1. Introduction

Many studies of the local institutional dynamics in common-pool resource governance focus on the agency and organising practices of community members. That is rightly so. However, the governance of such natural resources often takes place through cross-level interactions between multiple actors and different forms of organisation (Mwangi and Wardell 2012). This includes interactions between communities and meso-level organisations such as local government and local branches of the central state. Actors in such organisations typically interact regularly with local communities, but also reach horizontally to other localities and vertically to other scales. Understanding their agency is therefore important if we are to grasp the everyday dynamics of how water, forests, land, wildlife and other natural resources are governed on the ground.

In this article we focus on a particular group of such actors, whose everyday agency has only recently gained attention in development studies – namely the technical staff who work in the government departments of agriculture, water, forestry, environment etc. at the local level (Blundo and Glasman 2013). We are talking here about the public servants who are based in the decentralized (often deconcentrated) departments of central government ministries at district, county or similar levels and who thereby represent the “local state” (Olivier de Sardan 2014). Charged with implementing and enforcing national policies and laws on the ground, these “street level bureaucrats” (Lipsky 2010) potentially have significant influence on the options and limitations for community management of collective resources. Even where policies and laws promote inclusive and community-led natural resource governance, the local technical officers1 tend to remain key actors whenever communities require the collaboration, approval or goodwill of the state.

Because of their positions in the state structure, it is tempting to see such technical staff as mere extensions of the central state. From such a viewpoint, they may at best seem of limited analytical interest, and at worst as the embodiment of autocratic state-led approaches. However, the local institutional context of natural resource governance is often far more dynamic than a dichotomous “state-versus-community” approach would imply (Blundo and Le Meur 2009). In this article we seek a more nuanced understanding of the agency of local state actors in environmental governance, and how they go about implementing and enforcing statutory frameworks on the ground.

We do so through a case study of how local Environment Officers (EOs) have implemented the national environmental law in a rural county in southern Kenya. We find that their role is far from straightforward and involves making their own discrete interpretations of rules and procedures, and forming strategic alliances and informal agreements with communities on the use and management of natural resources. As a result, the environmental law is to a large extent implemented through a blending of formal and informal mechanisms and practices, thereby showing that not only communities but also local state actors are active bricoleurs (Douglas 1987; Cleaver 2012) in everyday natural resource governance.

Our aim with the article is thus two-fold: Firstly, we hope to cast more light on the role of the “local state” in the everyday governance of common-pool resources, with a particular emphasis on how front-line bureaucrats engage with communities in the implementation and enforcement of national policies and laws. Secondly, we seek to contribute empirical evidence to critical institutionalism in studies of natural resource governance (Cleaver 2012), by illustrating how bricolage can form an important part of the everyday practices whereby the “local state” seeks to reproduce its legitimacy and authority. In such situations, processes of bricolage are best understood as the product of interactions between and across the domains of state and community, rather than something “outside” the state and exclusive to communities.

2. Methods

The study is based on fieldwork conducted in Taita Taveta County in 2005, 2012 and 2013, thereby spanning most of the period from 2004 to present that Environmental Officers (EOs) have been active in the area. A total of 47 individual, qualitative interviews were conducted with (i) three of the four Environment Officers who have been active in the area over the period, (ii) other state employees working in agriculture, water, forestry, wildlife and community development, and (iii) other local actors including Councillors, members of Community Based Organisations (CBOs) and government chiefs who form a local level of the national administration. Focus group interviews with farmers and fishermen in the area complemented the qualitative interviews.

The government officers interviewed in Taita Taveta were the Heads and Deputy Heads of their local line departments, as well as their extension staff. Initial interviews focused on getting an overview of how they spent their days, what their work consisted of in practice, who they interacted with, etc. On this basis, particular events and situations were selected for more in-depth exploration through re-interviews and interviews with other involved actors (Mitchell 1983). Outside of Taita Taveta, a further 11 interviews were conducted with senior staff at various levels in the Ministry of Environment and the National Environmental Management Authority (NEMA), providing a backdrop to the organisational culture of which the EOs were part.

3. The local state and environmental management in Kenya2

From the late 1890s onwards, colonial governance in Kenya evolved around the well documented mechanism of a Provincial Administration and indirect rule (Berman 1990). At the head of the subnational structure was a handful of Provincial Commissioners who governed large areas through colonial District Commissioners and their technical officers. Everyday law and order was maintained through local Chiefs who were either selected on the basis of their authority and influence, or “invented” where none existed. District boundaries were drawn by the colonial administration and in the mid-1920s Native Councils and associated customary courts were established. The subnational colonial administration was based on a degree of semi-autonomous discretion, and Provincial and District Commissioners were allowed a good deal of liberty in everyday decision-making (Berman 1990; Anderson 2002). Moreover, the mandates of the Native Councils and courts tended to be relatively fluid, and served as arenas for local contestations over how authorities, rules, and land rights should be interpreted (Berry 1992). The exact nature of the formal local institutional environment was therefore not cast in stone, and ad hoc adapted arrangements were frequent.

As the colonial government continued to compress humans and livestock into African Reserves, concerns over land degradation developed in the colonial administration. The result was a conservation discourse, claiming that intervention was needed to enhance African agriculture through measures such as soil conservation and more effective grazing (Anderson 1984; Rocheleau et al. 1995). Conveniently, this also legitimized restrictions on African agriculture in favour of settler production (Mackenzie 2000). Accordingly, it became part of the Provincial and District administration’s tasks to promote conservation and what would today be called “sustainable” farming and land use in communities. However, for the colonial District Officer the task was often a balancing act between conflicting mandates: They had to further the interests of the central administration and settlers, while at the same time placating African communities discontent with the loss of farming land, pasture, water and other resources (Berry 1992). A situation which, in some respects at least, forewarned the situation faced by today’s government field officers in Kenya.

Following independence in 1963, President Kenyatta maintained rural control through a combination of clientilistic structures and a continuation of the colonial administrative system. In 1983 President Moi initiated the so-called “District Focus” decentralisation policy, designed to curb the increasing power of Provincial Commissioners and MPs, and build the new regime’s own support structures in the countryside (Barkan and Chege 1989). While largely maintaining the existing administrative framework, the new policy moved the emphasis of state support from provincial to district level (GoK 2008). The district level thus gained greater importance as a locus for planning and implementation, and as an arena for local actor’s access to funds and influence. Yet most districts were understaffed compared to their mandates, and in decentralisation terms the structure was at best one of deconcentration, but by no means devolution (Devas and Grant 2003).

Until the late 1990s, state authority in district level natural resources management rested only with the traditional sector departments such as agriculture, livestock, forest, water, wildlife. Then, in 1999 the cross-cutting Environmental Management and Coordination Act (the Environmental Act) was introduced (GoK 1999). Prepared by a group of activist lawyers and helped through parliament by donor conditionalities, the act was relatively progressive. It stipulated citizen rights in environmental management, and established an institutional framework based on inclusive approaches. At the national level, this included a multi-stakeholder Environmental Council, a Public Complaints Committee and a National Environmental Tribunal, to whom citizens could complain over environmental grievances. At the local level, District (later County) Environment Committees were established as the main vehicles for environmental planning and management, with a strong representation from the public and civil society.

The Environmental Act formed a new state organisation to coordinate and supervise its implementation, namely the National Environmental Management Agency (NEMA), with offices at national, provincial and district (later county) level. The latter were to be manned by local Environment Officers (EOs), to be stationed at district level. From the outset, their mandate was broad. On the one hand, the act provided them with a role as supervisors and coordinators of inclusive environmental planning processes, and ensuring public rights in environmental management. This included somewhat ambiguous tasks such as protecting “indigenous property rights of local communities in respect of biological diversity” (GoK 1999). On the other hand, the act also provided them with the responsibility to enforce environmental protection across multiple sectors such as agriculture, forestry, water and urban and industrial development.

The provisions in the act were further compounded by a new Constitution adopted in Kenya in 2010, which emphasized the principles of benefit sharing and inclusive approaches in environmental management (Mwenda and Kibutu 2012). The new Constitution also provided for a substantial restructuring of Local Government: The old District Councils were scrapped in favour of a new system of County Councils and elected Governors, and a greater degree of fiscal devolution.3 Consequently, NEMA’s local EOs, charged with a somewhat broad and ambiguous mandate, had to work in a changing institutional environment. In the following, we describe how they have gone about their task in one particular County, namely Taita Taveta in southern Kenya.

4. Taita Taveta County

Taita Taveta County is located in south-eastern Kenya between the border with Tanzania and the Nairobi-Mombasa highway. This largely rural county covers approximately 17,000 km2 and is home to some 285,000 people. The county is mostly known for the Taita Hills, an arable mountainous and densely settled area where the Taita people engage in small-scale agricultural production. However, the majority of the county (89%) is designated arid or semi-arid land, distributed along the foothills of Kilimanjaro and in low-lying rangelands inhabited by Tavetas, Masaai and other groups. A variety of land uses and livelihoods are found in the area. Subsistence production of maize, beans, and dairy dominates in the Taita Hills while livestock range production and horticulture through irrigation dominate the lowlands. Lake Jipe in the southernmost part of the County is an important resource for small-scale fishermen. The county also hosts several large-scale agricultural estates producing sisal, dairy, beef and fruits. Some of these are owned by highly influential families in the Kenyan political and economic landscape, who obtained the land in the years following independence (Onoma 2010). A significant proportion of the county is furthermore designated as national parks, wildlife sanctuaries and forest reserves, including the sprawling Tsavo National park. Lastly, there is a vibrant (mainly artisanal) mining industry for gemstones and other mineral resources, including expanding sand mining activities.

Taita Taveta is thus in many ways a microcosm of Kenya’s competing land uses, and the colonial history of the county remains imprinted on its map: The protected areas occupy a full 62% of the county’s land area, most of which is formally state land. The privately owned large estates account for 24%, leaving just 14% of the land area for the rural population (Njogu and Dietz 2006). The latter consists of a mosaic of tenure forms: Farmer’s individual plots are typically inherited, and while some households do have title deeds they are far from widespread. Pastoralism is in some parts of the county based on group ranches, although in other areas pastoralists are leasing out their land to in-migrant small-scale cultivators. Access to farming land and water is thus a significant point of contention in the county (Waswa et al. 2002). Meanwhile the previously poor road infrastructure is being improved, leading new interests into the County, including small-scale miners of sand and gemstones.

5. The World of the Environment Officer

It was into this context of dynamic and competitive resource use that the first NEMA Environment Officer arrived in 2004.4 He worked alone, and was later followed by two others in succession, also working alone. It was not until 2013 that a second position was opened in the area as a result of the constitutional reform, thereby creating a team of two EOs. During the period studied, a total of four EOs (three men and one woman) have thus been posted in Taita Taveta, each serving 2–4 years before being rotated to a new area. The EOs have been at different stages in their career – one was in the first posting, another in the second posting, and two were mid-career with a significant experience as government employees. Their style and personality traits have differed, some being more outgoing and active than others in their engagement with other actors. Nevertheless, in overall terms the practices described here were common among them, and were passed on between them.

On arriving in Taita Taveta, the EOs had to position themselves in a local institutional landscape that was rather more complex than their employment contract suggested. The formal mandates of EOs are shaped partly by the Environmental Act, and partly by the prevailing policies and approaches at central level. Their main tasks include: (i) screening development plans and projects to ensure they are not in breach with the Environmental Act, (ii) facilitating Environmental Impact Assessments, (iii) monitoring and enforcing the Environmental Act at community level, (iv) supporting the work of the Environment Committee in environmental planning and decision-making, (v) providing guidance and awareness to communities on environmental management through participatory approaches, and (vi) serving as a contact point for public complaints and grievances under the Environmental Act. To carry out these tasks, EOs have for the most been alone (although latterly with a colleague) in charge of an area covering 17,000 km2, with poor roads, many remote communities and considerable variety in land uses, ecosystems and environmental issues. Initially, the EO had no transport and had to hitch a ride with other government officers when going to the field. At the time of writing, a vehicle has been provided, but the task of covering the entire county remains a major challenge.

Formally the EO reports to NEMA’s headquarters in Nairobi, but these are 5–6 hours’ drive away, and even with a mobile phone most decisions on the ground are left to their own discretion. This includes how to actually interpret the Environmental Act they are supposed to implement and enforce. Few specific regulations exist and as the act itself is quite broad and of a cross-sectorial nature, its scope and practical implications is in many cases open to interpretation. This is even more pronounced when it comes to how policy directions from NEMA headquarters shall be translated into practice. For example, in training courses EOs have increasingly been instructed to apply “a combination of stick and carrot” – but exactly what this implies on the ground is not clear. In terms of fulfilling her/his formal role on an everyday basis, much is therefore left to the EOs own discretion and interpretation of the law, and to her/his ability to find pragmatic solutions to the practical problem of limited budgets, time and operational scope.

Apart from the formal mandates and organisational framework, EOs also have to position themselves vis-à-vis the “informal” structures and dynamics of the state (Olivier de Sardan 2014). In Taita Taveta as elsewhere, underlying turf wars and power relations within the state apparatus play a major role in the everyday working life of technical officers. The issue is particularly pronounced for EOs because of the cross-cutting nature of the environmental law and the mandate of NEMA. For example, the law provides NEMA and its officers with certain responsibilities in terms of issuing environmental licenses, but these are in conflict with licensing authorities provided to state water officers. Similar conflicting mandates exist in relation to the Kenya Forestry Service and other agencies. In the ensuing everyday struggles over authority and work domains, EOs are relatively poorly positioned because their mother agency lacks the clout and history of the more productive sectors, such as agriculture, forestry and mining. In navigating these and other informal features of the state, EOs typically have to juggle their relationship to NEMA headquarters in Nairobi with power relations in the local institutional context. The former is critical in terms of personal career strategies and continued employment, and meeting expectations from superior officers at HQ. At the same time, however, fulfilling these expectations is rarely possible without some form of local powerbase and network – not only within the state structure, but also outside it.

In Taita Taveta as elsewhere, the local institutional context is continuously changing. In the past 20 years, the number of Community Based Organisations (CBOs) has grown substantially, and at the time of writing there are more than 52 registered CBOs in Taita Taveta working on natural resource and environment issues. This is partly a result of support from aid agencies and conservation NGOs to the area, which has had a strong emphasis on community based approaches in agriculture, water, forestry and wildlife management. At the same time, local state agencies such as the Department of Gender and Social Development have promoted CBOs as modalities for service delivery and funding channels to community activities. Meanwhile CBOs serve a variety of purposes for community level actors themselves: While some are merely platforms for access to funding, others are driven by broader grievances over rights to land and/or developmental concerns over a degrading natural resource base.

In their relationship with CBOs, EOs have a strong position as law enforcers, and as gate-keepers to the state and associated funding. However, they are also expected to facilitate inclusive and incentives-driven approaches to environmental management, and to live up to the state rhetoric on civil servants as public service providers. With reference to this, CBOs in the area continuously put pressure on EOs to address their interests and grievances, and actively engage EOs as part of forum-shopping strategies (von Benda-Beckmann 1981). For example, CBOs and individual community members often approach EOs in Taita Taveta with grievances over forest access rights, while at the same time pursuing the same issue with staff from the Kenya Forest Service and Local Government Councillors.

At the same time, many CBOs in the area are increasingly seeking to assert their authority as managers of the environment in their own right. A number of CBOs are nevertheless engaged in (often donor-funded) pilot projects as managers of water, forest and wildlife resources. In these situations, CBOs are typically eager to present themselves as managers on behalf of “the citizens of Taita Taveta” or “the Kenyan people”, thereby claiming a degree of “public authority” in natural resource governance (Lund 2006). Local Governments, too, provide an important factor in the changing local institutional context. For EOs and other technical officers, the reform of Local Government under the new constitution has led to a necessary reorientation towards the previously very weak Local Government structures: They now have to work with Local Governments – and in particular the powerful new Governors – and ensure a good relationship with them, while simultaneously maintaining their independent authority as representatives of the central state.

Like many other local technical officers in rural Africa, EOs in Taita Taveta are thus confronted with a context of increasing institutional multiplicity, whereby a growing number of different types of organisations perform functions and claim authority within natural resource management (Ribot 2007). On the one hand, this provides certain opportunities for EOs and other technical officers, but on the other hand it also potentially challenges their role as local authorities in the management of public goods. The role of EOs in Taita Taveta is thus everything but straightforward: They must implement a national law and associated policies as both law enforcers and facilitators of inclusive approaches, and they must do so single-handedly with very limited means across a large area with multiple and competing forms of land use. At the same time, they must navigate in a local institutional landscape characterized by informal competition and power struggles within the state itself, and in a broader institutional environment characterized by growing pressure and expectations from CBOs, and challenges to their authority from both civil society and local government.

How have EOs in Taita Taveta approached this? In the following we discuss three key features of their practices, namely (i) working through personal networks, (ii) tailoring informal agreements, and (iii) delegating public functions and authority to civil society. In order to illustrate this, we first discuss how the EOs engage with other local actors in the local Environment Committee, and then describe two particular events that illustrate their practices and modes of engagement with communities.5

6. Working through personal networks

A key element in the practices of EOs in Taita Taveta has been to establish and work through informal relationships and networks with other stakeholders. This is illustrated by the way EOs have acted in relation to the local Environmental Committee, which is formally the main forum for planning and decision-making on environmental management at the County level (NEMA 2012). The Environmental Committee formally consist of representatives from the local line agencies, but also of four representatives of “farmers, women, youth and pastoralists” (GoK 1999), two representatives from the local business community, two from CBOs and two from NGOs. In principle the public are thus quite strongly represented in the Environmental Committee. EOs are formally tasked with facilitating and advising the committee, but over the years the committee’s meetings became a bit of a headache for both the EOs and the participating members. Apart from being costly affairs that exceeded the allocated budget, meetings in the committee were prone to conflict and heated debate among the participants: Some of the “public” representatives used the meetings to exert pressure on the government staff, complaining over insufficient service delivery and claiming increased community control over/access to forests, land and wildlife. At the same time, some of the technical officers (including forestry) were wary of collaborating on the environmental plans, which they saw as a potential threat to their authority. In consequence, technical officers from the line agencies began avoiding the meetings, with the result that the Environmental Committee lost what limited status it had in the local planning hierarchy.

In response to this, the EO and some of the members of the Environmental Committee began instead to liaise informally with one another outside the formal meetings. This included one of the CBO representatives, the farmer’s and women’s representative, as well as the Development Officer and the Mining and Agriculture officers. Common among this group was that they considered one another pragmatic and “constructive” to work with. By contrast, one CBO representative, as well as the pastoralist representative and other technical officers were not included in this informal network. They were considered un-cooperative and “difficult”. Over time, this informal liaison developed into a routine: The EO would look up the informal link members of the Environmental Committee individually at their homes or elsewhere to discuss issues that needed resolving under the committee, and gain their opinion. If there was disagreement, the EO would act as mediator until consent was obtained from those involved. Although the EO took the initiative for this practice, it was actively co-produced by the involved community members and technical officers. Indeed, the CBO and farmer’s and women’s representatives began coming directly to the EOs office on their own account, to bring up issues that would otherwise have been discussed in the Environment Committee. Alongside these informal relationships the Environmental Committee continued to meet formally, but now with a parallel “shadow” structure of relations that were able to dominate proceedings.

Environmental planning and management in the Environmental Committee is thus to a large extent executed through informal networks between the EOs, individual citizens and other state actors. This approach is also reflected in the EOs’ engagement in other fora. For example, when EOs are concerned about the environmental consequences of new development plans in the County, they prefer engaging informally with the responsible planning officer to find a solution, rather than addressing the matter in the County Planning Committee which is formally the appropriate forum. Significantly, this does not mean that the formal organisational arrangements are irrelevant, as it is through these that the functions and authority of EOs and Environmental Committee members are legitimized in the first place. Rather, formal and informal procedures and modes of organisation are interlinked: Without the environmental law, local stakeholders would have no mandate to be present on the Environment Committee, and the EOs would have no power to take part in the County Planning Committee and screen local development plans for environmental issues. Yet it is the informal networks and procedures that make these organisational arrangements “work” in the everyday reality.

7. Tailoring informal agreements

A second key element in the EOs approaches to carrying out their mandate in Taita Taveta has been to establish agreements with communities and other stakeholders on rules and practices for resource use.6 Many of these agreements are not written down or formally sanctioned by law or higher authorities, and yet they play an important aspect in the EO’s practical implementation of his mandate.

One example of this is the rules relating to sand mining in Taita Taveta. During the 1990s and 2000s, small-scale sand mining grew as a new economic activity in the area, as it did in Kenya more widely. It was carried out by small-scale contractors from the coast, who moved in on an ad hoc basis with pick-up vehicles and dug sand from open pits or rivers, for sale to the construction industry in Mombasa. At the time no formal license was required for sand mining, and as such the contractors extracted the sand for free. If done excessively or in sensitive areas, sand mining leads to erosion and siltation of rivers, and may hamper local water supplies such as from shallow wells. This – and the loss of potential revenues to outside contractors – led local CBOs to complain to the EO that sand mining by outsiders should be stopped.

As a relatively new activity in Kenya, sand mining fell “between the gaps” in formal law, and it was thereby also unclear who had authority to address it. However, as pressure from the CBOs mounted the EO decided to take action. He liaised with the informal network of community members and other technical officers within the Environment Committee, and between them it was agreed that sand mining should be banned for non-residents, while in other areas it would be subject to approval by the EO. The community representatives in the Environment Committee further proposed that local sand mining cooperatives should be established, thereby allowing local communities to access and benefit from this new economic activity.

The members held meetings with communities and CBOs to introduce the plan, and sand mining cooperatives were duly formed. On their own account, the cooperatives quickly took on a role as local monitors of sand mining in the area: When community members encountered external contractors digging sand, they informed them that this was now illegal, and that the sand was managed by communities under national law. Contractors who refused to abide were referred to the EO or members of the Environmental Committee, who confirmed the rule and referred to the Environmental Act.

In this way, the informal ban on sand mining was in practice backed by the formal authority vested in the Act, the EO and the Environmental Committee. As small-scale entrepreneurs with no formal sand-mining license to defend, the sand mining contractors had little choice but to relocate their operations elsewhere. In reality however, there was no clear backing in national law for the ban on external contractors at this time, and although some EC members referred to it as a “bylaw” it had no official status as such. Indeed, the ban was initially never stated in writing: It rested only on verbal agreement and local consensus among everyone involved, a joint product of mutual local interests vis-à-vis outsiders.

For the sand mining cooperatives and CBOs, the ban provided an additional income and an opportunity to claim public authority as natural resource managers, while for the EO it reduced pressure from CBOs, and extended his mandate into a legally “unoccupied” domain. This situation lasted 3 years, after which the rules around sand mining gradually became formalized in writing. This was partly facilitated by local government who were looking to capture revenues from the local cooperatives, and partly a result of the new national Mining Bill of 2013, which addressed sand mining and provided for community schemes such as those in Taita Taveta. The arrangement thus now has formal legal backing, but the nature of the ban and the sand mining cooperatives continued to be the same as when they were informally developed.

The case of sand mining illustrates how EOs and other actors in Taita Taveta produce self-tailored “unwritten” agreements that rest on ad hoc interpretations of the Environmental Act – but which nonetheless have local significance, and obtain a degree of durability (and in this case eventual formalisation) as the involved actors seek to develop and strengthen them to their advantage. In the following we discuss another such agreement, which illustrates a further aspect of the EO’s practices.

8. Delegating public functions and authority

A third main element in the EOs’ implementation practices is the delegation of public functions and authority to civil society. One example of this is the conflict between fishermen and farmers around Lake Jipe. This substantial lake straddles the border between Kenya and Tanzania in the southern part of Taita Taveta, and is an important source of livelihood for local fishermen. Encouraged by local government, small-scale farmers in search of arable land are moving increasingly close to the lake shores, and some are drawing water from the wetland and its feeder river for irrigation purposes. This has led to concerns among local fishermen, who fear that the lake is being drained and that fish stocks are being polluted by the use of agrochemicals by some farmers.

Throughout the late 2000s, fishermen in the areas complained repeatedly about this situation to the local fisheries and agricultural officers, as well as to the EO and members of the Environment Committee. In 2009, the situation escalated further and CBOs and EC members put pressure on the EO to address the issue. An overriding concern for the EO and other technical officers was the need to prevent the situation from developing into open conflict. This would raise attention and might require higher-ranking tiers of government to be called in from Mombasa or Nairobi. During our interviews, technical officers stated that this would have annoyed their superiors, and would imply that they were not doing their job. As the fishermen appeared ready to pursue their case regardless of the consequences, the technical officers were inclined to accommodate them as far as possible. This had to be done cautiously, however, partly to avoid the local farmer’s CBO turning against them, and partly because the local government was involved in promoting the settlement of farmers near the lake. With local government gaining greater influence over local decision-making and budgets under the new constitution, it would not be wise to take a head-on confrontation on already sensitive issues of land and water.

The approach applied by the EO and other technical officers was therefore to contain the conflict as much as possible. This was done through a pragmatic approach that avoided any direct use of force or other forms of explicit assertion of the officers’ formal mandates as law enforcers. Instead, they took on a facilitating role, seeking to “coach people in the right direction” as one officer put it during our interviews. The officers asked EC members to meet the Community Fisheries Cooperatives and farmer’s CBOs, and enlisted the local Chief as mediator. On this basis an area was negotiated where it was “acceptable” to farm, and another where it was “not acceptable”. In the latter area, some of the offending irrigation channels were closed. Agricultural extension staff furthermore began encouraging farmers to settle away from the lake, in an area where they had secured funds for small-scale water development schemes. As local government also offered land title deeds in these areas, this “deal” was not entirely unattractive for farmers.

These arrangements did not follow the formal land use plans for the area, and no formal bylaw was developed. In our interviews, technical officers explained that a formalisation would have required approval in Nairobi. That would not only raise attention, but would also be time consuming, and would bring up the thorny issue of who actually held jurisdiction under which law in a situation such as this: Environment, fisheries, water or agriculture? A bylaw would furthermore have been impossible to enforce in practice, given the overlapping mandates, limited staff and resources available to the technical officers. Instead, the involved technical officers enlisted the Community Fisheries Cooperatives and farmer’s CBOs themselves to monitor that the agreements were upheld, and to work out solutions on their own as far as possible. As a result of these activities, tensions in the conflict abated to some degree. At the time of writing, the underlying conflict remains, and new farmers still occasionally settle near the lake. When this happens fishermen invoke the prior agreement, and go to the Community Fisheries Cooperative and/or farmer’s CBOs to resolve the issue, and in some cases the Chief.

The events around Lake Jipe indicate the importance that EOs and other technical officers attach to avoiding “unrest” among communities. It also illustrates how EOs delegate state functions to CBOs and fishermen’s cooperatives: In principle the development, monitoring and enforcement of rules over natural resources is their task, but in most cases this is practically impossible on an everyday basis, given their limited time and resources. Instead, local civil society organisations take over, and thereby for their part gain a share in the “public authority” associated with performing state functions (Lund 2006). For the EOs, the risk of “sharing” this authority are far outweighed by the legitimacy it provides, the benefits of avoiding conflict, and the ability to report to superiors that inclusive approaches have been applied as per their mandate. Accordingly, when a new EO or other technical officers arrive in the area, they are informed of these arrangements both by their colleagues and by the communities involved.

9. Discussion

As the above cases illustrate, EOs in Taita Taveta seek to negotiate their complex roles through highly pragmatic everyday practices. One might ask, don’t these cases merely represent EOs doing “collaborative resource management”? In a broad sense they do but not as a planned programme of interventions, and not within the confines of the legal-rational framework of the Environmental Act or the formal state apparatus: The cases demonstrate approaches that draw from a continuous blending of formal and informal practices and means, whether it is the mixing of formal and informal decision-making processes, the tailoring of ad hoc rules and agreements, or the delegation of public functions and authority to “civil society” community organisations. A key aspect of the EO’s approaches is thus a continuous blending of formal and informal practices, rules and modes of organisation that cut across statutory, communal and personal domains.

The EOs in Taita Taveta are thereby bricoleurs who – faced with limited resources and an unrealistic mandate – seek to address the challenges at hand with available means and in creative ways. In this they are not alone: The cases above illustrate how other technical officers from line agencies collaborate with the EOs in these practices. For example, the Mining Officer has established informal agreements with communities in order to settle disputes over small-scale mining practices, while Wildlife Officers allow Maasai herdsmen to graze cattle in Tsavo National Park in return for information on poachers. Importantly, these practices do not reflect incompetence, disinterest, or a happy-go-lucky approach to things. During interviews, the EOs and other technical officers often discussed their informal approaches in fairly explicit and strategic terms, describing them as the best possible way to get things done under the conditions. Some directly referred to creativity as an unavoidable part of being a government representative at the local level.

This latter point is significant: The EOs in Taita Taveta do not act in opposition to the state as such, or erode their own underlying authority. Rather their efforts can be seen as a way of making the state function in a context of limited state “reach” and capacity. As one officer – referring to his ministry – said with a smile: “We are just trying to keep this old truck on the road”. From this perspective, the everyday practices of the EOs are in fact a prerequisite of the African state, a necessary means for it to function in the local context in the first place. In this there are similarities to the district officers of colonial times, who were also expected to act at their own direction, implement conflicting mandates and avoid unrest among the populace. Indeed, some of the measures employed by the EOs – such as the delegation of functions and authority to resolve resource conflicts – echo Berry’s (1992) description of how indirect rule played out in practice in Kenya’s colonial times.

The blending of legal-rational structures and informal relations through which the EOs work is well in line with the neopatrimonial qualities often ascribed to the African state (Erdmann and Engel 2006). However, simplistic interpretations should be avoided: It would be problematic to view the EO as the supreme head of a hierarchical clientilistic structure in Taita Taveta. For example, even the “collaborating” EC members did not always agree with the EO, and on some occasions vetoed his suggestions or took up an issue with other technical officers on their own account. Moreover, as Therkildsen (2014) has pointed out, to equate all informality in African bureaucracies with political mobilization and a collective plundering of the state would be a gross simplification. For the EOs in Taita Taveta, the formation of informal relationships with EC members and communities is first and foremost about being seen to deliver by their superiors and colleagues. To do this means to “get the job done” which – like their colonial predecessors – entails reproducing authority while avoiding trouble from communities. In such a situation, forming strategic alliances with those who have shared concerns and are collaborative has proven an effective approach.

While the practices of the EOs are thus embedded in a particular Kenyan context, they also reflect more universal traits of front-line bureaucrats. Their behaviour is thus similar to what Mathews (2011) and Blundo (2014) found among forest officers and game rangers in Mexico and Senegal respectively, and what Hamani (2014) calls the “inventive practices” of gendarmes and policemen in the district courts of Niger.7 There are many parallels, too, to the “street level bureaucrats” of western bureaucracies described by Lipsky (2010). The informal aspects of the EO’s practices in Taita Taveta are in other words not accidental or unusual, but rather reflect what Olivier de Sardan (2014) calls the “practical norms” of public actors, which form part of the habitus (Bourdieu 1990) of frontline bureaucrats.

The practices of the EOs in Taita Taveta can thereby be seen as part of an ongoing bricolage of formal and informal modes of governance and intervention by the local state. Does this mean that the resulting hybrid arrangements are simply the dictated products of the EOs? Not quite: The efforts of the EOs to negotiate their complex position and carry out their mandate take place in an ongoing interface with communities and other local actors, who actively seek to shape state interventions in resource governance (Long 2001; Funder et al. 2013). Moreover, community members are themselves active bricoleurs (Cleaver 2012; de Koning 2014), for whom the ongoing maintenance and reconstruction of social relations is a key part of securing access to resources (Berry 1993; Ontita 2012). Indeed, although our focus here has been on the agency of the EOs, the cases discussed also show how communities contribute to the arrangements in question. The hybrid arrangements around the environmental law in Taita Taveta are thereby to a large extent a joint product of the interaction between EOs, communities and other local actors. Although this interaction rarely takes place on a level playing field, it emphasizes that we should avoid notions of state organisations as somehow detached from local society and processes of bricolage (Anders 2009).

It is also important, however, to consider the two-sided nature of the practices of bricolage employed by the EOs. In Taita Taveta, the most obvious example of this is the so-called “squatter” issue: With a large share of land owned by people from outside the County, some communities in the area claim that they have been alienated from their historical lands. In consequence, they have settled on the private farmlands owned by some of the most influential families in Kenya (Onoma 2010). Although this formally makes them squatters on private land, they have called for land rights and titling with the local land administration authorities. Under the watchful eye of the powerful landowners and Kenyan media, land administration officers have been quick to defer the issue to national level. In response, communities have instead voiced their grievances with the EO and Environment Committee, referring to the environmental law and its somewhat hazy references to indigenous property rights. This time, however, there was no deal. The approach taken by the EOs has been to steer away from the issue, claiming that these are not “environmental” grievances and therefore fall outside the Environmental Act. The situation thereby illustrates how the EOs’ discretion in interpreting the law and its boundaries can also work against communities, and how the extent and outcomes of state-community interaction in bricolage will vary according to the specific political dynamics around a given resource in a given location (Wardell and Lund 2006; Mwangi 2010).

It is furthermore important to emphasize that the pragmatic and discrete practices of the EOs do not always work. For example, the EOs frequently received complaints from communities over restrictive forest use laws in certain parts of the Taita Hills. Attempts by the EOs to broker a pragmatic informal agreement between communities and the Kenya Forest Service have failed. There are several reasons for this: Firstly, the forest areas in question are close to Wundanyi, the former district headquarters, and play an important symbolic role in the identity and authority of the Kenya Forest Service in Taita Taveta. It is in other words an area where formal appearances are important to maintain. Secondly, the issue is quite heavily politicized as it has become a platform for a particularly vocal CBO to express wider grievances over community rights. Finally, national level turf wars between NEMA and the Kenya Forest Service (GoK 2013) means that forestry staff in Taita Taveta have a guarded relationship towards the EOs.

The hybrid arrangements for environmental management in Taita Taveta are thus not all-pervasive: Typically they do not apply where there is much external attention, where local technical officers have conflicting interests or lack control, and where more powerful actors rule – such as on the large privately held estates and farm lands. Nevertheless, in other areas they exist in several varieties: After 9 years of implementation in Taita Taveta, much of the Environmental Act that relates to communities and common-pool resources consists of arrangements and practices that blend informal structures and agreements with formal ones. As illustrated in the cases above, this includes public decision-making structures that mix representative democracy with informal networks; agreements over resource access and control that are purely verbal and informal, yet are backed by the authority of the legal-rational state; and management mechanisms where public authority and state functions are delegated to and assumed by civil society organisations. More fundamentally, perhaps, it also includes a blending of different norms and cultures for how the state and the public interacts in environmental management, as conventional technocratic planning procedures mix with informal personal relationships, and in turn with participatory approaches facilitated by a “service-providing” state.

While these arrangements clearly reflect hybridity and are the products of bricoleurs at work, it remains to be seen whether they are also sufficiently enduring to become more deeply entrenched institutions (Cleaver 2012). In this respect it is interesting to note how new EOs are introduced to these arrangements by communities and other EOs as matter of fact, and how they are so far being actively reproduced by the involved parties. This seems to suggest at least some degree of institutionalisation, although clearly the environmental law is still very young if measured over the longue duree. What seems clear, however, is that practices of bricolage are themselves an institution – a practical norm employed as much by the frontline bureaucrats of the local state as by the citizens with whom they engage.

10. Conclusion

Institutional bricolage is not only the domain of communities and other “non-state” actors. Our findings show how Environmental Officers and other representatives of the local state in Taita Taveta are actively involved in processes of bricolage in the everyday governance of natural resources. In a context of limited state reach, unrealistic mandates and ongoing competition over resource control and administrative jurisdiction, these front-line bureaucrats draw on both formal procedures and informal practical norms as they go about implementing national laws and policies, and seek to negotiate their complex position between local communities and the central state. In this space of compromise and pragmatism, the interests of technical officers and communities may converge and result in the production of informal agreements and rules, hybrid organisational arrangements, and a transfer of public functions to civil society. However, the informality and ad hoc discretion that characterizes the practices of local state officers can also be used against the claims of communities, and it can be understood as a means whereby the local state reproduces itself on the ground – and thereby also its presence and authority in environmental and natural resource governance. This highlights that while bricolage may be enabling and even liberating, it also has constraining qualities.