Introduction
The question of land tenure in sub-Saharan Africa (SSA) is bound up very strongly with issues of identity and power, and yet is central to the generation of rural livelihoods and the dynamic of state formation. Studies have addressed the relation between agrarian transformation and state formation from several perspectives. One is the critique exploring how neoliberal reforms support the formalisation of land tenure shape inequalities and processes of commoditisation, and yet reconfigure access to land and natural resources (Bernstein 2002; Berry 2002; Peters 2009). In recent years this debate has focused on the role of institutional measures promoting the legalisation and titling of land rights. Policies of land registration are currently under implementation throughout SSA with the support of international donors and, as part of larger programmes of land administration, their objective is improving individual security of tenure in the absence of private property rights. These reforms are inspired by the work of de Soto (2000)1 and are largely considered as a ‘new wave’ of land reform aiming to formalise land tenure without modifying the agrarian systems in place – and the underlying structures of power they are embedded in (Deininger and Binswanger 1999, 651).
The argument for promoting the registration of land rights is that titles improve ‘individual security of tenure’. Greater security is meant to positively impact on productivity, land rental markets and investments on land improvements (World Bank 2008, 139–141). According to Deininger and Binswanger (1999, 249), ‘secure individual property rights to land would therefore not only increase the beneficiaries’ incentives and provide collateral for further investment but … would automatically lead to socially and economically desirable land market transactions.' Land registration is also sponsored as a low-cost mechanism of land reform, whereby a clearer demarcation of land borders is expected to reduce conflicts. However, these assumptions have been criticised from several perspectives (von Benda-Beckmann 2003; Kingwill et al. 2006; Sikor and Lund 2009). First, in non-western societies security of tenure is not just dependent on the provision of rights, but is about how competing politico-legal institutions negotiate claims to land and resources. By empowering certain institutions, registration reconfigures power relations and access to resources in the rural milieu. A second assumption is that land registration leads to the formation of pro-poor markets. Research has shown that registration often creates new inequalities or reproduces existing ones, and yet land users do not necessarily conform to the market rules (Lund 2006). Third, the capacity of the state to guarantee property rights and provide infrastructures and capacity to enforce market rules in rural areas is often overestimated (Sikor and Lund 2009).
Registration of land rights may be regarded as one of the strategies employed by the state to legitimate claims on resources and exercise political control. In this article, I look at land registration as a form of territorialisation of state rule in the rural milieu (Vandergeest and Peluso 1995; Boone 2007; Lund and Boone 2013). Scott (1998) notes that making space and resources legible is a deliberate strategy employed by all modern states in order to govern, control people and consolidate power. Following Sack (1986, 19), territorialisation is ‘the attempt by an individual or group to affect, influence or control people, phenomena and relationships by delimiting and asserting control over a geographic area’. Territorialisation intersects with commoditisation, but they are not the same process and can occur separately. In SSA, whose transition to agrarian capitalism either is incomplete or follows a different trajectory than the western model, the territorialisation of state rule and the institution of a land market are often two distinct processes that might occur separately (Vandergeest and Peluso 1995, 415). Land markets may emerge with or without the direct involvement of state institutions. Commoditisation is therefore not the only driver of social stratification and inequalities in access to resources. Policies of land titling and registration are a clear example of this. By formalising the rights of different users, and providing local administrative structures or existing bodies with the authority to legitimise certain claims, they shape the local politics of resources and are central to the negotiation of access to land.
In this article, I argue that contemporary policies of land titling, regardless of whether they achieve their developmental objectives or not, reconfigure the local arena where political authority over land is exerted, yet open up new spaces for negotiation and contestation of the rights of access to resources. This article analyses the registration of rural land in Siraro, one of the administrative districts of Oromiya region in Ethiopia. I argue that in the case of Ethiopia, land registration reproduces state–peasant relations and, reflecting an increasingly hybrid discourse of land reform (Crewett and Korf 2008), serves to further extend the power of the state in the rural milieu. Land registration in Ethiopia is a strategy of territorialisation to the extent that the formalisation of land-use rights sanctions state ownership of land and, at the same time, empowers local administrative structures with increased authority over land. Land registration is legitimated through two political narratives of rural development, ‘equity and fairness’ (E&F) and ‘efficiency and productivity’ (E&P), which reflect a discourse of land reform that is increasingly hybrid, and constructed through logics drawing both on a socialist version of the ‘developmental state’ as well as on more liberal logics of decentralisation and privatisation. Land registration in Ethiopia is a clear example of how political power is expressed through spatial control and is therefore central to the overall project of state formation and negotiation of access to resources. The case of Siraro shows that this process is also contested, although not overtly, and yet reveals that a counterhegemonic discourse performed by peasants is at play.
Land registration and the territorialisation of state power in Ethiopia
Territoriality is an important element in the exercise of authority in the rural milieu, and studies acknowledge that the state is not the sole actor performing territorial strategies (Lund 2006; Lund and Boone 2013). Rather, overlapping politico-legal institutions attempt to see their authority recognised by employing different territorial strategies (Lund 2013, 16), and yet competing for local power. This institutional pluralism characterises many non-western societies, and SSA is no exception. Ethiopia is to some extent a peculiar case as it is often described as an unusually centralised and unitary polity if compared to other African countries (Vaughan and Tronvoll 2003). As noted by Markakis (2011, 12), by the turn of the twenty-first century the administrative coverage of the ‘highland centre’ and the ‘highland periphery’ of the country, ‘though uneven and ragged at places’, is comprehensive.2 Although Markakis (2011, 356) also notes that the process of state formation in Ethiopia is today far from being complete and that the question of land is central to its solution, the state is by far the most important actor employing territorial strategies to secure control over resources and people and to consolidate political power. Land registration, so far implemented in the four most populous regions of the country, is a clear example of this. Land registration is regarded here as a form of internal territorialisation (Vandergeest and Peluso 1995) pursued by the Ethiopian state to enforce control over land and the people who use it.
The distinctiveness of the Ethiopian case has to do with the historical dynamic of state formation which, being strongly tied to the land question, generated state–peasant relations characterised by the subordination and exploitation of the peasantry over time by the elites in power (Rahmato 2009). Ethiopia lacks a significant colonial legacy that in most other SSA societies considerably constrained the ability of the post-independent state to perform effective strategies of internal territorialisation. Historically, the ‘national question’ in Ethiopia has been framed around the attempt of the Abyssinian centre to secure the control over the southern, western and eastern peripheries of the country (Markakis 2011). This generated a long and controversial history of domination, violent conquest and coercion which yet gave rise to centre/periphery relations informed by a hierarchical system of distribution of political power (Abbink 2002, 157), and based on economic exploitation and cultural subjugation of the societies incorporated into the Abyssinian Empire. Since the reorganisation of the state under Haile Selassie after World War II, the territorialisation of the local administration has been an important facet of the internal territorialisation of the Ethiopian state. Central to this process is the attempt by successive governments to bring land administration under the close control of the state apparatus. This culminated when the military regime of the Dergue implemented the 1975 land reform, a reform which was both radical and effective in centralising politico-legal authority, particularly over land, under the newly established local administrative structures (the Peasant Associations) (Rahmato 1984). The land reform provided for the nationalisation of all lands and their redistribution to the farmers as a way to redress past inequalities (Clapham 1988).
More recently, when in 1991 the Ethiopian Peoples' Revolutionary Democratic Front (EPRDF) came to power, the territorialisation of the state entailed a development strategy based on smallholder agriculture to achieve two main objectives: achieving national food self-sufficiency (Lefort 2012, 681) and promoting the structural transformation of the economy towards industry (FDRE 1993, 5). In the 1990s these two objectives were framed under the Agriculture Development-Led Industrialization Strategy (ADLI), whereby ‘peasant farmers and pastoralists’ are considered as the cornerstone of the country's development strategy. In order ‘to bring about a structural transformation in the productivity of peasant agriculture’, and yet diversifying the economy in the long run (Ibid., 1), ADLI is based on a political narrative of E&F and reflects much of the principles inspiring the 1975 land reform, whose main structure remains unchanged since 1991. The narrative of E&F draws on state ownership of land, and considers smallholder agriculture as the most socially desirable and efficient system of production for rural development. E&F is based on a socialist understanding of the ‘rural question’, whereby strengthening small-scale farming through strong state interventionism in rural affairs is necessary to promote balanced growth and development. Support to smallholder farming is achieved through a number of national programmes providing agricultural inputs, extension facilities and food aid distribution. The narrative of E&F is central to land registration to the extent that the formalisation of land use rights sanctions state ownership of land. More broadly, E&F legitimates state interventionism and constitutes a powerful discourse for the territorialisation of the Ethiopian state in the rural milieu.
While in the 1990s the government's agenda was centred on the support to smallholder agriculture and a narrative of E&F prevailed, more recently the strategy for rural development has been reformulated towards a narrative of E&P. After the purge of the so-called ‘left wing’ of the Tigrayan People's Liberation Front in 2001 and the unexpected outcome of the 2005 elections (Lefort 2007; Aalen and Tronvoll 2009a; Clapham 2009), a new vision of sustaining the structural transformation of the rural economy on the way to promoting a ‘managed or inclusive’ form of agrarian capitalism emerged within the ruling elite. In 2005 the party/government adopted a new federal land law, the Rural Land Administration and Use Proclamation (RLAUP) (FDRE 2005a), which, together with a proclamation ruling the Expropriation of Landholdings (FDRE 2005c), mark a general shift in the country's land policy. The RLAUP is informed by two main political priorities. First, by emphasising the urgency of establishing an information database identifying rural landholdings, the land law speeds up land registration and provides for a general framework of its implementation. Second, in order to improve the efficiency of rural development, decentralisation of land administration is prioritised. The RLAUP can be considered a main point of departure from a rural agenda previously dominated by E&F. Yet, it legitimates land registration and decentralisation as important strategies of territorialisation of the state in the rural milieu.
In the same direction goes the 2002 Sustainable Development and Poverty Reduction Program (FDRE 2002), which for the first time prioritised agricultural commercialisation, as opposed to food self-sufficiency, for rural development. This programme was followed by the 2005 Plan for Accelerated and Sustained Development to End Poverty (PASDEP) (FDRE 2005b), whereby an ‘accelerated growth strategy’ is possible only by capturing private initiatives of farmers and by supporting diversification and commercialisation of agriculture (Ibid., 27). Under PASDEP, land productivity, commercialisation and structural efficiency of the local administrative structures are given much more attention than before. While reaffirming state ownership of land, less emphasis is given to smallholder agriculture. An accelerated growth strategy is to be achieved through the ‘intensification of marketable farm products’, both for domestic and for export markets. Shifting to high-value crops so as to exploit market niches, individuation of high-potential areas, support to large-scale commercial farming and stimulating market development by providing basic infrastructures for cash-crop agriculture are among the challenges envisaged by the programme (Ibid., 46–50). The implementation of the Growth and Transformation Plan (2010–2015) goes in the same direction and commercial agriculture is seen now as the backbone of economic and social development at all levels, on the way to making Ethiopia a middle-income country by 2025 (FDRE 2010).
The narrative of E&P emerging in recent years is based on an export-oriented economic model to provide a solution to the long-standing productivity crisis of the Ethiopian rural areas. More broadly, this reflects a growing commitment in the ruling party to promote a structural transformation of the rural economy on the way to achieving a ‘managed’ form of agrarian capitalism. Stimulating the productivity of smallholder farmers and connecting their production to the market are among the main political priorities underpinning this narrative. Far from being based on a neoliberal model, the overall idea is creating a market-based economy pushed by a strong ‘developmental state’ (Lefort 2012; Zenawi 2012). For the elite in power, state interventionism is essential to perform and control commoditisation, market formation and socio-economic differentiation. Decentralised land administration, based on the deconcentration of powers and authority to regional and district government bodies, constitutes a main pillar of this narrative (Chanie 2007). The narrative of E&P underpins an idea of rural development oriented towards the formalisation of the economy, and primarily of land tenure, by means of the territorialisation of the developmental state in the rural milieu.
Land registration is legitimated by a narrative of both E&P and E&F, as it aims to create the institutional conditions for a ‘structural transformation’ of the rural economy driven by a strong developmental state that keeps the monopoly of the administration of property rights to land. Land registration is central to the strategy of internal territorialisation of state rule and, as I will show in the following paragraphs, it reconfigures how authority is exerted in the rural milieu, while opening new spaces for negotiation and contestation of the rights of access to land.
Land registration in Ethiopia
The Ethiopian government has implemented land registration since the late 1990s in collaboration with some international donors, most notably the United States Agency for International Development (USAID) and the World Bank (for instance ELTAP 2008), but more recently also the UK Department for International Development. The programme covers the four most populous regional states – Amhara, Tigray, Oromiya and the Southern Nations, Nationalities and Peoples region (SNNP) – and aims to provide each farming household with a land-use certificate in the name of both the spouses. The programme does not undermine the 1975 land reform and reaffirms the principle of state ownership of land. For instance, land transfers are limited to short-term leases. A local land registry is kept at village (kebele) level, and the programme follows two stages of implementation. At first, plot demarcation is implemented through so-called ‘traditional’ tools. A more detailed cadastral survey – involving Global Positioning System (GPS) technology – is scheduled for the second phase of the programme. Whereas Phase 1 is largely completed, with the exception of a few pilot initiatives in Tigray and Amhara Phase 2 has not been implemented yet.3 In Oromiya and SNNP the distribution has been slowed by problems such as the fragmentation of landholdings, shortages of resources and the redefinition of administrative boundaries of kebele and wereda (district) (Holden and Tefera 2008, 32).
Land registration is implemented by Land Administration Committees (LACs) nominated at kebele level. With some differences from region to region, LACs are mixed bodies, composed of local officials, elders and other community members. These institutions therefore play a pivotal role during the implementation process, as they have the difficult task of ensuring strong local participation and mediating potential conflict during demarcation. The 2005 land law empowers LACs with even more ambitious tasks: performing as court of first instance for land dispute and setting out land use and environmental protection planning. The kind of certificate issued by the LACs varies from region to region, and for instance in Tigray consists of a single piece of paper, while in Amhara and Oromiya it is a small book, called ‘the user book’.
Land registration in Ethiopia aims to improve individual security of tenure, strengthening productivity and food self-sufficiency (Rahmato 2009, 181). A number of quantitative studies in the field of agriculture economics have identified significant positive impacts of the programme on reduction of land border disputes (Holden, Deininger, and Ghebru 2010), land rental market formation (Deininger, Ali, and Alemu 2011; Holden, Deininger, and Ghebru 2011), household welfare (Holden, Deininger, and Ghebru 2011), gender equality (Holden and Bezabih 2008; UN-HABITAT 2008; Holden, Deininger, and Ghebru 2011), investment and land productivity (Deininger and Jin 2006; Holden, Deininger, and Ghebru 2009). These studies report that registration was characterised by a strong community involvement, and that implementation was quick and low cost, and welcomed by farmers. A critical element of the programme is the relation between registration, security of tenure and productivity. Following Deininger et al. (2008, 1802–1803), growing individual security of tenure generates incentives for investments in agriculture, thereby increasing land productivity. Thus, land registration is expected to improve short-term land rental market participation. Certificates are distributed to build farmers' confidence that their land will be returned once contracts expire (Holden, Deininger, and Ghebru 2011, 33–35). Other qualitative studies show that land titling raised considerable expectations among farmers throughout the country and generally the distribution was welcomed by the vast majority of farmers (Rahmato 2009, 183). While these quantitative and qualitative studies have often tended to address specific components of the programme, in the following paragraphs I aim to discuss land registration as a strategy of territorialisation of state rule in the case of Siraro (Figure 1).
Map of the study area at wereda level. Source: Central Statistical Agency, Addis Ababa, Ethiopia, 2010. (The fieldwork methodology encompassed semi-structured interviews with 100 farmers in Siraro. I made use of qualitative methods of data collection and analysis. An open questionnaire was divided into two main sections. The first aimed to obtain general information about landholding, household size, and crop and farming practices. The second included questions about policies and programmes relevant to land management and administration, including land registration.)
Land registration in Siraro
Siraro is a wereda in the West-Arsi zone of the Oromiya region, tracing for most of its territory the border with the SNNP region. Its main economic activity is smallholder agriculture, which employs over 90% of the total population. Agriculture in Siraro suffers from a number of problems. Population pressure over land is growing constantly, and land shortage and poor productivity of agriculture are the main factors behind the recurrent food crisis.4 On average, households farm between 0.5 and 1 hectare of land to sustain large families (7–8 members).5 Given that the population grows fast and land is mainly transmitted through inheritance, landlessness is high especially among young farmers. According to the wereda agricultural office, 67% of farmers received a land certificate by May 2012 (Phase 1, involving ‘traditional’ methods of land demarcation).6 Interviews with farmers and local officials highlight at least three long-term implications of land registration.
The first concern is about the feasibility of constantly updating the land registry created at the agricultural office. Given the highly fragmented system of landholdings, it is certainly a difficult task to keep the registry up to date. The pattern of landholdings is constantly changing every few years and the techniques of land demarcation used by the LACs raise concern about the long-term impacts of the programme. One official of the wereda agriculture office noted:
Keeping the land registry up-to-date is probably one of the most challenging issues of the programme. Plots are very small in our district, and the pattern of landholdings changes fast due to the inheritance system … . I am not entirely sure this is going to work, but we are trying our best. (Interview with a wereda official from the ARDO, Loke Heda, Siraro, 5 May 2012)
Second, the first phase of land titling has involved the use of so-called ‘traditional tools’ of demarcation. In Siraro this meant the use of ropes, or the identification of physical objects (such as trees or stones), on the plot. Sometimes borders were estimated only by sight. These data are then copied down to the certificate given to each farmer, but it is hard to conclude that they provide a clear and undisputable system of plot demarcation. Paradoxically, flawed demarcation made the programme a success because no major conflicts were reported, as well as the programme being fast and low cost. A third concern has to do with the second phase of land registration, which is scheduled for the future and involves the use of GPS technology. Establishing an effective cadastre will require huge investments and technical expertise. Phase 2 is likely to entail higher costs and will require huge investments in both human and financial capital.
In general, the majority of the farmers interviewed were satisfied with land registration. The main reason is that the certificate helps them to ‘prove ownership’. The following quote represents well how farmers normally respond when asked for a general comment about the programme:
I welcomed the certificate. Now I have a piece of paper showing that the land I cultivate belongs to my family. … As they actually prove whom the land belongs to, certificates help in reducing conflicts. (Interview with a farmer in Ropi, Siraro, 27 November 2009)
Nonetheless, more in-depth interviews also show more controversial aspects of the registration process. A common concern of many farmers was that, following the registration, the land tax would rise consistently. Surveying and mapping land and natural resources is central to collecting taxes more effectively and to controlling people's activities and their access to and use of local resources. Yet, given that land demarcation techniques do not provide undisputable borders, the relation between local officials and peasants strongly influences the ways political authority shapes access to land. Land registration has significantly extended the presence of the state in the countryside. More broadly, the politics of land registration reveals that the programme opens up new spaces for the negotiation of local resources and reconfigures the use of political authority over land.
The politics of land registration in Siraro
In the last part of the article, I will explore in more detail three aspects of the politics of land registration in Siraro. I will first explain why land registration fits into the broader context of the territorialisation of land administration, and then address the relation between registration and security of tenure, and state–peasant interactions.
The territorialisation of land administration
Land registration has important repercussions on state–peasant relations as the programme fits into the broader context of the territorialisation of land administration, which in recent years has entailed the decentralisation of prerogatives from federal to local administrative structures.
The literature acknowledges that, rather than being characterised by devolution, decentralisation in Ethiopia is rather a deconcentration of power from upper to lower bodies of government (Chanie 2007; Emmenegger, Keno, and Hagmann 2011). This has to do with the structure of the ruling party, which in Ethiopia mirrors the local administrative structure, and yet strongly influences its decisions and undertakings. Ethiopia is characterised by a de facto single-party rule, whereby local officials are first and foremost party members (Aalen and Tronvoll 2009a). At federal level this system is dominated by the EPRDF, the ruling political coalition composed of ethnic-based parties and other affiliates at the regional level. The administrative apparatus at regional, wereda and kebele levels mirrors the organisation of the party structure. Decisions taken by the party can seldom be questioned by the local officials, who are left only with the prerogative of implementation. Local administrative structures, and particularly the levels of the wereda and kebele, are primarily concerned with implementing government-initiated policies rather than representing the interests of their constituencies, in a context where upward accountability is dominant. In Siraro this is for instance reflected by local officials reproducing the government discourse of land registration as informed by the narrative of both E&P and E&F. This relation is empirically depicted in this statement by one of them:
Land registration serves two main purposes. First it aims to improve agriculture productivity, making it easier for farmers to invest in land improvements. Second, the titles protect farmers against arbitrary eviction. Thus, given that land use titles cannot be transferred but only inherited by other family members, and that the land is public property, land certificates inhibit land concentration and therefore protect the poorest strata of the peasantry. (Interview with a local official of the ARDO, Loke Heda, Siraro, 4 February 2010)
The federal and regional land proclamations implementing land registration reflect the idea behind this quote, by which the ‘developmental state’ is the manager of rural development, and its actions are intended to ‘protect’ the peasants as well as to support local production and growth. As part of this project, the 2005 land law and the enacting regional proclamations shift important prerogatives from central to local bodies of government, providing local officials with increased authority over land. The federal law simplifies the procedures for local bodies to promote land expropriation if the land is not properly conserved and protected (FDRE 2005a, Art. 10). This principle is confirmed in another federal proclamation ruling expropriation, affirming that the ‘competent authorities’ have the power to expropriate rural landholdings for public purposes or when the land is required for ‘a better development project to be carried out by public entities, private investors, cooperative societies or other organs, or where such expropriation has been decided by the appropriate higher regional or federal government organ for the same purpose’ (FDRE 2005c, Art. 3). In Oromiya, the regional land law confirms that land-use rights can be terminated if the land is required for more important public uses or when it is not properly conserved (ONRS 2007, 7–8).
The increased authority of local administrative structures is reflected in everyday practices, as well as in how different discursive repertoires are used to legitimate and resist land registration. Farmers complain that local officials use the power of land allocation, the selection of farmers for development schemes and the allocation of agricultural extensions to reward political allies and to strengthen patronage networks. For instance, some farmers noted that in order to receive the land certificate they were asked to take party membership (interviews with farmers in Siraro, 15 November 2009 and 12 February 2010). Aalen and Tronvoll (2009b, 115) seem to confirm that expansion of the party membership after the 2005 electoral turmoil was a deliberate strategy pursued by the party/government in order to secure consensus.7 Expansion of the party membership through land registration seems to be an important facet of the territorialisation of state rule in Ethiopia.
More broadly, many of the farmers interviewed maintain that ‘the government’ retains the power to dispossess them from their land at any time. While this is seldom put into practice, threat of dispossession is an instrument often employed to maintain farmers' support for government demands, and to not question any decision. An example in this regard is the claim of land for ‘development purposes’. One farmer observed:
The kebele often ask me to participate in community works, such as road construction and tree planting. All farmers are asked to do so, not only me, and without any remuneration. … They claim this is for the benefit of the community, and I do believe it is. … If you fail to attend these activities or if you question these decisions you are blamed to be against your community. … Then they tell you that your land may be used for the development of the community, and that you can be given other lands. (Interview with a farmer in Ropi, 18 April 2012)
In the context of Siraro, where land shortage is an overriding problem and landlessness is considerable, threat of dispossession is a powerful tool to pressure farmers. While claims for ‘better development purposes’ or ‘community development’ are justified in relation to the narrative of E&F, similar re-elaborations are raised in relation to E&P. One farmer noted:
Local officials often ask us to attend a number of different activities for the development of our community such as maintaining roads and terracing … Meetings and work for the kebele is time-consuming. … They also want us to be productive and we have to follow the example of model farmers … . But the most important thing is doing what they ask. I want to keep good relations. If you don't participate they can blame you that you are not using your land productively. (Interview with a farmer in Ropi, 19 April 2012)
This quote shows that threat of dispossession is sometimes legitimated by arguing that land can be used in a ‘more productive manner’, or when it is not ‘properly conserved’. The discourse of rural development in Siraro is characterised by multiple and overlapping narratives of land reform which are selectively employed by local officials to legitimate the territorialisation of land administration in the rural milieu. This has important consequences for the relation between registration and security of tenure.
Land registration and security of tenure
The territorialisation of land administration is an important facet of the internal territorialisation of state rule in Ethiopia. Yet, it is important in setting the conditions in which land registration is implemented. Significantly, the case of Siraro shows that the question of ‘security of tenure’, rather than depending exclusively on the provision of a certificate, has to be contextualised into the broader and dialectical relation between state institutions and peasant farmers. Farmers' main source of ‘insecurity’ seems to be that local officials have now greater leverage over land allocation than before. One farmer noted:
I think the programme is not adding much to what is already in place. … We can use it in case of land dispute, but if in case of conflict with other farmers the certificate may be helpful in settling the dispute, if the government wants to take my land it will do it anyway. (Group interview with farmers on the topic of the Productive Safety-Net Programme, 7 December 2009)
Also, farmers make a distinction between the impact of land registration on the settlement of land disputes, for which they believe registration is positive, and on land productivity. Farmers are confident that registration will limit the amount of border disputes, thus establishing a more reliable system of conflict resolution. For instance, this will enable them to spend less time in local courts and more time farming their plots. Nonetheless, they also argue that land registration ‘is not bringing more land’, and they do not see how the programme will bring benefits in terms of increased productivity. One young farmer noted:
Three months ago they came here measuring my land and then I collected my certificate … I don't see how this piece of paper is going to make my maize grow faster. The amount of land I have is the same as before and fertilisers and seeds are expensive (Interview with a farmer in Ropi, 21 April 2012).
Access to agricultural inputs, resulting in constrained productivity, is one of the recurrent themes emerging from the interviews in Siraro. While a detailed discussion on how inputs are provided to farmers is beyond the scope of this paper, what is important to stress here is that ‘security of tenure’ is a context-dependent and subjective perception which is not only affected by the extent to which land is registered or not. The opposite is also true: security of tenure is only a single variable affecting productivity, land disputes and market formation.
The implementation of land registration shows that the very notion of security of tenure is context dependent and farmers constantly depend on their relations with local officials to secure their access to land and natural resources. Land registration, by sanctioning state ownership of land and providing farmers with land-use rights dependent on a host of conditions, has strengthened an already hierarchical system of land administration, yet increased the subordination of the peasantry to the state.
Land registration and state–peasant relations
Interactions between peasant and local officials can be analysed further by considering the role of the LACs, the local bodies in charge of issuing the certificates and retaining the power to settle land disputes during the demarcation process. In Siraro they partly replaced community elders dealing with minor land cases requiring a simple conciliation. This is a further example of how surveying and mapping land reconfigures patterns of political authority, and yet it is central to controlling people's activities and their access to and use of local resources. LACs seem to hold a certain degree of legitimacy among farmers. These bodies are composed of both elders from the local community and local officials from the kebele. While land registration has not involved major conflicts and ‘decisions are taken in a consensual manner,’ this is because the programme does not provide for a complete and clear demarcation of land borders. Yet, interviews show that farmers sometimes try to avoid registration or prefer keeping a certain degree of flexibility in their strategies of access to land. Where land borders between farmers were contested, registration either slowed down or was not implemented at all. One informant noted:
I know many cases where the registration was deliberately avoided by the LAC … The reason is that they wanted to avoid conflicts in the district. Land registration has to be a best-practice. … Some other times the border is deliberately left unmarked … or movable objects are used as landmarks. (Discussion with an informant, Siraro, 18 April 2012)
It seems that while land registration fits into a context where local officials are granted greater power on land allocation and use, this does not stop farmers from contesting the authority. This emerges in other interviews. One other informant noted:
Some farmers tried to refuse the LAC to register their land … They were afraid this would result in paying more taxes. … Often they gave up … They were told: either get the certificate or you will lose your use-right. (Discussion with an informant, Siraro, 15 February 2010)
This shows that the territorialisation of state power in Ethiopia is also contested, although farmers in the end tend to conform to the demands of the state. More broadly, farmers are frustrated by the implementation of land registration because they perceive it as the last in a long series of government programmes imposed from above. As noted by one farmer:
I am compelled to attend meetings at the kebele every month, sometimes every two weeks … , for instance they explain why land registration is important for the community … They mostly keep repeating the same things … but you have to go, you are forced to follow what they say. (Interview with a farmer in Siraro, 29 November 2009)
Farmers are confronted with the inescapable reality that they must engage with the state in order to secure access to land. Even if this is sometimes ‘a waste of time’, their access to resources depends largely on maintaining good relations with local officials. There is therefore little option but to comply with what they are told to do. Farmers remain committed to any decision made by the state. One informant observed:
Farmers are in a very difficult position … If they do not comply with what they are told to do, they can be blamed as being against the government … or the development of their very community. (Discussion with an informant, Siraro, 17 April 2012)
The territorialisation of state power provided by land registration generates two distinct dynamics. First, it strengthens the capacity of state institutions, particularly the kebele and wereda, to assert a monopoly on the means of enforcement to control people's activities effectively as required by state laws and policies. This confirms that control over land arising from registration of land rights does not only reflect pre-existing authority, but rather reconfigures the local arena where the rights to resources are negotiated and authority is produced (Lund and Boone 2013, 2). Second, as the state increases its capacity for enforcement, a counterhegemonic discourse performed by the peasants has emerged as a consequence. This recalls the distinction between public transcript and hidden transcript made by Scott (1990). The former describes the open interaction between dominant and subordinate groups, the latter the discourse taking place ‘offstage’, beyond the direct observation of the people in power. The existence of a hidden transcript highlights that the dominant discourse is also contested. As noted by Vandergeest and Peluso (1995), territorialisation is often unstable. In Ethiopia, the people most directly affected by the state's resource territorialisation have tried to undermine it by practices that ignore the state's goals. As a result, although the Ethiopian state has increased its capacity to enforce territorial controls, at a discursive level this project is also challenged to the extent that peasants try to minimise the contacts with the state, and conform only to the inescapable demands made by the local officials.
Conclusion
In this article I have argued that land registration, being a main facet of the contemporary strategies of internal territorialisation of the state in rural areas, reconfigures the exercise of political authority and yet opens up new spaces for negotiation and contestation of land and resources. In Ethiopia, land registration reproduces historical state–peasant relations characterised by subordination and domination, and yet serves to further extend the reach of the state in the rural milieu. Land registration is legitimated through a complex discursive repertoire, drawing on the narratives of E&F and E&P, and is an important facet of the broader project of rural development. In the eyes of the ruling elite, a strong ‘developmental state’ is necessary to promote ‘inclusive growth’ in rural areas, driving the country towards a ‘managed’ form of agrarian capitalism. Land registration is central to this project and, by sanctioning state ownership of land, is ostensibly intended to support the efficiency and the rationalisation of agricultural production, as well as the control of key economic activities and the people performing them. The case of Siraro shows at least two issues. First, the question of security of tenure has to be contextualised into the broader social and political relations between peasants and state. Second, although land registration and the territorialisation of land administration improve the capacity of the Ethiopian state to enforce territorial control in the rural milieu, the increased authority over land creates a counterhegemonic discourse, a hidden transcript performed by farmers, which potentially undermines the impacts of the programme.
More broadly, land registration has to be regarded as a deeply political project structuring the exercise of power and reconfiguring how authority over land and resources is claimed and legitimated by the different actors. The persistence of a ‘techno-managerial’ and depoliticised attitude of international donors – and particularly the World Bank – towards land reform in the contemporary global context misses the central point that institutional reforms shape power dynamics and inequalities in access to resources. In the past, this approach has often failed to generate significant developmental outcomes as it has ignored the social and political context of its implementation, and in particular has underestimated the importance of power relations (Berry 2002; Boone 2007; Peters 2009). The case of Ethiopia shows that land tenure is embedded in complex social and political relations and that each experiment of land reform has to be evaluated in light of national and local politics and of state–peasant interactions. ‘Security of tenure’ is dependent on multiple factors, and the provision of a land certificate – an ‘institutional fix’ – does not necessarily lead to growing security and socio-economic development.