Sustainable Urban Agriculture in Ghana: What Governance System Works?

Urban farming takes advantage of its proximity to market, transport and other urban infrastructure to provide food for the city and sustain the livelihoods of urban and peri-urban dwellers. It is an agricultural activity which employs more than 50% of the local urban population with positive and negative impacts on local and national development. Urban agriculture is an informal activity not supported by law but in practice is regulated to a certain extent by state institutions, traditional rulers, farmers and national and international non-governmental organisations. Tamale’s rapid population growth, exacerbated by the unplanned development system and institutional conflicts, are factors contributing to the present bottlenecks in the urban agricultural system. In this paper, these bottlenecks are conceptualised as problems of governance. These issues will be illustrated using ethnographic data from land sales, crop-livestock competition, waste-water irrigation, and markets. I will explain how conflicts which arise from these different situations are resolved through the interactions of various governance systems. Informal governance arrangements are widespread, but neither they nor formal systems are always successful in resolving governance issues. A participatory governance does not seem possible due to actors’ divergent interests. A governance solution for this sector is not yet apparent, contributing to food and nutritional insecurity.


Urban Farming in Tamale
The importance of urban agriculture varies considerably between cities in Africa. Urban agriculture is still considered a rural activity in many African cities as it does not fit in with the modern city model [1], as the case of Tamale. Urban agriculture is mostly practiced by the urban poor, but the middle class are gradually getting involved, especially to suppliment their incomes, as well a for esthetics [2]. Urban agriculture in the Tamale metropolitan area in Northern Ghana could potentially contribute to food and nutritional security at the local and national level [1,3] when it is integrated into the spatial planning system of the country [4]. Urban farming is a subsistence and commercial activity which provides food for households and also generates income for urban dwellers [5,6]. More than 70% of fresh vegetables in the city are supplied by urban farmers [3]. Urban agriculture involves not only crop farming but also livestock production where animal waste is transformed as manure and used later to fertilise vegetable fields [7,8]. However, there exist certain bottlenecks which constrain urban farming in many African cities, and Tamale is not an exception [1,9]. These could be conceptualised as ambiguous institutional governance systems which exist at multiple scales and levels due to the way resource politics plays out especially in the urban sphere. This paper explain everyday governance processes that shape urban agricultural practices, thereby meeting the gap described by [1]. It will further describe how the diverse interests of actors influence the functioning of

Notion of Governance
Before examining the data, we shall situate governance in the anthropological literature by reviewing its analytical, descriptive and normative meanings. In this section, we shall describe certain forms of governance systems that are relevant to our study situation. Governance does exist and has been defined by political scientists [16] in normative and substantial terms. These scientists insist on the horizontal nature of governance, thereby increasing the private-public divide, which raises questions on representation and participation. According to [17], if governance is stripped of its normative nature then it will offer an escape path from "popular political expressions". He traces the origin of this "heterogeneous" concept from the English Middle Age period. During this period, there existed corporate governance which dealt with transaction costs and later urban governance which was promoted as a social movement during the Thatcher's era [18,19]. Later in the 1980s, the popular good governance was born, and now we talk of global governance, which is considered in the 21st century as the "substance of a truly built concept" with the World Bank and the International Monetary Fund as the vehicles of this governance system. Governance, therefore, has an "uncertain conceptual status, as a multiform phenomenon", as described by [16].
Governance, according to [12], is one of those concepts that have been reconstructed and marketed in developing countries as a "catch-all category". The normative load is problematic and calls for a new lens in studying it to understand mundane everyday life in society. Governance is a paradoxical development parlance, which is neither recent nor solely dedicated to reforming dysfunctional governments to provide quality public services for the population [12]. According to [20], the sociology of governance is when: "the object of investigation is understood as an emergent pattern or order of a social system, arising out of complex negotiations and exchange between 'intermediate' social actors, groups, forces, organisations, public and semi-public institutions in which state organisations are only one-and not necessarily the most significant amongst many others seeking to steer or manage these relations." Governance is also defined as "structures and processes of power and authority, cooperation, and conflict that govern decision-making and dispute resolution ( . . . ) informal, organisations and social institutions", where interactions with formal institutions including those of the state are recognised and deemed essential [21]. According to [22], this concept is "empty" and neutral, leaving the social actors to fill it with their (emic) theories and practices of governance. Indeed, governance acts as a "cultural market" recognised amongst ''developers" inasmuch as it is a descriptive and analytical concept [23].
Institutions with different ideologies have defined governance differently, emphasising different elements of the concept which fits with their interest or agenda. Governance in the African setting is often defined as the provision of public goods by the state. Contemporary development discourses portray an image of corrupt countries and a need for international intervention to provide or subvert public goods and services [24]. The World Bank's definition of governance illustrates this perspective: "the manner in which public officials and institutions acquire and exercise the authority to shape public policy and provide public goods and services" [25]. The normative definition of governance by the World Bank has been criticised for being polluted and misleading as it masks neoliberal ideas and depoliticises public affairs to the benefit of a technocratic and managerial vision [12]. Governance can broadly be defined: "As a set of interactions (conflicts, negotiations, alliance, compromise, avoidance, etc.) resulting in more or less stabilised regulations, producing order and disorder (the point is subject to diverging interpretations between stakeholders) and defining a social field, the boundaries and participants of which are predefined". [12] This definition refers to how different institutions operate depending on their positionality at a particular time and space, which gives them legitimacy and authority. Governance here is used as a practical tool in reference to the delivery of collective services. As mentioned above, I shall look at governance in urban agriculture from an analytical and descriptive angle, focusing on the delivery of social and collective goods and services to the population. The emphasis and relevance of this definition are that collective actions associated with the state is today implemented by different institutions and actors with specific functions in Africa [15].
Multiple governance systems, with their set of rules, may act simultaneously, complementing or conflicting with each other [26]. State-led governance systems coexist with other informal sectors, which are private or charity inclined. For example, in Northern Ghana land governance is managed administratively by the state and traditional institutions, leading to overlaps and conflicts but in some instances complimenting each other.
Authorities who enforce rules in a given system are rarely neutral. They impose their normative ideas about how society should function [26]. Most players, for example, involved in urban agriculture would be favoured more by a particular governance system than another. These favours are obtained through negotiations and forum shopping [27], as actors seek for individual and collective actors with similar interests to gain public recognition and legitimacy [28].

Forms of Governance
Here the focus will be on "local governance" known to take place in a local arena, where the social reality of the people can be better understood [28]. There exist multiple institutions (internal and external) which sometimes confront and compliment each other with ripple effects on policy at the national and international levels [29,30]. This choice to focus on local governance is also because local governance fits well with the decentralised and subsidiarity administrative typology found in many African states, of which Ghana is inclusive [15]. Governance is examined not only from a power and authority angle, but it also from a managerial, technical and procedural dimension. It deals with concrete action in delivering collective goods and services to the populace while interacting with them. On this note, one can term a local mode of governance as an organised form of public delivery of goods and services by a local institution (carrying out a project for an external actor or not) which can be formal or not. For example, New Energy, a local non-governmental organization in Northern Ghana, is changing farmer's practice through the introduction of solar-powered irrigation facilities which could irrigate over 30 acres of land in a project funded by the United Nation Development Programmes and Ghana Energy Commission.
Olivier de Sardan [15] identified eight modes of local governance in Niger, West Africa. Five of these eight modes of governance have been identified here as relevant to the urban farming system in Tamale. The first system is the formal government, comprised of local and regional agencies responsible for designing, implementing and enforcing policy and law, which directly or indirectly influence urban agriculture. Some of these regional agencies include state land agencies, municipal assembly, the Ministry of Agriculture and Health, among others. These institutions have diverse functions, some of which overlap. In this mode of governance, rules on delivering services are enforced through intermediaries like the police and other local brokers [31]. Also, due to its decentralised state, politicisation has taken over efficiency and accountability, leaving space for different interests to interact and interplay, constructing new rules according to the main actor in control.
Secondly, the chiefly mode of governance is one where the traditional rulers have legitimacy given through historical memoir but also reinforced by the formal government [32]. The powers of the chiefs have eroded in some African states [15], but in Ghana, they have been reinforced where according to [32], chiefs are not accountable to the government nor their subjects in their administration and management of land. This traditional institution holds land in trust for communities according to the 1992 Ghana Constitution and uses the proceeds for their selfish interest even though they have a social obligation to their people. Land purchasers or leasees in Tamale ostensibly need the acknowledgement of chiefs; through whom they can get an allocation note and land titles later from the Lands Commission. Olivier de Sardan [33], Mahama [34] and Arthur [35] all note the politicisation of the chieftaincy institution in Dagbon, which is a tool of underdevelopment.
The associational mode of governance comprises of civil societies known as the third group. They abound in Northern Ghana and Tamale in particular, which is often referred to as the "NGO city". Civil societies stem directly from the community itself, in the form of mutual self-help or social organisations like the ataaya groups. The Urban Agricultural Network (Urbanet), for example, was instrumental in the establishment of the Northern Regional Vegetable Farmers' Union (NRVFU) with the aim to resolve bottlenecks inherent in urban farming. Urbanet encouraged the union to apply to the Business Sector Advocacy Challenge Fund for funds to support advocacy training, which will tackle land use and management issues relating to urban farming that arose especially in 2007. The union also performs one of the functions of the civil society, acting as the voice of the marginalised seeking a livelihood [35]. In wielding their power, they create spaces within which other governance systems can flourish and also hold the formal government accountable. Diamond [36] has distinguished "civil societies" from associations, in which the "social" mechanisms of governance is the main focus. However, Godsäter and Söderbaum [37] and the United Nation Development Programme Partnership Bureau [38] also state that civil societies are not necessarily formally organised. They are sometimes composed of more traditional structures with a weak material and financial base.
In keeping with this, our data has prompted us to define a fourth form of governance, different from civil society. We refer to this form of governance as the informal governance system. The informal governance system's rules and regulations are imposed through informal mechanisms. Individuals and institutions have to adhere to these rules and norms which are reproduced through everyday transactions and interactions. Interpersonal relationships between people serve to reinforce the ideas about what is expected and what is not. For example, when individual farmers borrow each other tools, they return them without being told to do so. Informal governance systems provided spaces within which agricultural activities and practices can be reshaped and transformed. Informal organisations are not necessarily civil society because it does not always have to do with citizens' relationship to formal government. The idea here relates to that of culture, referring to the tradition, norms and belief system of a people. Helmke and Levitsky [39] and Soysa and Jütting [40] see informal institutions as stemming from cultural roots or simply being "non-state". Armstrong and Gilson [41] also note that its objectives vary between cultural settings, according to different narratives and discourses about development and democracy. This idea would, therefore, seem to be more about enforcing social order than providing public services.
I conceptualise markets as a subset of informal social structures. The market can be seen as a macro-force, as some of its characteristics permeate interactions at most scales which include ideas of trade, competition, exchange, and risk. The expectation of being able to use market relationships for financial gain leads people to observe certain ideals about how to behave to maximise that gain, or at least minimise risk. Wholesale traders, for example, adhere to pricing conventions and goods standards. Food vendors similarly attempt to present an image of cleanliness and hygiene to attract customers.
The extraction of this last market subset from the broader category of informal social interactions is an excellent way to illustrate that, in fact, these systems of governance overlap and interact [24]. Different institutions arrange peoples' behaviour simultaneously and may complement or antagonise each other in the process. There may be contemporaneous sets of rules that potentially allow people to behave in some alternative ways in any given situation. Similarly, there may be a number of different authorities claiming the right to enforce certain norms of behaviour. Multiple groups of stakeholders, such as those mentioned in this section, may each be affiliated with one or many of them [41].
Participatory governance is our fifth governance system. This idea acknowledges that within any situation there are a variety of stakeholders with potentially conflicting normative ideas. It conceptualises a system where all parties concerned may contribute towards defining how governance should be performed, reaching a consensus or compromise. Thus, Kiwanuka [42] describes how decentralisation can lead to "good" governance when all stakeholders in a local area agree on an issue. Schneider [43] defines participatory governance as the "missing link for poverty reduction", as it has the potential to improve information sharing and accountability. In practice, participatory governance is not always optimum or possible, as discussed by Lee [44] and expounded upon in the conclusion of this paper.
The patronage, sponsorship-based and religious governance systems are not typical in Ghana, although the Presbyterian and Catholic churches have to some extent been shaping the governance system in Tamale. For example, the Catholic and Presbyterian churches became popular in the northern region of Ghana, particularly through their agricultural interventions in the 20th century. When it comes to urban agriculture to date, there is little or no contribution from these three governance systems.

Case Study Location and Context
Tamale is the administrative capital of the northern region of Ghana as shown in (Figure 1) below. It has no major river, has few seasonal streams and a low water table [3]. These environmental features constrain agriculture in the metropolis. Compounding this, the local government Act 462 section 51(3b) prohibits farming in settlements where the population is more than 5000 without permission from the local government. This law is not adhered to in Tamale where 74% of the population live in the urban area and practice one form of agriculture or another [45]. Nevertheless, informal urban agriculture does take place in backyard farms and open space dry season vegetable farming sites [5,46,47]. It has no major river, has few seasonal streams and a low water table [3]. These environmental features constrain agriculture in the metropolis. Compounding this, the local government Act 462 section 51(3b) prohibits farming in settlements where the population is more than 5000 without permission from the local government. This law is not adhered to in Tamale where 74% of the population live in the urban area and practice one form of agriculture or another [45]. Nevertheless, informal urban agriculture does take place in backyard farms and open space dry season vegetable farming sites [5,46,47]. Farmers cultivate more recently cabbage (Brassica olerace) and lettuce (Latuca sativa), alongside more long-standing species such as okra (Abelmoschus esculentus), amaranthus (Amaranthus spp.), tomatoes (Solanum lysopersicum), roselle (Hibiscus sabdariffa), and ayoyo (Corchorus  It has no major river, has few seasonal streams and a low water table [3]. These environmental features constrain agriculture in the metropolis. Compounding this, the local government Act 462 section 51(3b) prohibits farming in settlements where the population is more than 5000 without permission from the local government. This law is not adhered to in Tamale where 74% of the population live in the urban area and practice one form of agriculture or another [45]. Nevertheless, informal urban agriculture does take place in backyard farms and open space dry season vegetable farming sites [5,46,47]. Farmers cultivate more recently cabbage (Brassica olerace) and lettuce (Latuca sativa), alongside more long-standing species such as okra (Abelmoschus esculentus), amaranthus   These sites can be found on public, communal and private lands, with formal or informal usufruct arrangements [8]. Limited access to the ever decreasing farmlands, water, agrochemicals, as well as less tangible inputs, such as knowledge and market access, has and is still influencing farmers' practices.

Methods
An ethnographic study consisting of participant observation, informal conversations, focus group discussions and interviews was carried out from November 2013 to September 2014 in Tamale. The empirical evidence gathered from these techniques is used to describe four social situations, each of which illustrates the everyday interaction between different modes of governance in Tamale and its effect on the urban agricultural sector. These examples, presented in the results section are derived from two main case studies produced from 20 site visits in which I conducted 40 semi-structured interviews, eight focus group discussions, and three informal conversations. These interviews were carried out with farmers, chiefs, agricultural officials from the ministry and irrigation unit, and staff from the Ghana Water Company, Tamale metropolitan chief executive officer, land agencies, agro-dealers and representatives from civil society organisations implicated in urban agricultural activities directly or indirectly in their different offices or place of work.
Interviews provided general information on the dynamics of urban agriculture, including: the origin of urban agriculture; problems faced when practicing urban agriculture like access and ownership of resources such as land, water and seed; maintaining access and control over resources, and; recommendations for a sustianable practice. The information from interviews were then verified in focus group discussions with farmers where multiple responses were gotten and discussed on the issues raised from the interviews. Focus group discussions (FGD) were carried out with farmers on 10 vegetable sites. These FGD consisted of six to eight farmers, two translators and myself. It was conducted in Dagbani and translated into English, with the help of my assistants. Information from the focus group discussions was triangulated with that from some in-depth interviews and later validated with information from informal conversations. The triangulation process used, reduced bias inherent in social science work. These bias could be as result of answers influenced by group dynamics where answers depend on the presence of certain individuals and can change if those individuals are absent. Interviews, focus group discussions, and some informal conversations were recorded after permission was sort from interviewees. These recordings were later transcribed using F4 transcription by the researcher to get immersed in the data.
Data were analysed from the emic perspective, it started from the data collection stage and ran throughout the research process. Data transcribed was used a reference point to continually probe interviewees on issues I did not understand or what was not clear to better understand the social reality of the farmers from their point of view. This process was repeated throughout the research process in what [47,48] refer to as progressive focusing. After this process, themes and concepts reflected in the transcripts were identified and continually refined as new information was added until I got to the point of theoretical saturation. This, according to [49], "is the building of rich data within the process of inquiry, by attending to scope and replication, hence, in turn, building the theoretical aspects of the inquiry". Theories and concepts developed were used as a tool to check if there were gaps in data collection and if so more data was collected and new themes contextualised. This continuous process of processing data from participants to data analysis and back to participants, made this research work theoretical saturated.

Insecure Access to Land
Urban farmers gain access to land in a variety of ways. Some farmers cultivate on government lands, communal lands (by accessing plots of land through chiefs, family heads or clans) and individual landowners' parcels. These farmers may be squatting, or using these site with the owners' knowledge to prevent encroachers from building on it [50]. There are sometimes overlaps between these categories of land access. As an increasingly scarce commodity in the urban zone, land is often an issue of conflict. Communal lands are entrusted on chiefs for the people and meant to be used for the benefit of the populace [51]. However, many chiefs in Tamale are taking advantage of the lucrative land markets to sell lands, including those under cultivation [52] for commercial and housing purposes. This situation in Tamale, caused a staff member from the regional Ministry of Agriculture to say to me that a Dagmba man has now "become a landless man".
It is worth explaining the standard route to gaining access to land for farming with intention to later get a title. (Figure 3). Sustainability 2017, 9,2090 8 of 17 these categories of land access. As an increasingly scarce commodity in the urban zone, land is often an issue of conflict. Communal lands are entrusted on chiefs for the people and meant to be used for the benefit of the populace [51]. However, many chiefs in Tamale are taking advantage of the lucrative land markets to sell lands, including those under cultivation [52] for commercial and housing purposes. This situation in Tamale, caused a staff member from the regional Ministry of Agriculture to say to me that a Dagmba man has now "become a landless man". It is worth explaining the standard route to gaining access to land for farming with intention to later get a title. (Figure 3). Land developers can obtain leasehold rights to land, even though this process is lengthy, complicated, risky and expensive. Acquiring land for residential or commercial farming purpose starts with identifying the land one is interested in owning. When a plot of land is identified by anyone interested in it, one can contact the Lands Commission to check if the chief in question has the credibility to allocate that land and whether the land has been zoned for the purposes aligned to that of the developer. This information can also be obtained from the Survey and Mapping Division of the Lands Commission, the Town and Country Planning Department and the Customary Land Secretariat. If these criteria are met, then one can approach the chief in charge of the said land for an allocation note which later needs to be confirmed by the divisional chief if one is dealing with a junior chief.
A traditional token consisting of kola nuts and a cash price which is decided based on the land area is presented to the chief, who issues the allocation note. The kola nut and cash are considered a token collected for sacrifices to be carried out on the land before it is allocated. During fieldwork, the transactions between a land buyer and chiefs were often referred to as "land sales" by land "buyers", although traditionally land is leased and not sold. When the chief gives an allocation letter to the new landowner, he then takes this allocation note to the Survey and Mapping Division at the Lands Commission. The Survey and Mapping Division goes back to the site to check if the site plan reflects the reality on the ground, as private surveyors used by chiefs are not always competent and may place boundary markers at the wrong locations. The team from the Survey and Mapping Division, if satisfied with the site plan, prepares a cadastral plan from the layout. Land developers can obtain leasehold rights to land, even though this process is lengthy, complicated, risky and expensive. Acquiring land for residential or commercial farming purpose starts with identifying the land one is interested in owning. When a plot of land is identified by anyone interested in it, one can contact the Lands Commission to check if the chief in question has the credibility to allocate that land and whether the land has been zoned for the purposes aligned to that of the developer. This information can also be obtained from the Survey and Mapping Division of the Lands Commission, the Town and Country Planning Department and the Customary Land Secretariat. If these criteria are met, then one can approach the chief in charge of the said land for an allocation note which later needs to be confirmed by the divisional chief if one is dealing with a junior chief.
A traditional token consisting of kola nuts and a cash price which is decided based on the land area is presented to the chief, who issues the allocation note. The kola nut and cash are considered a token collected for sacrifices to be carried out on the land before it is allocated. During fieldwork, the transactions between a land buyer and chiefs were often referred to as "land sales" by land "buyers", although traditionally land is leased and not sold. When the chief gives an allocation letter to the new landowner, he then takes this allocation note to the Survey and Mapping Division at the Lands Commission. The Survey and Mapping Division goes back to the site to check if the site plan reflects the reality on the ground, as private surveyors used by chiefs are not always competent and may place boundary markers at the wrong locations. The team from the Survey and Mapping Division, if satisfied with the site plan, prepares a cadastral plan from the layout.
According to law LI144, any piece of land to be leased out has to be approved by the regional surveyor for registration purposes. The regional surveyor, who is the head of the Survey and Mapping Division, would then approve the site plan. The new landowner is expected to attach the site plan to the allocation letter and table it to the Lands Commission for a lease. If land requested is for a residential purpose a 99-year lease is given, and if it is commercial or industrial a minimum of a 50-year lease is given. If the land is for farming of permanent crops like cocoa and shea nuts, a 50-year lease is given. However, if the land is for farming of seasonal crop like maize or rice, a 25-year lease is granted. Foreigners are given not more than a 50-year lease for residential or commercial agricultural activities. When the site plan, allocation note and cadastral plan are presented at the Lands Commission, a lease form is produced. This lease form gives columns for the signatures of the divisional chief, the paramount chief, the new landowner, and two witnesses. The document is then submitted back to the regional chairman of the Lands Commission after all signatures have been appended, and he then gives his approval of these documents.
Before the lease is finalised, records are checked to make sure that the allocation is not in conflict with earlier allocation notes by the chief or another chief. If it is encumbered, the lease is sent back to the chief who issued the allocation note. If there is no conflict, then a last trip to the site is conducted to make sure no conflict cropped up on the ground during the processing period. If all is in order, a processing fee is demanded, before this lease is registered, showing that the process for a lease is completed.
It should be noted that whereas land purchasers from outside the community often undergo this procedure, indigenes rarely do; this is partly due to the complicated administrative procedures and high fees, but also because of their perception of the hereditary right to land, which is that all they have use rights to all customary lands [53]. A staff member from the Regional Ministry of Agriculture said "chiefs don't want to sell lands to the indigenes. The indigenes will never allow it because it is their ancestors who owned the land. It is their inheritance; how can you buy your inheritance".
Buipela farmers in Southern Tamale, for example, had usufructs rights to their farmlands allocated to them by the chief. In 2012, much of these lands were sold to commercial and residential developers by the same chief, obliging farmers to leave in search of lands at irrigation sites and peri-urban areas. I interviwed a female farmer in Buipela who also lost her land to chief's sales. I asked her why she did not buy her plot of land from the chief to continue farming and she said: "If chiefs inform the farmers that they want to sell the land, they will buy the land. What chiefs do is that they will just sell the plot of land out to another person and you will only know this when the person is ready to build. In Buipela they don't sell lands to farmers." In Tamale, a chief has a legal right to lease land under his control, even if such leases displaces farmers. A relevant question, posed by Ubink [53] relevant to this study is: does the state declare chiefs as fulfilling their traditional duty to their people? A question which has multifaceted answers depending on the interest of the government institution or focal person answering the question. In Buipela, the chief has allodial rights over land and, according to traditional norms, is meant to use it for the good of the community. However, traditional authorities currently are jeopardising their communities' livelihoods by the allocation of agricultural land for other purposes, as a result of the interaction between traditional and market forms of governance. Due to the lack of accountability, chiefs do not carry out any developments in the community but use the money for their benefit [53]. A staff member at the land's commission paints a vivid picture of chief's lack of accountability to their community. He stated: "I have always been saying that, hmmm in Ghana we are underdeveloped because of our land tenure system. It is part of the problem. Here is the case where there is a sitting chief. Land is money and government can task land and raise a lot of money for development. An area becomes ripe for development then a chief goes in and zones the place. If we have about fifty plots and you are selling each plot ten thousand Ghana Cedi, multiple it and see how much money the chief can raise from these sales. You see, but he sells the land and then spends the money on himself and the immediate people around him, without being held accountable for how he spends the money. There is no law compelling him to account, except that, one enlightened person in the family or in the palace may ask; where is the money? If you go to the palace, you may find a child who is unable to pay his school fees. I hope you understand. If you go to Accra, it is worse what they do. He sells the land; the next thing to do is to marry another wife. I am not saying all the chiefs are bad, some of them are good. The sad thing is that you know the money that is accrued from the sales; he does not use it for the development of the area. He would sell the land without any development and then expect the government to come in to provide the infrastructure services; road, water, electricity. Do you think this is fair? Is not fair I can tell you that when you go to Accra, chiefs are getting a lot of money they are selling lands, for millions of dollars but where do these monies go . . . , they are not being held accountable." Woodhouse [54] also characterises the situation in Tamale as resulting from chiefs' appropriation of a status resembling freehold right, rather than trusteeship, as explained above.
The Gumbihini neighbourhood contains two large farming sites, both positioned on state land. In 1989, a nearby dam burst, destroying houses and offices in these locations. Subsequently, the government declared these sites as disaster zones and a green belt prohibiting any constructions of houses. Farmers have intensified farming on these sites, but chiefs still had interest in leasing out these lands even though the state had prohibited it. The first sign of a lease is the placement of a pillar in the plot, which serves as a notification of intention to build and thus a physical symbol of power. It is a form of claiming land through an inscription on the landscape. Farmers uprooted such pillars as warnings to the land purchaser not to build by taking away the symbol of power by the chiefs. Farmers went further to destroy some of the foundations of houses when buyers do not heed to their advice not to build on these sites. The above incident is one of the reasons why Urbanet facilitated an advocacy training for farmers to handle such issues more diplomatically in future, and in a way that will be sustainable.
Intermediaries such as letters, site plans, and notice boards are used by different institutions to lay claim directly or indirectly to land in a variety of ways. One example is the case of the Gumbihini old dam. This case involves a dispute over a letter, alleged to be from the Regent of Dagbon authorising the Gumbihini chief to sell land in the green zone area. This letter was presented to farmers as an authentic letter from the Dagbon Regent. However, when the farmers met the Regent, he denied any knowledge of such a letter. The Gumbihini chief's explanation later was that his officials had delivered this forged letter to him and that he was unaware that it was not genuine [55]. This example illustrates that the struggle for power over land is not just between chiefs and farmers, but also at different levels of the traditional hierarchy-local chiefs and paramount chiefs. Indeed, some farmers suggested that certain sales are transacted by the chiefs' sons and secretaries without the chiefs' knowledge.
Another route to land acquisition implicates workers from the Town and Country Planning Department (TCPD). Representatives of the TCPD admitted that there were fraudulent staff members in their department. These staff had accepted bribes from chiefs in return for rezoning land as residential areas. The issue above was evidenced by the altering of the Gumbihini Planning Scheme (where lines drawn in pen were found on the plan). Government officials, therefore, assist in accrediting more power over land resources to chiefs.
Farmers claimed that many parcels of land have been sold to multiple developers, with the chiefs' associates collecting a fee from each. An example is the parcel of land on the Gumbihini Waterworks site purchased by lawyer Albert Luguterah. Luguterah bought land on the Waterworks site and discovered that the land had been sold to other developers, and so he opened a case against chief Naa Fuseini Nabila (the late chief of Gumbihini) in court. He put a signpost in the farming site to announce this fact. The signpost had on it a court induction letter from the courts to any one concerned. This signpost represents his attempt to use the formal judicial system to assert his power over this land. Unfortunately, traditional institutions are accountable to themselves and not government institutions, reasons why the chief refused to go to court. This issue was later resolved out of court by a new Gumbihini chief.
Each of these examples from the Gumbihini sites shows how actors have used symbols and documents as tools to assert their power over land resource and other stakeholders. These symbols and documents have rearranged power between governance systems, whether they are traditional, communal or bureaucratic. The resulting practices and the different processes these governance systems undergo shows that everyday land politics present in Ghana to a greater extent is continually reshaped by interactions and agendas of government institutions, traditional leaders and other stakeholders with interest like NGOs and farmers.
In Gumbihini, however, chiefs' sale of land is illegal, as it is a government-owned green belt. Chiefs are therefore circumventing the formal governance system, with the complicity of some government employees. From the farmers' perspective, the outcome is the same: restricted access to the resources they use to support their livelihoods. The points mentioned above relate to the Centre for the Future State (CFS)'s [11] characterisation of monopolistic, dictatorial traditional authorities. CFS characterises such institutions as more likely to be democratic after interaction with "external influences". In this data set, however, the opposite occurs. State and traditional authorities collude within a market system to appropriate resources from individual and public actors. As farmers uproot pillars, a physical symbol of power, and push down buildings, they challenge this traditional governance and governmental systems to begin functioning as they are supposed to.
In 2007, after a house was destroyed by farmers, a participatory meeting was held, involving the MoFA and Urbanet, the regional security force, some chiefs, the TCPD, the mayor among other actors, to solve the problem of access to public land for farming. After the 2007 incident, the farmers' union was formed, and Urbanet helped them solicit funds for advocacy training, an example of what Gödsater and Söderbaum [37] describe as civil society resisting regional governance. Gödsater and Söderbaum [37] describe a situation where a non-governmental civil society organisation has disingenuous intentions. The picture in Tamale is much more complicated: state and traditional authorities act somewhat in concert, with civil society attempting to resist their power, more akin to [15] the idea of a multiplicity of actors who provide collective service alongside the state, or sometimes without the state.

Conflict between Crop Farmers and Livestock Owners
Competition for land has exacerbated the urgency of other issues, notably that of conflict between crops and livestock farmers. This section details two of such instances.
The crop-livestock conflict has been ongoing amongst the community of Zagyuri in North Tamale and its adjacent communities for several years. Cattle from nearby villages stray onto the state land upon which the Zagyuri farmers crop ayoyo (Hibiscus sabdariffa). This cattle invasion of vegetable sites is problematic especially in the dry season when pasture is hard to find elsewhere. The situation has spilt over into violent conflict. In 2005, the cattle of the chief of Sugashee strayed onto the Zagyuri site. Zagyuri residents went to that chief and entreated him to confine his animals. The chief stated that dry vegetable farming is not prestigious; he described vegetable farming as useless and stated that they are low in nutrition compared to maize and rice grown in the wet season. His opinion was that in the dry season the correct use of the land was for grazing, not crop farming. This statement by the chief angered the farmers, and they vowed to kill any livestock that strayed onto their farms. Indeed, farmers from Zagyuri, Sugashee and the neighbouring Katariga all agreed that several animals are killed annually for such a perceived misdemeanour.
Katariga poses a particularly complicated case for Zagyuri farmers. These two villages (Katariga and Zagyuri) share common ancestors. Therefore, it is less acceptable for Katariga's cattle owners to attack Zagyuri residents for the killing of their animals. The indigenes of Katariga and Zagyuri are aware that they are not entirely blameless when such incidents occur. Their children may not guard the cattle well in the dry season, and the animals do escape from their tethers at night. Often, the owner of a butchered animal simply accepts the blame and takes the butchered animal to a meat trader if the animal is not yet dead. However, if the animal is dead, the meat is disposed of as it cannot be consumed according to Islamic law. The chiefs of Zagyuri and Katariga have both advised cattle owners to guard their herds better. In 2014, the situation boiled over again in an incident where two cows strayed onto the farm of a youth, and he killed one of them. The cattle owner in return came to the Zagyuri site and attacked the boy, who was later sent to the hospital. The parents of the boy took care of his bills and did not involve the authorities. When the youth recovered, he later fenced his ayoyo plot.
The state has in fact put in place a mechanism for resolving such farmer-grazer conflict in Tamale. Paragraph 2 and 3 of the local government bylaw of 2 March 2007 dictates that animals should not loiter in town. Animals that stray into farms in the urban or peri-urban areas close to a business centre, school, hospital or other public areasshould be brought to the environmental health officer at the Metropolitan Assembly. There, the animal owner will pay a fee to the owner of any destroyed property. Also, the assembly can retent the animals if the owner refuses to pay. At Zagyuri, however, this law has never been enforced. When I interviwed the chief on how he resolves the crop-livestock conflict he said, "I am not aware animals are destroying their crops or that they are having problem, nobody ever come to tell me that he is doing a vegetable farm. So I am not aware that anything is going wrong there". It was clear from the interview that the chief knew the problems vegetabe farmers were facing but did not want to interven directly. This might be explained from what the Queen mother of Katariga said, "The problem is that even the farmers over there when the Zagyuri chief them to fence their fields, they would not listen to him. if they do not listen to him what of somebody from here. The people farming there do not listen to anybody's advice or anybody's suggestions on how to keep their fields. When the chief sometimes call on them to advise them, they talk to him as if he is not a chief. Sometimes, nobody would answer the palace call, so the chief just leaves alone." In this example, no system of governance can successfully mediate a compromise between livestock and crop farmers' interests by enforcing the law. Traditional and informal land usufruct arrangements do not prevent such disputes. Through their attempts to use such arrangements, farmers are directly undermining the mechanism the state has put in place to mediate such conflicts. Private enclosure by individual farmers is a possibility, but one that may attract reprisals if they appropriate the resources of other actors to do this.

Wastewater Irrigation
There is public concern about the health implications of irrigating vegetables with wastewater [8]. Concerns about health have heightened especially after the cholera outbreak in 2010. Local radio stations were able to interview officials at the Ministry of Health and survivors of the outbreak, who had eaten raw vegetables from fast food vendors before falling sick. Farmers, however, refused to comment on the issue. Some chased reporters off their farms, fuelling the story of their culpability. Raw vegetable sales dropped after this. Farmers at Gumbihini new dam then began to agitate for cleaner water facilities to irrigate their vegetables.
Farmers asked for pipe-borne water in particular, rather than wells or dugouts. The NGO Urbanet assisted farmers in the Gumbihini area to install water pipes in both sites. Media and civil society reaction thus sanctioned the potentially harmful irrigation practices of some cultivators. This loose form of social governance was prompted and mediated through organised civil society and diffused social mechanisms. Interestingly, the state played no role in mitigating the health issues.
The Zagyuri vegetable site is known for the use of wastewater from the nearby barracks [3], and consumers and marketers disparage crops from this location. Focus group discussion with farmers and observation of farmers in this site reveals that they cultivate only indigenous vegetables, which must be boiled before consumption, killing any bacteria. Therefore they have not been complaints about specific incidents of disease arising from consuming vegetables produced here. Some of these farmers, however, have grown raw-eaten vegetables in the past, specifically cabbage. They were able to sell them and did not always make the provenance known to buyers. A farmer in Zagyuri said the following: "Anytime the market women take the vegetables to the market, the retailer would ask the woman where she bought it. If the market women make a mistake and mention that theybought it from Zagyuri, because of the nature of the water they use to water it, alot of retailers will either buy the vegetable at a low price or reject it. So when market women are asked where the vegetable was bought, they don't tell the retailers or consumers that they get it from here they would rather tell them they get it from a different source so they can sell it easily or fast. Some of the sources mentioned are government irrigation sites like Libga Irrigation Scheme." However, they have stopped growing cabbage not only because of consumer pressure but because of water scarcity and poor crop performance [56]. The example above represents another type of market-driven governance system, in this case, one that some farmers have occasionally undermined but which is seen in corporate governance through market transactions.
The state's potential governmental role is evident in a system where public health is a concern. The lack of state involvement in this situation means that informal systems of regulation have developed in order to provide the public with some form of assurance about the safety of the products they are buying. However, a market-led system of public health assurance can easily be circumvented.

Imperfect Marketing Opportunities
Some farmers I interviewed felt that they had successfully negotiated with market traders to reach an understanding about prices. Others gained information from peers to bargain with traders. A small minority of farmers complained that they did not know market prices which gives market traders an advantage over them, resulting in farmers being cheated. Farmers' union representatives noted that they had attempted to tackle the latter issue by starting a marketing cooperative. This would have been an effort to impose a formal market system, mediated through a civil society organisation. However, the plan never materialised. The reasons given for this included lack of funds for storage facilities and lack of interest in a cooperative marketing strategy. The existence of familial ties between marketers and farmers guarantees trade for both farmers and marketers when the market is slack, and act to cement other bonds within the household. Indeed, marketing by kin is relatively commonplace in the sites I have visited. A farmer in Zagyuri said, "that is the business of my mother. My mother sells 'ayoyo' , tomatoes, okro, pepper and other things which she buys from me and others".
An informal system of governance, partly fuelled by market relations amongst kin, therefore currently operates, and civil society has as of yet had much less of a role to play. Nevertheless, conflict persists. The informal system favours those traders and farmers who have the power to control market transactions. Although a formal mechanism would advantage farmers in an economic sense, there are social benefits of the current marketing system that mean some favour it instead.
The idea of a marketing board is in some ways reminiscent of the days of parastatal organisations. Neoliberal discourse traditionally reacted strongly against such government intervention in market mechanisms. However, current value chain approaches to agricultural commercialisation encourage private and cooperative attempts to organise into market-oriented groups (see, for example, Yuksel [57] and Dempsey and Campbell [58]). Such group organisation resembles the shape of civil society, yet differ in its market-facing function. Markets are firmly embedded in social relations, whether in a system involving cooperatives or one where familial and informal social ties dominate. Therefore, although the market system of governance has rules, people alter them according to those superimposed by other interacting and contemporaneous governance systems.
The four situations explained above have shown that the different groups involved in the urban farming system have various, often conflicting, ideas about how affairs should be run and who should run them. Where there is no formally established governance system, an informal one can develop. If it is unsatisfactory, or if existing formal governance systems are not functioning smoothly, power clashes occur. Each group has an agenda they wish to promote. Each body therefore also has the governance system it may wish to promote, one that they believe would favour their interests in providing public goods and services. These issues raised have a long-term effect on the socioeconomic activities of the people. As the population continues to grow and the climate changes, there is a need to not only focus on the effect on urban agricultural activities and health-related risks but a new planning scheme which looks at all these complex issues at different levels and scales.

Conclusions-Formal or Informal Governance?
At an UrbanFood Plus stakeholder workshop in January 2014, actors made suggestions as to how the everyday governance of the urban farming system could be reformed or improved. Members of the farmers' union proposed a participatory form of governance, involving dialogue between all bodies involved and an agreement of each stakeholder's role. The proposal from the farmers union members fits with the union's ideology of bringing authorities to account through advocacy and engagement. However, an important question relates to the body that would finally make sure the different actors in a participatory governance model stuck to terms they had agreed on. Finding an ultimate impartial governor would be difficult as it stands. The union itself has a constituency which makes it an unviable candidate. The NGO Urbanet was originally a platform for dialogue between all stakeholders. However, it is now closely associated with the farmers' union hosting their meetings. It can therefore also be perceived as too close to the interests of one group of stakeholders to efficiently facilitate such a participatory process.
More fundamentally, Lee [44] identifies that participatory governance is appropriate only when certain "baseline" conditions have been met: first, all participants must be motivated toward a common goal and, second, they must recognise what the other participants contribute to that goal. These include the existence of common goals which in this case is sustainable agriculture contributing to income generation, food and nutritional security and respect for each other's contributions. The case studies detailed above demonstrated that the second condition is not met in Tamale. The majority of stakeholders share the goal of sustainable agriculture contributing towards income generation and food security, but farmers are especially not part of the decision-making process which could potentially influence policies or laws. There is potential to reconcile the goals of, say, farmers and the health service, market people and consumers and possibly even crop and livestock farmers. However, a minority of powerful stakeholders have as their primary objective the accumulation of personal power and wealth, for example, through the development of a land market, which is one of the bottlenecks to a participatory solution. As the second of Lee's conditions for participatory governance is not met, it is unlikely to be an appropriate remedy in the current situation. On another note, in reality, a common goal might not exist, but a range of practical goals might exist, in this case, a compromise which benefits all stakeholders has still not been evidenced. Another suggestion at the workshop was that urban agriculturebe formalised, and an urban farming policy developed, which ties in with [44] studying down, studying up and studying in the middle. In Tamale the government is unlikely to form an effective governor, in this case, bearing in mind its lack of intervention enforcement and implementation, except in a case where an urban farming policy is introduced by multilateral cooperations or international non-governmental organisations as a prerequisite for certain agricultural interventions or aid, such as the case of the gender-sensitive agricultural policy in irrigation sites [44].
So the question of how to govern this dynamic and promising field is still unanswered. Weak governmental enforcement means alternative informal mechanisms, including those acting through markets, have developed. If current informal systems of governance are working in some areas, is there any need to formalise the system at all? It could be feasible to let it continue as it is, with the balance of power between diverse groups permitting functionality. On the other hand, not all informal mechanisms are working, and as land pressure increases and the taste for raw vegetables grows, it is likely that the sector will see further conflicts which will be difficult to resolve through informal mechanisms that can be easily subverted. The state, traditional leaders and civil society organisations are bodies that can be invoked to implement more formal management systems. There are instances where it has proven difficult for each of them to provide solutions to governance issues, indicating that devolving power to any single body could be problematic. A participatory solution would require actors' objectives to align, indicating a relinquishing of power. Whatever the solution to the impending problems of governance in urban agriculture will be, it will be difficult to implement them, as actors interest are varied and conflicting. For example, chiefs are taking advantage of urbanisation and its valuable land markets to allocate agricultural lands for residential and commercial purposes, reducing lands used by farmers for farming.