs directly profit from game quotas: often a fixed rate is allotted to them and sometimes they get extra-bonus quota game. This game is thought to furnish gatherings at the chief's place with food. They are also an inroad for the influence of NGOs and ministries on communities. Both directly give advice to and instruct committees. In many ways committees are intermediaries of power between the local level, state organs, and global actors.
Conservancies hence display a hybrid governance structure. On the one hand conservancy committees adhere to the management structures set out in the conservancy legislation. Committees are oriented along principles of modern governance: democratic elections, accountability of elected representatives, and co-supervision of committee members. On the other hand they pave the way for the continued influence and dominance of traditional authorities, since committees depend on affirmation of their plans by the traditional authority.

Environmental governance through conservancies
How do conservancies govern social-ecological relations, and what fields of human-environment relations do committees address? How do they define the common-pool resource? What visions and practices are there to distribute the benefits accruing from this common-pool resource? Five major efforts are notable: (a) internal zonation facilitates communal land use management; (b) sizeable core-conservation areas are established, and currently also enforced; (c) the monitoring of wildlife and human-wildlife conflicts captures the dynamics of humanwildlife interaction, and contributes to a rapidly developing expert knowledge of game mobility and a growing intimacy with wildlife; (d) the setting of quotas and commercial hunting determines the population dynamics of game; (e) the tendering of land to outside investors transforms a resource (land) into communal income; both core conservation areas and terrains allotted exclusively to tourism contribute to significant land-use changes.
Let us first shed some light on the boundaries a conservancy has to establish in order to become gazetted. The territorial boundaries must be specified before the application for conservancy status is submitted to the MET. In most cases such boundary discussions between prospective conservancies take a long time. Drawn-out boundary disputes are typical features in the emergence of conservancies. While the colonial chiefly territories did not demand fixed and mappable boundaries, such boundaries are nowadays the precondition for the gazettement of a conservancy. In many instances chiefs dispute each other's territorial claims and boundaries. They usually do so with historical reasoning: whose ancestors have first settled here, who only came later. Often boundary disputes also erupt because of internal conflicts for chieftaincy. All along the eastern boundary of the Kaokoveld's boundaries another type of dispute comes into play: in a number of conservancies at the eastern rim of the Kaokoveld Oshiwambo-speaking farmers are settling; their access rights to ground are much disputed and usually they are denied membership in conservancies. However, the conservancies' boundaries often include their extensive fields within the boundaries themselves, leading to an impasse between conservancy boundary and membership. Day-to-day herding takes place mainly within these boundaries. In a survey of ninety-five households in twenty conservancies in 2012 I found that in normal rainfall years most of the grazing takes place within a conservancy. This is a strong hint that many if not most conservancies -while mandated to manage gamefirst of all define grazing territories, and by implication stipulate boundaries around these grazing territories. The boundaries of these grazing territories are still porous though: A third of the households surveyed in 2012 said that they would have access to pastures beyond the boundaries of their conservancy through kinship ties in times of need. Schnegg and Bollig (2016) showed that during the drought 2013/2014 quite a number of households crossed conservancy boundaries in their search for pasture. Events occurring Towards an Arid Eden? 787 in 2015 however hint at progressive exclusionary practices: Himba who had migrated south of Opuwo and had settled there with Herero communities had been denied access to conservancy membership. They were now petitioned to leave these areas as they had allegedly transgressed internal zonations of conservancies.
The zonation of a conservancy is an instrument to differentiate core-conservation zones from zones for touristic use, for commercial hunting, and for subsistence herding and/or farming (see Map 2). Such zonation planning usually results in digital management maps. Resource management planning takes place at the interface between committees, and NGO and GO staff, who offer advice on the topic. Committees do know that a rough categorization of the conservancy area is needed in order for it to be governmentally acknowledged -so a core conservation area and settlement areas should be singled out. In the past, core conservation areas were often placed at the margins of conservancies, where herding had been rare. These are often areas which are far away from roads and which do not have boreholes. As these areas have rarely been grazed they are of considerable importance as dry-season grazing reserves however. Often conservancies hope to attract a tourist campsite or even a lodge to such exclusive areas. Table 2 details the zonation plan for Ehirovipuka conservancy, and Map 2 shows the zonation plan for Omatendeka Conservancy. Four different types of tourism areas were established: in some of them the grazing of livestock is minimized or totally forbidden. Further, hunting areas and multi-use livestockfarming areas are designated. The kind of activity allowed in each zone is clearly spelled out. Table 3 shows how much space is devoted to a specific type of usage. It is significant that all four conservancy management plans analyzed here devoted substantial areas to wildlife and hunting, in many ways precluding substantial land-use change: about a third of the land is given to conservation, with the hope of reaping communal benefits later on.
In several conservancies I asked in 2012 whether there were still households staying regularly in core-conservation areas. In several instances there were still a few (or single) households, and it was very unclear whose responsibility it was to act to address the situation. The conservancy committee ostensibly had no right to punish wrongdoers or to expel households from the core conservation zone. The general idea in 2012 was that the committee would have to report people trespassing over zonation boundaries, and especially those staying in core conservation areas, to the traditional authorities, who (in theory at least) had consented to these zonation boundaries when the conservancy was established. The traditional authorities, however, were often not in general agreement with the zonation, or seriously disagreed amongst themselves on the placement of zone boundaries. In interviews conducted in 2012 I also found that the zones' boundaries were not well known among herders. While knowledge of zonation boundaries was somewhat unequally distributed, with a good knowledge of such boundaries e.g. in Ehirovipuka (where such boundaries were physically marked) and less well developed knowledge in Okangundumba (and most other conservancies), generally the institution of zonation was only incipient.
In 2015 conflicts around core conservation zones had escalated. A number of conservancies had convinced traditional authorities to engage a lawyer in Windhoek to act on their behalf and to seek eviction orders with formal courts to push out homesteads from core conservation areas. It was rumored that in one case a foreign investor had insisted that he would only invest in a lodge within the core conservation area if that area was actually devoid of people. In an interview the lawyer conceded that eviction would be difficult, as the people in question were often locals who had long made use of those areas that were later designated core conservation areas. Only if it could be proven without doubt that all members of the conservancy and the traditional authority had consented to the complete evacuation of the core conservation area was there a chance to evict people.
The core conservation areas are probably the clearest spatial expression of the new commons. The rumors about the investor who would not spend money unless  individual pastoralists had withdrawn from the area clearly shows how closely these new commons are linked to global inputs and external capital investment. In many ways these new commons prepare the way for a commoditization of land without privatizing it and without fragmenting it into freehold farms. 17 Zonation and core conservation areas are discussed in a number of meetings. Usually a conservancy committee prepares an initial idea about such zones (typically in conjunction with external advisors) and has its ideas then discussed in a larger meeting of conservancy members. Attendance of such meetings is highly irregular however, and many inhabitants of a conservancy apparently had never heard about the existence of such boundaries before they were gazetted.

General income situation of conservancies
Commons need to be economically successful in order to be resilient. Unlike resources held in private ownership they must not only produce sustainably, but must also facilitate the just distribution of benefits. Agrawal (2001Agrawal ( , 1661 sees unpredictable benefit flows and unfair allocation as having adverse effects on the durability of institutions of common-pool resource management. The allocation of specific communal rights in game and land to conservancies by government decree requires private business partners to turn these assets into a benefit. I will first of all address the question of income in general before touching upon intracommunity benefit flows. The total income of Namibian conservancies increased from N$600,000 (c. 55,000€ in 2010) in 1998 to N$39.5 million (c. 3.5 Mio €) in 2010 (NACSO 2012, 4). In 2013 the total income was up to N$68 million (6.2 Mio €) (NACSO 2014). The income of single conservancies differed greatly. Suich (2009, 18) reported that in 2006, 32 conservancies earned cash income ranging from a meager N$7200 (Kunene River) up to a high of N$927,950 by Torra conservancy, and that while 18 (out of 50) conservancies did not receive any income at all, some 13 conservancies covered their operating costs through revenue generated by conservancy income. In 2010 some 45 out of 59 gazetted conservancies had some kind of cash income, and some 23 conservancies operated independently of donor support, i.e. paying salaries and operations from their income. In 2013, 65 out of 80 conservancies generated returns, 36 covered their operational costs from their own income, and 38 distributed cash or in-kind benefits to members or were able to invest into community projects (NACSO 2014, 31). These are certainly impressive figures. However, they say little about how incomes are distributed in the community and whether such incomes contribute e.g. to poverty alleviation. Where does conservancy income (i.e. the return on common-pool resources) come from? Table 4 gives an overview of various sources of income in Namibian conservancies for the years 2007 and 2010. Table 4 shows that about half of the income in cash and in kind is generated from joint ventures in the tourism sector (i.e. mainly rents from lodges, and wages paid to conservancy members in such enterprises). A substantial amount of income is earned through trophy hunting; in 2010 some 28.2% resulted from this activity and in 2013 some 31% resulted from trophy hunting. With the ban on trophy hunting in Botswana these figures are likely to grow. In 2010 trophy hunting generated some N$13.9 million (1.3 Mio €). Out of this 80% were cash payments (N$11.1 million), and 20% resulted from game meat distributions. 18 Conservancies in north-western Namibia "produce" two commodities jointly: game, and wilderness areas. They need external operators, commercial hunters and clients to turn both these resources into income -to transform wildlife into huntable game, and landscapes into marketable wildernesses. Both commodities are peculiar in many ways. Conservancies receive annual game quotas. These are set in annual meetings in which conservancy members, officers of the MET, NGO staff, and also trophy-hunting companies participate. 19 About 20% of the quota is designated for trophy hunting, whereas 80% is kept for own-use hunting (see Table 5). The latter category consists of animals assigned to traditional authori-18 Trophy hunters, or more often their helpers, usually only cut off the "trophy part" of the animal that has been shot. The meat is left with the community for distribution. 19 Since 2015 the annual quota setting has been transformed into a three-yearly quota setting. ties to furnish meetings with meat, animals traded in shoot-and-sell contracts to butchers from the wider region, and animals exchanged with local agencies for their services. Bollig and Olwage (2016) report that animals from this part of the quota were also given to the local police in payment for some services, and to road contractors for extending a road.
For those animals assigned for trophy hunting and for shoot-and-sell hunting a buyer has to be found. While in theory trophy-hunting quotas should be publicly advertised, in practice conservancies are directly approached by trophy hunters. The exact contract between hunter and conservancy is negotiated between a number of stakeholders. The trophy-hunting company may only buy part of the designated trophy-hunting quota. They usually then guarantee a fixed number of game animals which will be hunted, and an optional number which will be paid for per individual animal hunted. The designated quota is paid in several instalments over the course of the year, directly to the conservancy account. A number of conservancies complained that they had not managed to attract a trophy-hunting company yet, i.e. they had failed to reap any benefits from their commons. Indeed, figures presented in the NACSO Annual Report for 2013 (NACSO 2014) show that only some conservancies in northern Kunene had contracts with trophy-hunting companies. In an interview the director of one such company remarked that trophyhunting quotas in some conservancies are too small, or simply have the wrong set of animals on offer to make it worthwhile to contract them (see Table 5).
Much of the remainder of the quota is given to shoot-and-sell contracts. Here local butchers come and buy large numbers of game from the quota. They drive into the area with a cooler truck and then often shoot large numbers of animals. Due to the Red Line regulations (Miescher 2013) they are not allowed to export game meat south of the Red Line. Hence, lucrative game-meat markets in Namibia's centre are not accessible to them. The main market for game meat from Kunene is the rapidly growing urban area of Oshakati/Ondangwa, where an urban middle class distinguishes itself through the consumption of game meat. There is a conflict of interest between trophy hunters and shoot-and-sell hunters. The trophy-hunting companies allege that shoot-and-sell hunting is taking place at times and in places where their clients are not present. Bollig and Olwage (2016) show that trophy hunting and shoot-and-sell hunting have gained increasing relevance for the income of conservancies. Communal property (i.e. quota game) is commoditized, and via contractual agreement with commercial hunters directly transferred into private property of hunting companies. Figures suggest that the potential of this transfer is as yet not fully made use of: by no means is all game put on the quota saleable. On average only about 20% of the value ascribed to hunting quotas is actually cashed in (see Table 5), and a number of conservancies are only able to sell less than 10% of the quotas allotted to them.
A major part of conservancy incomes is invested in salaries and the general upkeep of operations of a conservancy. A NACSO report on the year 2010 establishes (NACSO 2012, 25) that conservancies all over Namibia covered the majority of the costs of 619 conservancy management jobs from income gen- erated through conservancy-related activities. In Kunene North, in twenty-eight conservancies, some 100-150 people were employed. Unfortunately I do not have exact figures on those employed: a majority of them, however, are male and aged between thirty and fifty. In interviews with committee members in some ten conservancies we found a tendency to employ more staff when income generated by the conservancy increased. Typically game guards are employed first. In the past, if a conservancy had not generated sufficient funds, the IRDNC stepped in and supplied limited funds to pay game guards during an initial phase. These payments were then phased out once a conservancy proves to be financially viable. Most conservancies had at least four game guards. These guards regularly patrol the area, making entries in event books on game sightings and human-animal conflicts. They are also meant to report on e.g. poaching cases or allegations of poaching; they report to the committee. Additional staff positions are created when the income of a conservancy increases: usually a program officer, a field officer (coordinating the activities of game guards), and a financial administrator are employed. The salaries of these positions varied. Usually the program officer was the best-paid position (with some N$2500-3000 per month, 230-270€ 20 ); game guards were found to earn between N$500 and 1000 per month (46-91€). These salaries are very moderate according to Namibian standards, and the salaries for game guards were usually below the minimum wage fixed by law (N$722 in the agricultural sector in 2014, SACAU 2014).

Conclusion: the new commons, conservation, and globalization
The legal reforms of the 1990s established a new form of commons. While game had been the property of the state for nearly a century and had been anxiously guarded by state officials, now specific, official rights to game and land were transferred to mandated communities, the so-called conservancies. In conjunction with NGOs the government established procedures stipulating how such conservancies had to be organized (clearly defined membership, elected committees, management plan, spatial boundaries). It is certainly important to note that only management and transfer rights to game and land were devolved. Communities and administrative staff together fixed a quota for each huntable game species, and partitioned the quota into valuable trophy-hunting game, and less valuable "own-use" game. While the conservancy was to manage all the game in its territory, only the specified quota was to be used. With respect to land, the conservancy was permitted to transfer use rights in the form of land rentals to private investors. Contracting between conservancies and private investors was guarded and monitored by state officials. In many ways, the newly emerging common-pool resource is different from traditional commons. The new commons are complex judicial entities, are partial, and their use is monitored by the state. In this way, however, they are no different e.g. from Acheson's lobster fisheries in Maine: here too resource use is defined conjointly between the community of users and an administrative body (Acheson 2006). Despite these restrictions, conservancies have spread rapidly in north-western Namibia. Obviously they do not only play a role in defining and facilitating the use of a (new) common-pool resource; their innate structures (membership, boundaries) can also be used for local political purposes: they provide a mechanism for territorializing political ambitions and for defining inclusion and exclusion. Meant as a vehicle for nature conservation, they are used as a tool to legitimize the emergence of village territories under the joint guidance of an elected committee and traditional authorities. Committee members and employed staff as well as traditional authorities usually profit personally from conservancies. Whenever profits are made they are in a primary position to gain from benefit distribution. They are also used to organize land use. Nambia's new commons of game management open up communal lands for an international commodity market. Land can be rented under specified conditions, and game may be hunted by commercial trophy-hunting companies. Often those buying the commodities produced by these new commons are wealthy Westerners (as hunters) or globally operating tourism entrepreneurs. Price tags are attached to entities which were previously non-negotiable and did not have a market. I have shown that this commoditization process is ongoing and incomplete. While the income in some conservancies is sizeable, in many conservancies there is as yet little income from the new commons. The distribution of benefits from these new commons is still problematic, and a point of concern not only for those planning and facilitating conservancies but also for local activists. How can income meaningfully be distributed, what is a just and equitable mode of benefit dispersal, and how should costs accruing from the rising number of game animals and the establishment of core conservation areas as no-go zones be handled? These are open questions that are eagerly discussed in the local context.
To what extent have new communal institutions been developed, and what do these new commons look like? This contribution has detailed three salient processes linked to the establishment of the new commons. The emergent institutions of conservancies are deeply intertwined with existing institutions and organizational patterns of decision-making regarding access to and management of natural resources. Traditional authorities have considerable influence on the development of conservancy programs. They are also entangled with NGO and donor facilitated processes and government agendas. While these entanglements with other scales, institutions and strategies are strong, I have argued that conservancies do create a new platform for young, educated males and females to further their own ends and the needs of their communities (however they may perceive them). I suggest that the argument that "CBNRM is thereby clearly positioned as a state-, NGO-and donor-facilitated process of outsourcing access to significant public natural/wildlife resources and potential income streams to private-sector (frequently foreign) business interests" is perhaps somewhat simplistic. Communities do regain some control over natural resources which they had lost in colonial times -even if this control is transient and not easy to handle. It is certainly true that the rights to game and land re-communalized through the conservancy program only result in sizeable benefits if marketed. This creates new dependencies and new challenges, but also new potentials. The Kaokoveld comes from a colonial period in which the interaction of local farmers with wider