Practical Approaches to Good Governance: Working with the Grain?

Our point of departure for this article is the idea that governance involves a plurality of arrangements that are adapted by different stakeholders. Much contemporary thinking about African development is concerned with the hybrid nature of governance in which official rules and mechanisms combine in various ways with local practices and the norms of moral economy, including ideas about mutuality and the right to subsistence (Hyden Reference Hyden2006; Chabal 2009). Increasingly, instead of being seen as dysfunctional, these are being investigated as “arrangements that work,” practical hybrids that can secure best fit between development policy imperatives and local practices. As such they seem to provide a way of “working with the grain”—of adapting to an existing context and extending the realm of governance from formal domains of professionalized decision-making into everyday interactions (Booth Reference Booth2011; Reference Booth2012).^{Footnote 1}

Such thinking is an adaptation of mainstream development policy approaches, which suggest that, within broadly neoliberal models of economic growth and democracy, getting institutions right can help promote good governance and vibrant economic activity (Grindle Reference Grindle2007). Where such policies are applied to natural resource management, two intersecting trends are apparent. The first is one of selective formalization enacted through legislated rights to land and water, the registration of associations, the design of tariffs, and the codification of rules and sanctions to regulate resource access (Palotti 2008). The second is a strategy of normalizing informality through support to community decision-making. Here the deployment of culturally acceptable norms, roles, and practices and the facilitation of both associational and entrepreneurial livelihood activities are intended to generate social and economic capital (Osei-Kufuor Reference Osei-Kufuor2010; Cleaver 2004).

The dominant policy vision, then, sees the potential of harnessing both “formal” and “informal” arrangements in pursuit of “good governance.” In relation to the management of natural resources, analysts suggest that this involves recognizing polycentricity: institutions form a mosaic of interconnected arrangements in which there is no single governing authority (Lankford & Hepworth 2012; Andersson & Ostrom Reference Andersson and Ostrom2008). From a plurality perspective, the governance of resources like land and water is enacted through a variety of stakeholders (state and nonstate actors) with plural configurations of laws, rules, and procedures (coexisting customary and modern legal regimes) and with a diversity of uses and values (Merrey et al. Reference Merrey and Molden2007). Policy approaches emphasize that such governance arrangements can be adapted to individual contexts through stakeholder representation in public decision-making (Ramalingam et al. Reference Ramalingam2008).

Recently emerging academic literature analyzing the “real” and “practical” nature of governance in Africa tends to be more nuanced than these policy approaches (while also partially informing them). This literature stresses the everyday improvisation and informal negotiations that produce governance arrangements that are neither entirely customary nor wholly bureaucratic, but something new and different (Olivier de Sardan 2008). The ambivalences and tensions of incorporating hybrid institutions into state structures are highlighted, for example, in studies of the public/private delivery of basic services like health and education (Titeca & de Herdt 2011) or the provision of security through semiformalized vigilante groups in markets and neighborhoods in Somalia, Nigeria, and South Africa (Menkhaus Reference Menkhaus2008; Fouchard Reference Fouchard2011). Such approaches emphasize the normative plurality and plasticity of such arrangements and the differential resources of authority from which actors draw in negotiating them. “Real governance” is formed, negotiated, and contested in the street, the clinic, the market, and the press as well as in the formalized public decision-making arenas of community and local government (Hagman & Peclard 2012).

Critical Institutional Perspectives and Real Governance

Such ideas about the nature of “real” (as opposed to idealized) governance embedded in everyday relations overlap significantly with those in the literature exploring critical institutional perspectives on natural resource governance (Mehta et al. Reference Mehta, Leach and Scoones2001; Cleaver Reference Cleaver2012). Here dominant policy perspectives on good governance have been critiqued for their narrowly instrumental focus on institutional-managerial solutions (Franks & Cleaver Reference Franks and Cleaver2007). Critics also suggest that the mainstream institutional ideas prominent in policy, which often draw on the ideas of Elinor Ostrom (Reference Ostrom1990, Reference Ostrom2008), are overly optimistic about the possibilities of crafting robust local institutions. Mainstream ideas about stakeholder representation and communities are seen as often overly romantic and lacking meaningful power analysis (Ribot et al. Reference Ribot, Agrawal and Larson2006). Those writing from broadly critical perspectives question whether plural governance arrangements really expand the “room for maneuver” for nonstate actors or whether, instead, they increase opportunities for some while multiplying disadvantage for others (Neef Reference Neef2009; Odgaard Reference Odgaard2002).

Such critical institutionalist analyses emphasize, variously, the historical formation of institutions and the complex interplay between modern and traditional, formal and informal arrangements. From such viewpoints rules, boundaries, and scales are “fuzzy” and overlapping, social identities are complex and unequal, and power relationships shape the outcomes of resource management (Johnson Reference Johnson2004; Lund Reference Lund2006). Drawing from this broad school of thought, we adopt an “institutional bricolage” perspective as a way of understanding how institutions are formed and to what effect. Derived from the French, bricolage means to make creative and resourceful use of whatever materials are at hand, regardless of their original purpose. We use this as an analogy for the way in which people patch together institutions from existing social and institutional arrangements. The concept of “bricolage” deployed here draws on insights from anthropology (Levi-Strauss Reference Levi-Strauss1962; Douglas Reference Douglas1987) and from structuration and critical realist social theory (Giddens 1994; Archer Reference Archer2000; Sayer Reference Sayer2000). “Institutional bricolage” consists of the processes in which people (consciously and nonconsciously) draw on existing social formulae and arrangements (rules, traditions, norms, roles, and relationships) to patch together institutions in response to changing situations. Such innovations and adaptations are legitimized by reference to tradition, socially acceptable ways of doing things, and existing relations of authority. The institutions so produced are dynamic hybrids of the modern and traditional, the formal and informal. They are negotiated and structured, uneven in functioning and impact (Cleaver Reference Cleaver2012; de Koning 2011).

In this article we aim to bring nuanced perspectives on “real” or hybrid governance into engagement with critical institutional theory. In particular, we suggest that deploying the concept of institutional bricolage helps us understand just how hybrid arrangements come about, and why some people win and some lose in real governance situations. We deploy a bricolage lens to scrutinize the functioning of local governance practices involving pastoralists in the Usangu Plains in Southwest Tanzania.^{Footnote 2} This case has wider relevance for understanding the ways that the governance of natural resources is deeply imbued with power relations (Raik 2008). We suggest that a bricolage perspective can illuminate the nature of hybridity as a dynamic and uneven process, and illustrate the ways in which hybrid institutions are formed in the social and material practices of daily life. From this perspective we can also see how the scope for negotiating governance is restricted, and the room for maneuver is less for some actors than for others. Individual “bricoleurs” dynamically create and contest governance arrangements, but, we argue, they are also limited in their agency and constrained by their social environment. In exploring these points we raise questions as to how far hybrid institutions, formed through bricolage, offer a promising way of “working with the grain” in African development.

Profiling Resource Governance in the Usangu Plains

The Usangu Plains, situated in the Great Ruaha river basin in Tanzania (some 800 km Southwest of Dar es Salaam), are an area of high natural resource potential and of key significance to Tanzanian economic policy. Characterized by fertile soils and abundant grasslands, and adjacent to a major highway and railway line, the area has proved attractive to farmers and pastoralists for decades. The area derives its name from the Sangu ethnic group, but over the twentieth and twenty-first centuries there has been a history of considerable inmigration of Sukuma and Maasai pastoralists from the north and farmers of various ethnic groups (predominantly Nyakusa, but also Bena, Hehe, Wanji, and Safwa) from the south and west of the country. Despite this appearance of diversity, the population size and claims to indigeneity of irrigated rice farmers mean that they exercise disproportionate political influence (Walsh Reference Walsh2012).^{Footnote 3}

The area is the site of prominent national and international debates about people‒environment dynamics. The swamp and seasonally flooded grasslands (the ihefu) at the center of the plains have been favored by pastoralists for grazing and are considered by conservationists to be an important site for biodiversity. Water from the ihefu empties into the Great Ruaha River, and this flows downstream to the Mtera hydroelectric dam providing power for Dar es Salaam and other urban areas. The Usangu Plains are renowned for their agricultural potential, and both large-scale and smallholder irrigated rice farming have received considerable support from government and international donors. Part of the area was recently designated an agricultural growth corridor under a government and private-sector initiative aimed at developing commercialized smallholder farming and modernized (nonmobile) cattle production (see sagcot.com; see also SMUWC 2001; Franks et al. Reference Franks, Lankford and Mdemu2004, Reference Franks2013 for a review of development of the plains).

In 1993 the unprecedented drying up of the Ruaha River at the end of the dry season sparked a public panic about environmental degradation and depletion. This event and other drivers (such as a focus on irrigation efficiency, and an emphasis on community management following the Rio Summit) shaped a number of development projects in the plains in the 1990s. These included the River Basin Management and Smallholder Irrigation Project (funded by the World Bank), the Sustainable Management of the Usangu Wetland and Its Catchment (SMUWC) project (funded by the British Department for International Development), and the WWF-implemented Ruaha Water Programme (funded mainly by the European Commission and the British Department for International Development). These projects (as well as others) involved designing and formalizing community or group-based institutions for resource management including irrigation associations, pastoralist associations, water user associations, and village natural resource management committees. In 2001 the Tanzanian prime minister issued a statement committing the government to the restoration of dry season flows in the Great Ruaha River within a ten-year period (Walsh Reference Walsh2012).

The impacts of pastoralist livelihood practices on farming and on the environmental well-being of the plains have been contentious subjects in both Usangu and national politics (Charnley 1997; Malley et al. Reference Malley2008; Mattee & Shem Reference Mattee and Shem2006). Pastoralists in Usangu have often been partly settled in village areas for the purposes of farming, but young men and boys travel with their herds to seasonal grazing camps in rich grasslands of the ihefu. The movement of cattle through farming areas to access grass and water is a perennially contentious issue in the plains. The drying up of the Ruaha River was commonly blamed on the pastoralists and their cattle despite considerable scientific evidence generated through the 1990s that upstream irrigated agriculture was to blame (SMUWC 2001;Walsh Reference Walsh2012). By 2006 a number of interconnected pressures triggered a crisis that resulted in the government-initiated forced removal of pastoralists from large parts of the plains.^{Footnote 4} These factors included negative national public opinion about power cuts in urban areas (allegedly caused by too little water flowing to the hydroelectric power plants), government commitment to restoring water flows and modernizing livestock production, and pressure from commercial interest groups (tourist safari and hunting operators) who lobbied for an extension of conservation areas. Between 2006 and 2007 up to seventy thousand people (mostly pastoralists) and three hundred thousand livestock were forcibly removed from the fertile grasslands of the ihefu with the help of heavily armed police, an antipoaching unit, game wardens, and ground and air patrols. The evictions were made permanent by the expansion of the nearby Ruaha National Park and the destruction of those villages (often containing pastoralist settlements) within park boundaries. In 2009 actions to extend the Mpanga-Kipengera Game Reserve further limited grazing areas and required the resettlement of more villagers. During our fieldwork in 2011 the district authorities announced that they would enforce a herd limit of twenty head of cattle per household.

Many of the pastoralists evicted from the ihefu were resettled to eastern regions of Tanzania, though others shifted within the Usangu Plains and resettled themselves around existing villages. The expulsion and resettlement process was condemned by civil society organizations that alleged abuses including livestock theft, extortion of fines and bribes, the forced separation of families, use of violence, and the denial of access to basic food, shelter, and services.

How did these processes and events shape the dynamics of local-level governance? And how did they help to shape hybrid institutions for resource management? Let us begin to explore this through a scenario of “real” or “everyday” governance observed by two of the authors in one village office in the Usangu Plains in 2011.

A Real Governance Scenario

Picture the village government office, a two-room brick building in the center of the village. Here the village executive officer (a government employee) sits intermittently, and members of the Village Council, elected by the Village Assembly (composed of all adult residents) meet.

A young Sukuma pastoralist woman and her infant had been sitting in the inner office for some hours. The office attendant explained that the woman had been “arrested” as a hostage to secure the presence of her husband, who was accused of letting his cattle trample the maize growing on a farmer’s fields. The farmer had discovered the damage two days earlier, had allegedly traced cattle footprints back to the pastoralist’s compound, and then reported the incident to the village government.

The elected chairman of the Village Council had summoned the pastoralist to the office for a hearing, but he failed to attend. The members of the Village Defence and Security Committee sent men to arrest him, without success. The village government chairman then ordered the arrest of the wife as a strategy to speed up the pastoralist’s own surrender.

On avoiding capture by the village guards, the pastoralist had surrendered himself to the chairman of his subvillage, who is also a pastoralist and secretary to the local pastoralist association. The chairman then negotiated the surrender of the accused man to the village office.

When the accused pastoralist arrived at the village office, he addressed his wife angrily. Speaking in Sukuma, he berated her for allowing herself to be captured by the village government. He said that Sukuma women should go to school to learn about their human rights. The woman hung her head and cried, while the committee chairman instructed him to speak only in Swahili so that everyone could understand.

A heated discussion in Swahili between the pastoralist and the Village Council members ensued. In this the man denied the charges, challenged the leaders to produce any evidence, and justified his reaction to their original summons, claiming that to respond and surrender himself would have been seen as an admission of guilt. He also stated that he felt more answerable to the subvillage chairman/pastoralist association secretary than to the village government.

The village government leaders were cross with the man for evading their authority and for encouraging his wife to do the same. They pointed out their role in welfare provision—what would he do when he or his family needed help from the village? As the discussion went on it became less obviously confrontational. The chairman of the Defence and Security Committee, a farmer, and the accused pastoralist sat side by side on a bench, the pastoralist’s child toddling between them and being petted by both. When the discussion came to an end the wife was allowed to go. She tied her baby to her back and cycled off, clearly extremely angry with her husband. A younger member of the village council led the man away, hand on arm, in the direction of the nearby bar.

We later heard from village leaders that they were planning to hold a hearing on the matter. They thought it unlikely that the farmer could produce evidence, and so they would have to use their wisdom to mediate between the parties with the aim of an agreed-upon resolution. However, they also believed that the pastoralist was guilty and “in need of more wisdom” than the farmer. If the dispute could not be resolved at the Village Council hearing it would be forwarded to the police and finally to the primary court.

Analysis of the Scenario: Local Relationships and Broader Governance Processes

How do we understand what was happening here? This snapshot (we don’t know how the situation was eventually resolved) touches upon a number of issues of relevance to institutional hybridity and governance. At one level the situation concerned the competing interests of grazing and agriculture and how they are regulated at the local level through decentralized village government. However, other meaningful processes were also evident as the micro-level use of resources was enacted through authoritative local government and through practical negotiation.

In this snapshot we can see that the boundaries between the formal and informal, public and private are permeable. There is a blending of logics, an overlapping of domains and institutional arrangements at play. For example, the authority of the village government was exercised through the rather extralegal practice of hostage-taking. This “informal” exercise of state authority also overlapped with the marital relationship of the pastoralist couple. The village government and its committees, the pastoralist association, social networks, and household relations were all brought into play, thus confirming a plural and hybrid governance analysis.

We can also see, however, that hybridity is not just a convenient and instrumental blending of whichever arrangements happen to work, offering a handy solution to contentious situations. From a critical institutionalist perspective, negotiations and hybrid arrangements are not just about the service, resource, or property to be governed, but are concerned more widely with boundaries, political orders, claims to identity, and belonging (Lund Reference Lund2006, Reference Lund2012). People’s actions are also “disciplined” by authoritative societal discourses, arrangements, and norms (Agrawal Reference Agrawal2005). We can see that all these factors shaped the behavior of people in the village office scenario. In that interaction common African (“traditional”) ideas about the wisdom of elders and preferences for resolution and reconciliation (Maganga Reference Maganga2002) were interwoven with underlying assumptions of the guilt of one party. In Tanzania wider societal stereotyping demonizes pastoralists and their practices as “backward,” and Usangu is not the only area in which farmer‒pastoralist competition is politically prominent (Brockington 2001; Igoe Reference Igoe2010). Government policies aim to redirect the livelihoods of pastoralists toward “modernized” livestock production by reducing herd size, making settlements permanent, formalizing land title, enclosing common grazing land, and expanding agriculture (Benjaminsen et al. Reference Benjaminsen, Maganga and Abdallah2009; Mattee & Shem Reference Mattee and Shem2006). We can see how such discourses permeated interactions between people at the local level. Underlying the particular incident of crop damage were broader contestations over livelihood practices and claims to property, resources, and citizenship. In the Usangu Plains, an area of growing population and agricultural intensification, where land and water are increasingly appropriated by the state, private companies, and individual farmers, even everyday pastoral practices have become highly charged with wider meaning.

Statelike authority reaches into local institutions and encounters between citizens, into everyday politics and livelihood interactions. In the shifting alliances and accommodations through which political authority is exercised, statelike authority works as a resource or legitimating device for particular arrangements formed through bricolage (Lund Reference Lund2006; Jones Reference Jones2009). In situations where government malfunctions or outsources service provision to the private sector, NGOs, or communities, the idea of the state retains dominance as the image of authoritative regulation, thereby shaping negotiated practices and arrangements (Titeca & de Herdt 2010).

In our scenario, the pastoralist clearly resisted the authority of the village government (dominated by farmers) and made his doubts about their sovereignty over him clear. He forcefully reminded his wife of their rights as Tanzanian citizens, which he claimed the village government had violated. He used the Sukuma language as a way of asserting his own values and identity in resistance to the village government, and of chiding his wife. However, we can speculate that he knew that his scope for negotiation was limited, so he also relied on the pastoralist leader who, as one of the elite of the village government, was able to bridge pastoralist/farmer interests and draw on both ethnic identity and statelike authority to help broker a settlement.

Hybridity analyses tend to focus in this manner on the ways in which governance arrangements are negotiated or contested between different actors. In the village office scenario, the situation was negotiated by all parties. They variously drew on statelike authority and on social relationships, with claims made to local and national norms and discourses. However, the concept of negotiation, with an implied emphasis on decision-making effected through discussion, does not adequately capture the practical dimensions of this process. Physical embodiment and the practices of material resource use interact with discursive negotiations in a number of ways. In this incident, the growing crops were materially damaged by deliberate or accidental herding practices; footprints were tracked in order to identify a culprit; authority was asserted by the acts of arrest and hostage-taking and initially resisted by running away; a negotiated resolution was furthered by the use of sociable body language and physical contact in discussion.

In advancing hybridity analyses commentators tend to emphasize the dynamic, creative, and voluntaristic aspects of institutional formation. In doing so they background the broader factors that pattern governance outcomes such as demographic trends, political economy, and property regimes (Hagmann & Peclard Reference Hagmann and Peclard2010). Some commentators question the reach of a “negotiability” analysis, a point cogently made in relation to the persistence of structurally unequal land rights in Africa by Pauline Peters (Reference Peters2004). Dornbos (Reference Dornbos2010) points out that the persistent and deliberate negation of crucial stakeholder interests by those in positions of authority, and the inequalities of power and resources between actors in particular countries, may leave little room for any meaningful “negotiation” about statehood or citizenship. Similarly Etzold and colleagues (2012) point out that it is important not to limit institutional analysis to the “problem of interplay” between pluralities of coexisting institutions. When we analyze the ways in which institutions are formed, legitimized, and work, we see them animated in negotiations between actors, but also in the material and symbolic practices of everyday life and their articulation with societal allocations of resources.

So how exactly is power imposed, negotiated, accepted, or resisted to produce hybrid governance arrangements, and what is the balance between opportunity and constraint? In other articles we attempt to map the ways in which societal resources, governance mechanisms, and the actions of individuals interact to produce variable outcomes for differently placed people in the Usangu Plains (Cleaver Reference Cleaver2002; Franks & Cleaver Reference Franks and Cleaver2007; Franks et al. Reference Franks2013). Here we concentrate on deploying the concept of institutional bricolage to track just how hybrid institutions are formed and how we can understand them as both structured and negotiated.

What Does the Concept of Bricolage Add to Our Analysis?

In the dynamic and sometimes opaque processes of bricolage, institutional components (rules, traditions, norms, roles, and relationships) are continuously reused and refashioned. Critically though, for an institution to work, to fit socially, it must seem natural in some way; hybrids are unlikely to be merely a purposeful combination of coexisting arrangements. Hybrid institutions are legitimized by reference to socially acceptable ways of doing things, claims to tradition, identity, and rights, and appeals to existing or previous relations of authority. This involves ongoing processes of claim-making, adaptation, and reinvention in which institutions are animated by people continuously reacting to changing environments but limited by their imagination and life-worlds (Cleaver Reference Cleaver2012).

We can identify a number of key features that help illuminate the articulation of structure and agency in the creation of institutions through bricolage. First, institutions are formed in the necessary improvisation of daily practice, in the interrelationship of sayings and doings with material things in certain historical and social settings (Bourdieu Reference Bourdieu1977; Everts et al. 2011). People piece together arrangements to address their everyday challenges and to respond to changes in their social fields.

Second, such pieced-together institutions are often multipurpose, and they wax and wane over time. Even when institutions are designed for one purpose, they often evolve through bricolage to encompass others. Bricolage is a fundamentally dynamic process characterized by plurality, unevenness, and temporal intermittence, and as such is difficult to analyze (Berry Reference Berry, Lund and Marcussen1994).

Third, bricolage allows new configurations to seem legitimate, to socially fit. They may be naturalized by analogy with the “right ways of doing things”—notions of proper order derived from the social, natural, or spiritual worlds (Douglas Reference Douglas1987). Tradition is called upon and even invented to confer authority on a new arrangement (Hobsbawm & Ranger 1983). Meaning in the form of legitimizing symbols, discourses, and claims to authority can “leak” from one institutional setting to another—so, for example, official-looking stamps may be used to legitimize unofficial land transactions (Benjaminsen & Lund 2002).

Fourth, people are bricoleurs, shaping governance arrangements through conscious and nonconscious action (Giddens Reference Giddens1984; Cleaver Reference Cleaver2012). Ordinary people with hybrid identities (the farmer is also a mother, wife, citizen, and member of an ethnic group) shape institutions in public discussion, in routinized practices, and in everyday social interactions. People’s actions are also framed by changing moral worldviews (relating individuals to the community, the natural world, God, and the ancestors) and shaped by emotion. The search for meaning is at least as important in shaping individual and collective action as the quest for profit and power.

Fifth, however, bricolage is an authoritative process, shaped by relations of power. Authority, reputation, and assets of individuals all matter when it comes to making and breaking rules. Power shapes processes of bricolage through public decision-making, in everyday social interactions, and through broader social-structural factors. Power often works invisibly, embedded in the frequently taken-for-granted nature of the social order (Lukes Reference Lukes2005). Unequal power relations can be modified through bricolage (by design of mechanisms to promote equality, through public negotiation, and in the daily bending of rules), but the costs to individuals of challenging arrangements are often high—in terms of loss of reputation, goodwill and patronage, payment of fines, time spent resolving disputes, and restricted resource access.

Sixth, institutional bricolage takes place in historical, socioeconomic, and physical settings, in particular societal configurations of authority and material resources. These resources provide some of the material from which institutions can be formed (historical legacy, laws, policies, administrative mechanisms, financial allocations, prominent discourses), but they also shape and pattern these (Sehring Reference Sehring2009; Adger Reference Adger2001). Tracking these factors helps us think about the wider governance frame within which bricolage takes place and the material available to actors in making claims to authority, gaining access to resources, and forming hybrid institutions. Here, too, the often taken-for-granted nature of the social order constrains the ways in which people innovate or imagine alternatives.

In the following sections we return to our Tanzanian research to explore the evolution of Sungusungu, a particular hybrid institutional arrangement involving pastoralists in the Usangu Plains. In doing so we attempt to give due weight both to the patterning effects of societal structures and to the variation derived from the contextualized practices and negotiations of socially located actors. We examine how a hybrid arrangement for law and order and welfare among pastoralists (the Sungusungu) emerged through bricolage. This alternative security organization, emerging due to the perceived failure of the state to ensure law and order in rural areas, drew on claims to the tradition of pastoralist cattle militias, overlapped with other institutional forms (such as village government), and was at various times partially legitimized and partially excluded by authoritative state action. In tracking the Sungusungu and related institutional arrangements over time and placing them in the context of wider societal structures, we raise questions about how far such arrangements can be practical answers to local issues of natural resource governance.

Framing Natural Resource Governance in Tanzania

As a way of sketching out how the configuration of societal resources shapes hybrid local governance and the evolution of the Sungusungu, we discuss below some key aspects of Tanzanian policies, legislation, and administrative arrangements over time. For the purposes of clarity we present these selected themes in loose chronological order, while recognizing that in reality they overlap and coexist in sedimented layers that in turn frame the emergence of hybrid arrangements (Olivier de Sardan 2008).

The postcolonial Tanzanian government has pursued, sometimes concurrently, policies of centralization, decentralization, and recentralization that have resulted in a plurality of governance arrangements (Benjaminsen & Bryceson 2012). Nyerere’s nation-building project and his elaboration of an African socialist model of development in the 1970s left a legacy of top-down state‒society relations that continues to shape local institutions in the neoliberal era (Pallotti Reference Pallotti2008; Heald Reference Heald2006). During the nation-building period a system of government institutions reaching from the national to the subvillage level was established, land use planning was deployed on a grand scale, the concept of a Tanzanian national identity was promoted, and the state was heavily involved in the delivery of basic services. Like other African countries, the postindependence Tanzanian government faced the challenge of defining the place of customary law while building a modern nation-state. Customary and Islamic law were given general recognition and expected to develop organically, but strict parameters were set on their remit and formal written law retained predominant status (Juma & Maganga Reference Juma and Maganga2005).

From the mid-1980s onward the government publicly abandoned the African socialist model in favor of economic and political liberalization. Substantially supported by international donor agencies, key policy directions from the late 1990s included institutional reforms, good governance, decentralization, and community participation. In step with international policy directions, the government championed community-based conservation while also promoting top-down interventions—for example, in the continued expansion of protected conservation areas (Brockington Reference Brockington2008). The policy imperatives for economic growth and for conservation have thus become entwined, with significant consequences for the livelihoods of poor people and the ways that resource dilemmas are presented both nationally and internationally (Kinsella & Brehony 2009; Igoe Reference Igoe2010).

Over the past two decades policy and legislation have furthered the selective formalization of property rights (over land, water and mineral resources, forests) in the interests of both economic liberalization and good governance (Palotti 2008). Additionally, users are increasingly expected to pay for rights and services such as access to water (Lein & Tagseth Reference Lein and Tagseth2009). However, formalization and individualization processes are not straightforward; different strands of government policy can seem contradictory, and “traditional” and “modern” law often coexist (Maganga et al. Reference Maganga, Odgaard, Sjaastad, Derman, Odgaard and Sjaastad2007). In association with selective formalization goes the partial normalization of informal, customary, or unofficial arrangements, and there have been various attempts by the Tanzanian government to “work with the grain” of local sociocultural arrangements and to semilegitimize activities that can support the goals of economic growth and democracy (Dill Reference Dill2010). For example, land legislation includes the possibility of claiming a “customary” right of occupancy within a village, while strategies for growth and poverty reduction promote collective action, social forms of association, and local entrepreneurship in support of government initiatives.^{Footnote 5}

From the early 2000s onward, policy and legislation have increasingly focused on commercialization and intensification of agriculture. Strategies include the privatization of state farms and the promotion of commercial smallholder irrigation through support to microfinance and marketing. The adoption of an “Agriculture First” strategy rhetorically links both large commercial and small-scale agricultural production to the national project of food security—resulting in a privileging of farmers and farming in policy discourses. The government view is that this strategy can achieve both economic growth and sustainable natural resource management (Sokoni Reference Sokoni2008), although critical commentators see tendencies toward the appropriation of land, water, and mineral resources by particular interest groups (Woodhouse et al. Reference Woodhouse, Bernstein and Hulme2000). There are also concerns that certain groups such as pastoralists, hunter-gatherers, artisanal miners, and fisherfolk do not fit easily within a highly commercialized economic model of growth and are poorly represented both in “formal” government and selectively incorporated “informal” social institutions (Lange Reference Lange2011; Malley et al. Reference Malley2008).

Let us now examine how these key trends interact to help shape hybrid arrangements involving pastoralists in the Usangu Plains.

Sungusungu: The Evolution of an Invented Institution

Using a bricolage lens to examine the evolution of a hybrid institution (the Sungusungu) in Usangu, we can see how institutions may wax and wane, gain and shed functions, and be incorporated or excluded from state governance arrangements over time.

The militia that are called Sungusungu were not unique to the Usangu Plains but spread rapidly through East Africa in the 1980s and into the 1990s. In Tanzania the Sungusungu arose following the war with Uganda (1979) and in response to cattle raiding, general lawlessness, and perceived corruption and ineptitude of the police in rural areas. The Sungusungu institutions developed unevenly across the country, but the literature suggests that they often thrived in remote areas where state administration was weak. The response of the government to Sungusungu was mixed, with President Nyerere and the ruling party seeing it as a desirable manifestation of grassroots action and a pragmatic solution to limited government capacity, while the police and judiciary opposed it as threatening their jurisdiction. The Sungusungu was given quasilegal status under the People’s Militia Laws in 1989 and 1997 (Abrahams Reference Abrahams1987, 1988; Heald Reference Heald2002, Reference Heald2006; Nkonya Reference Nkonya2006).

During fieldwork for the SMUWC project in 1999‒2002, we found an active network of organized Sungusungu groups linking young men and boys herding cattle in the grasslands of the Ihefu wetland with one another and with their home villages. At first sight this seemed to be a Sukuma pastoralist militia whose main purpose was to guard cattle. However, this militia had developed through bricolage in ways that drew on traditional elements (age-differentiated roles, horns to warn and summon people, use of magic and charms) but also incorporated new roles and ways of functioning. In the villages studied, Sungusungu had either replaced or supplemented official village defense committees.

Such institutions had become partly cross-ethnic, with Sangu agriculturalists and Maasai as well as Sukuma (pastoralists) sometimes operating as commanders. The inclusion of Sangu farmers might have partly been a strategy to ensure working links to the elite of village society. The Sungusungu was responsible for cattle security and keeping order in the seasonal grazing lands; in some places this remit was confirmed by a meeting of the Village Assembly. The Sungusungu operated on the basis of demarcation of roles between elders and youth, a practice common to all ethnic groups, with the youth being the foot soldiers or guards and the elders acting variously to advise on tactics, identify wrongdoers, bestow charms and medicines, and dispense justice. This semiofficial militia was considered by members to be formally accountable to the (modern) village government, while its operation was shaped by the commanders and by socially embedded principles of conflict minimization and reconciliation.

The Sungusungu, as an institution formed through bricolage, was multipurpose, acting not only to maintain law and order but also as a communication channel and a welfare organization, ensuring the health and well-being of the herders in seasonal camps. Commanders collected fees, and when a herder was sick the fees were used to transport him (by cart, bicycle, or bus) to a hospital. Sungusungu networks also ensured that proper relations were maintained between the seasonal camps and the home villages, with messengers traveling to and fro between them. The Sungusungu foot soldiers, roaming around in the course of herding animals and carrying out their duties, also acted as village messengers, broadcasting information about meetings and events to dispersed settlements.

When we returned to Usangu in 2011, the form and functions of the Sungusungu had evolved further, following a period when the pastoralists had been forcibly evicted from the grazing lands. Many pastoralists with large herds of cattle were now relocated to other areas of Tanzania, but remnants of their families continued to tend smaller herds around Usangu villages. Some influential pastoralists held positions on the Village Council and on the formalized Pastoralists Association, but many were marginalized and felt alienated from the formal institutions of decision-making. We witnessed meetings of the Village Council at which destocking was discussed without the presence of pastoralist members, and informal discussions in which leaders of a water user association decided to take executive action and double the fines for cattle trespassing on the irrigation schemes.

In our study area the Sungusungu, now largely delinked from local government, was therefore operating covertly, meeting secretly in the bush and functioning as a form of resistance to perceived oppression. The main concerns of members were ensuring the practical well-being of families following the removals, pastoralist security (in the face of hostile national government), and the assertion of pastoralist identity under continued pressure to destock and disperse. Informants linked the breaking up of communities and social networks to the loss of collective power of the Sungusungu. One remarked that in the area in which he was resettled you could blow the traditional horn to warn of a stranger or a threat but could no longer rely on people coming to assist you.

Although the Sungusungu had contracted in function and membership to become primarily an ethnic identity organization, elements of incorporation into more formalized governance arrangements remained. For example, leaders of the official Mapogoro Pastoralist Association, formed by government and NGOs to ensure that livestock keepers’ interests were represented in water management decisions, named the Sungusungu as a closely linked organization in a network diagram. Despite such partial incorporation however, pastoralist‒farmer disputes remained prominent in local interactions over water and land (as shown above in our village office scenario) and another government initiative in 2012 aimed to craft a new institutional form to address these problems.

In November 2011 a police officer from faraway Arusha, an expert in farmer‒pastoralist conflict resolution, visited an Usangu village to sensitize residents at the Village Assembly about the need to set up a new conflict resolution mechanism. The task of the new institution, named the Collaborative Security Committee, was to resolve disputes between farmers and livestock keepers as well as to provide legal and advisory services to the community. Nominated members would participate in ward committees together with extension officers, police officers, and village executive officers. The committee was to refer unresolved conflicts to the village government, which in turn might refer them to the police, especially if physical confrontation or assault was involved. Subsequently the village government forwarded the names of five villagers to the Village Assembly for consideration. Two of these were farmers, two were pastoralists, and an additional nominee was a member of village government. There was a heated debate at the Village Assembly about the appropriateness of one of the pastoralists nominated. He was a commander of the Sungusungu and some questioned his hot temper. However, after discussion and rejection of some alternative nominees, he was confirmed in the post.

Hybrid Governance through Bricolage

These sketches of institutional formation and functioning suggest the plasticity and overlapping nature of governance arrangements and the ways in which they also reflect and reproduce dominant power relations in society. Deploying a bricolage perspective helps us to see how such hybridity of governance arrangements comes about.

The evolution of Sungusungu well illustrates the creative aspects of improvisation and adaptation as well as the authoritative role that claims to tradition can play in legitimizing institutions. The Sungusungu arose in Tanzania as an improvised response to cattle theft and the breakdown of law and order. Piecing together arrangements from existing elements of social organization (age-set traditions, clan organization) and livelihood practices (especially herding) meant that the Sungusungu rapidly became both multipurpose and embedded in everyday life. For example, when the young men herding cattle were made guards of the Sungusungu, their role easily expanded into that of community messengers. Abrahams (Reference Abrahams1998) argues that the emergence of the Sungusungu was possible and became systemic because of the rich practices of neighborhood association in everyday life among the Sukuma. Dance societies, cultivating teams, spirit possession, ritual associations, hunting groups, threshing teams, and informal courts all provided a rich store of associational arrangements for adaptation. Drawing on such institutional stock, adapting to circumstances and the exigencies of daily life, the Sungusungu institutions became multipurpose, dealing variously with law and order, theft and debts, welfare and social organization, land and water management, marriage and divorce (Nkonya Reference Nkonya2006; SMUWC 2001).

Sungusungu, then, drew on existing arrangements but is neither a purely customary nor a totally modern institution. In his fieldwork in the 1980s, Abrahams (Reference Abrahams1987, Reference Abrahams1998) could find only flimsy evidence for a direct traditional provenance, although this was widely claimed in contemporary press reports. He defined Sungusungu as a modern grassroots development based on reinventions of, or references to, customary forms. Sungusungu organizing principles clearly do resonate with Sukuma cosmology and culture—especially the delineation of roles according to patriarchal seniority, the part that witch doctors play in communicating with ancestors, and the use of traditional medicines and charms to make the leaders strong and fearless (Nkonya Reference Nkonya2006). In this hybrid institutional form, remnants of tradition (such as the blowing of a wooden horn to call for assistance, the involvement of diviners in identifying thieves, and the use of titles derived from chieftaincy arrangements) exist side by side with more official bureaucratic modes (the secretary takes minutes, cases are referred to the village executive officer or to the Village Council). Reflecting on the use of bows and arrows, Abrahams points out that this both draws on authoritative traditional ways (arrows as symbols of paternal ancestry) and is a pragmatic contemporary alternative to carrying other sorts of weapons, which would incur police or judicial sanction.

A bricolage analysis then suggests that hybrid institutions dynamically formed by combining and adapting existing norms and practices can be innovatory while maintaining the appearance and authority of tradition. Sungusungu arrangements arose and evolved through the exercise of human agency and so inevitably reflect the distribution of resources and power in society. Various commentators tracking Sungusungu in different areas mention the role of prominent individuals from pastoralist groups and from government as initiators, founders, or mobilizers of Sungusungu in different areas (Abrahams Reference Abrahams1987; Heald Reference Heald2006; Nkonya Reference Nkonya2006). Such individuals clearly helped to shape institutions in particular directions—we have seen how in the Usangu case, the character of one individual commander was a matter for public debate.

However, individual actions are shaped in wider societal contexts, and this interaction influences the form and functioning of hybrid institutions. Whose norms become prominent in the piecing together of institutional arrangements—which bricoleurs commanded the most authoritative resources to shape Sungusungu to their advantage? For example, Sungusungu draws on patrilineal social arrangements and rarely involves women; in fact, there is evidence of its working against the interests of women, in some cases enforcing the return to the marital home of those who have left their husbands (Abrahams Reference Abrahams1998).

The particular form taken by Sungusungu in the Usangu Plains is also shaped by factors like migration. In Usangu the Sukuma pastoralists are in-migrants who need to interact with “indigenous” Sangu farmers in pursuit of livelihoods. We can speculate that the inclusion of some farmers (probably themselves also cattle keepers) as Sungusungu commanders in the 1990s ensured not only that pastoralists were linked into village politics, but also that the village elite extended their reach into a newly formed institution. We have also seen how government policy trends in favor of conservation and commercialization of agriculture created a dominant discourse that disadvantages pastoralists. This, along with the physical removal and the break–up of pastoral families and the emphasis on creating new formalized institutions for local governance (the Pastoralist Association, the Collaborative Security Committee), accounts for some of the waning of the Sungusungu as a local institution over a decade or so.

So, the dynamism, innovation, richness, and negotiated nature of institutional formation should not blind us to the working out of broader patterns of inequality enacted through (and sometimes resisted in) these hybrid institutions. Hybrid institutions are multipurpose and operate intermittently, they overlap with both customary arrangements (assemblies, clan, and lineage) and with the bureaucratic and political workings of village government. Their boundaries become blurred and meaning leaks from one institutional domain to another. The twilight nature of such institutions (referring to their opacity, their existence in between formality and informality—see Lund Reference Lund2006) means they are adaptable, but also that they are open to manipulation, or to capture. How, then, do we understand the waxing and waning of hybrid institutions in relation to the state and to authoritative governance relationships?

Though the emergence and spread of Sungusungu across East Africa was rapid, it was operationalized in various forms as a grassroots initiative, an extension of local government, and a neotraditional form of social organization (Heald Reference Heald2006). In Usangu the Sungusungu appeared as a law-and-order and a welfare organization in the 1990s and was semi-incorporated into local government through the Village Council’s Defence and Security Committee. By 2012 it had waned significantly, to become a weaker form of ethnic organization meeting secretly in the bush. However, it clearly still held some authoritative weight, as illustrated by the deliberate (though contested) inclusion of the Sungusungu commander into the government-initiated Collaborative Security Committee.

The effects of the partial incorporation of the Sungusungu into state-led governance arrangements are debatable. Heald (Reference Heald2006) characterizes the attitude of the Tanzanian government toward Sungusungu as formed through an uneasy combination of pragmatism and ideology. Perhaps optimistically, she sees the Sungusungu (which she studied among the Kuria ethnic group in the Mara region in the 1990s) as a resistance to government centralization. According to her, they “reformed and reclaimed the state” (2006:282), and as a result the administration showed greater responsiveness and local community was allowed greater autonomy. But she also cites counterarguments claiming that the influence of the Sungusungu was only limited and temporary because their selective inclusion into local government hampered their ability to challenge relations of power and production.

There are many ambivalences and tensions in incorporating hybrid institutions into state structures through both formalization and normalization. Menkhaus writes about the semiformalizing provision of protection by vigilante groups in a market town on the Kenya‒Somalia border (2008) as evidence of the “mediated state.” Here the subcontracting by government agencies of certain functions to such a group is a pragmatic option for building governance arrangements where government is weak. In such cases there may be clear benefits to local people in terms of improved law and in the assertion of moral community, highlighting ideas of justice, social propriety, association, and belonging (Buur et al. Reference Buur, Jensen and Stepputat2007). However, hybrid institutions operate in shadowy ways, sometimes beyond the law, and moral communities may demonize or exclude certain categories of people. In Menkhaus’s case civic organizations deployed severe “customary” forms of punishment (blood compensation for murder, collective clan confiscation of cattle for theft) to establish law and order and social discipline. In some areas of East Africa the Sungusungu also imposed discipline by extralegal killings, beatings, and witchhunts, though we found no evidence of this in Usangu. So the perceived benefits of building governance arrangements through the selective incorporation of such hybrid institutions must be balanced against the potential for a loss of the “grassroots” energy, the legitimization of exclusion, the infringement of human rights, and institutional capture by predatory politicians.

Conclusion: Diversity, Hybridity, and the Patterning of Outcomes

The evolution of the Sungusungu in the Usangu Plains illustrates well the diversity and hybridity of governance arrangements. People partially and intermittently engage with a variety of institutions, dynamically adapting these through processes of bricolage. Social relationships, identity, belonging, and everyday livelihood practices clearly matter as much as public decision-making in the negotiation of hybrid institutional arrangements.

Power relations, embodied in individuals, differentially shape the ways such institutions evolve. Some bricoleurs command more authority than others, have a greater capacity to allocate resources, and are better placed to negotiate hybrid institutional arrangements. The outcomes of governance arrangements at the local level in terms of their effects on people’s livelihoods vary according to their social position (for example, the difference between being a farmer or a pastoralist, a youth or an elder, a first or a third wife), and people piece together strategies that involve various combinations of working with and resisting formal or bureaucratic governance arrangements.

However, we have illustrated that a focus on dynamism, negotiability, and contests for power among actors is not sufficient for understanding the nature and effects of hybrid institutions. If we adopt a bricolage perspective we can explore the ways in which macro- and micro-processes interact, shaping people’s actions and their ability to form and work with institutions. Taking a broader look at the outcomes of plural and hybrid arrangements for managing resources in the Usangu Plains, we can see an overall pattern in which mobile pastoralism as a livelihood strategy is being squeezed (to the point of nonviability). The extension of protected areas by government and the breaking up of pastoralist families and villages, the prioritization of agriculture, the demographic pressure on land and accumulation of irrigated land in the hands of fewer people (large companies and elite smallholders), and the land titling program all work to reduce the land available for grazing. The same factors limit the viability of pastoralist social organization and pastoralists’ scope for negotiation. On the whole, dominant actors perceive pastoralists as fitting poorly with contemporary models of the economy and of citizenship in Tanzania. As a result, pastoralists’ livelihood opportunities, their ability to exercise political voice, and their capacity to claim rights as citizens are frequently threatened in governance arrangements.

In focusing on the promise that hybrid governance arrangements offer of “working with the grain” for African development, we need to be aware that institutions formed through bricolage are constantly shifting, shadowy arrangements that often obscure the ways in which social patterns of inequality are reproduced. If both designing new institutions and working with the grain of socially embedded institutions serve to reproduce inequalities of power, then what are the possibilities of fashioning more socially just and effective natural resource governance in Usangu and elsewhere? Practical development experiences (such as those reported by Merrey & Cook Reference Merrey and Cook2012) suggest the need to anticipate the evolution of institutions in unexpected ways, to advocate for changes at the level of policy and macro-resource allocation, and to work to increase the ability of marginalized actors to shape governance arrangements.