Ethiopia: Reforming Land Tenure

An Ideological Battlefield: Land Policy in Ethiopia

Ethiopia is one of several post‐socialist countries undergoing an ambivalent transformation process towards some kind of capitalist economy. This applies in particular to land tenure regimes. After the fall of the derg (military socialist) regime of Mengistu in 1991, privatisation of farm collectives took place rapidly. Many international observers expected that in this process of ‘post‐socialist transition’, a move towards privatisation and registration of land titles would follow, which was regarded as a means to increase productivity of Ethiopia’s small‐holder agriculture. However, these expectations were soon disappointed. The Transitional Government of Ethiopia, despite being committed to the liberalisation of the economy, decided not to question state ownership (Adal, 2001a; Belay and Manig, 2004; Jemma, 2001). In its declaration on economic policy in November 1991 (TGE, 1991), it announced the continuation of the land policy of the derg regime. A referendum that was promised to decide the future of the land tenure system never took place (Jemma, 2001). Instead, the new constitution of 1995 approved and confirmed the state ownership of land in Ethiopia (FDRE, 1995).

Land policy has remained controversial in Ethiopia since the fall of the derg regime in 1991. Observers note that the debate is influenced by ideological considerations rather than being based on substantive empirical data (Adal, 2001a; Jemma, 2001; Hoben, 2000; Rahmato, 2004). Broadly speaking, we can distinguish two antagonistic political discourses on the land question: (1) the discourse of fairness and state protection that is arguing for state ownership, and (2) the discourse of privatisation and efficiency. The former is the position of the government that remains critical of privatising land holdings. The government expects privatisation to foster the concentration of land ownership in few hands by crowding out poor, destitute farm families from their land. Critics of the government’s position, such as the Ethiopian Economic Association (EEA), argue that state ownership of land prevents the development of a land market and thereby holds down productivity (EEA/EPRI, 2002). In addition, international policy discourses on decentralisation and community empowerment have infiltrated the political debate and subsequently land policy documents.

This article explores the two opposing arguments and how the principles of fairness versus efficiency are played out against each other. We then demonstrate how these arguments have influenced the formulation of policy and legal texts, in particular the federal and regional land policies with a particular view on the new Oromia regional land policy, namely the ‘Oromia Rural Land Use and Administration Proclamation No. 56/2002’ (RGO, 2002). We complement this with an analysis of ongoing policy implementation practices in Oromia Region. It is arguable that the regulative provisions are conceptually hybrid, accommodating both fairness and efficiency considerations, while implementation focuses on selected elements of these policy texts that promise to be popular with the rural peasantry.

The Debate: The Future of Land Tenure after the Derg

The dismissal of the derg regime on 28 May 1991 created a period of uncertainty about the future of land rights in Ethiopia (Adal, 2001a; Jemma, 2001). In this political void, both the option of privatisation of land ownership and the continuation of some form of state ownership were discussed (EEA/EEPRI, 2002; Rahmato, 2004). However, already in its November 1991 declaration on economic policy, the Transitional Government of Ethiopia (TGE) announced the continuation of the land policy of the derg regime (TGE, 1991). In 1995, state ownership of land was instituted in Ethiopia’s new constitution (FDRE, 1995), but despite the passing of the constitution, the debate over land ownership continued until the 2000 elections (Jemma, 2001; Adal, 2001a). However, the debate was partly suppressed thereafter with the Prime Minister publicly declaring land policy a ‘dead issue’ in meetings with donor agencies. After the 2005 elections (Hagmann, 2006), some of the proponents of privatisation have been jailed as members of opposition parties. The Ethiopian land policy debate focuses on the needs and economic pressures of the peasant economy of highland sedentary farming and is less concerned with the problems of pastoralist land tenure (e.g. Abdulahi, 2007) or the development of land markets in other parts of the country.

The Government’s Position

The government’s position builds on a social equity paradigm (Rahmato, 2004:1012) and tenure security considerations, which can broadly be subsumed under the fairness as egalitarianism principle. In this, state ownership is regarded as the most appropriate means to protect the rural peasantry from the negative side effects of market forces. In particular, the government claims that state ownership prevents (Adal, 2001a, 2002; Jemma, 2001; Rahmato, 2004) the accumulation and concentration of land in the hands of a small number of urban and bourgeois land owners, who acquire large tracts of land through distress sales by poor peasants, which would lead to:

The government maintains that its policy is tailored to equally distribute land to all who claim rights to ensure access for the needy, to provide them with the means to make a living as farmers in rural areas, and to protect them from selling or mortgaging their land, thereby safeguarding them from the grabbing hand of an urban bourgeoisie and rural elites. The government also explains that this land policy is aimed at preventing political unrest (MOIPAD, 2001).

The government’s position is supported by arguments pertaining to principles of fairness as historical justice. Some argue that privatisation of land would lead to a ‘total reversal of the 1975 agrarian reform measures’ (Mersha, 1998, cited in Adal, 2001a:57) and would negate the positive effects of the derg reforms for the rural peasantry by giving way to the eviction of the poor and the re‐emergence of feudal landlords (Bekele, 1997, cited in Jemma, 2001:48). This anti‐imperial narrative can be detected in the following quotation that places imperial landlords and capitalist farmers on equal par:

Overall, it could be argued, though, that the land tenure policy of the current regime should not come as a surprise. When the Tigray People’s Liberation Front (TPLF), the backbone of the current ruling regime (Abbing, 1997), fought its armed struggle against the derg regime, the TPLF, following its Marxist‐Leninist leanings (Gudina, 2003), implemented its own versions of land reform in the areas it controlled and conducted land redistribution schemes. There were hardly any ideological differences between the derg and TPLF regarding land policy (Beyene, 2004:60), since both gave only usufruct rights to land holders and created rural organisations ‐ the bailto were established by the TPLF and peasant associations (PAs) by the derg. The number of redistribution schemes in Tigray and the specific length of redistribution have varied across Tigray depending on who controlled which territory ‐ and redistributions normally were carried out when one party won over a specific locality from the other (Young, 1997). Hence, state control of land tenure and the policy of redistribution as a power instrument (as well as an expression of a radically egalitarian ideology) have been deeply entrenched in the historical experience of the TPLF leadership that is currently ruling Ethiopia (Young, 1998).

The Critics of State Ownership

Those critical of the government’s proposals on state ownership of land are most forcefully represented by some economists of the Ethiopian Economic Association (EEA/EEPRI 2002), donor agencies (e.g. IMF, 1998) and a number of Ethiopian and international researchers (Deininger et al. 2004; Rahmato, 1994, 2004). Most of those propagating privatisation of land ownership emphasise efficiency considerations (Deininger et al. 2004; EEA/EEPRI, 2002). Broadly speaking, it is argued that state ownership of land yields negative effects on land productivity and therefore produces lower efficiency levels than would be achievable with the working of a private land market. In particular, those in favor of private property rights assert that state ownership provides barriers to full‐scale efficiency, because:

Rahmato (1994, 1999, 2004) has been a particularly forceful critic who does not employ the state ownership vs. privatisation dichotomy, but suggests a third way of community or ‘associative’ ownership of land (Rahmato, 1994:13‐15). He argues that state ownership would have significant detrimental social and environmental effects, thereby creating the ‘conditions for a Malthusian disaster’ (Rahmato, 2004:13). He sees insufficient incentives to invest in sustainable land use practices and to adopt modern technologies due to tenure insecurity and small plot sizes coinciding with an increasing rural population and inter‐generational struggles over access to land. However, these arguments are seldom based on comprehensive empirical evidence.¹

In addition, some critics of state ownership have also brought forward fairness as historical justice considerations. Ethiopian land rights have traditionally been tightly linked with the exercise of power over the rural peasantry ‐ in the imperial period where large parts of the peasantry were denied land ownership and lived in precarious tenancy relations as much as during the derg regime, where collectivization and the increasingly tight grip of the state on the local peasant associations meant increasing state control of farmers. Adal (2001a:58), therefore, argues that privatisation of land holdings means the fulfillment of peasants’ self‐determination as part of the ‘unrestricted economic rights’ any farmer should enjoy. Several authors ascertain that since the derg reforms in 1975, peasants have become the ‘tenants of the state’ (Rahmato, 1992; Abebe, 1998) and therefore have been deprived of their (human) right to freely decide what to do with their farm land.²

Policy Texts: Federal & Regional Land Proclamations

This section outlines how the two discourses on fairness and efficiency have been translated into new legal texts and codes that were designed after the derg regime. We focus on provisions made in the 1995 Constitution and in the Federal Land Administration Proclamation of 1997 and then consider the legal implications from the regional land proclamation designed by the Oromia Regional Government in 2002. The 1995 Constitution approved state ownership of land as the statutory framework, while the 1997 Federal Land Policy provided space for some differentiation of constitutional and federal rules by giving each regional state the right to formulate a regional land policy. In the process of formulating these laws and regulations, however, the principles applied were not as clear‐cut as the antagonistic policy debate seemed to imply. As a result, these legal texts have become conceptual hybrids: efficiency, fairness as egalitarianism, and fairness as historical justice principles are all reflected in various provisions of the federal and regional documents.

Federal Legislation & Policy

Article 40 of the 1995 Ethiopian constitution states that:

Many scholars have questioned how far the 1995 constitution provided any change of land rights compared to the land reform proclamation of 1975 (Belay and Manig, 2004; Ege, 1997; Nega et al. 2003; Rahmato, 2004). There are, nevertheless, some notable differences between the formal legal rules of 1975 and 1995; namely the 1975 proclamation prohibited the lease of land and the hiring of labour and capped the maximum land holding per individual to 10 ha, which are absent in the 1995 document (Adal, 2001a:56). Furthermore, Article 40, section 7 of the 1995 Constitution specifies the rights to compensation payments for investment on land in case the ‘right to use expires’, which were new provisions:

The subsequent ‘Federal Rural Land Administration Proclamation’ (Proclamation 89/1997) was enacted in July 1997 (FDRE, 1997). While the proclamation transfers the authority for land administration, including rights to distribute land, to the regional governments and vests them with the power over the ‘assignment of holding rights and the execution of distribution of holdings’ (Article 2, 6) it clearly limits the space for regional administrations to exercise those powers. The preamble of the proclamation confirms the constitutional claim that land ownership rights are vested in the state and it rules out the possibility of mortgaging and selling land. It also continues to hold the redistribution of land holdings by the regional governments to be legal. Article 6 authorises regional governments to implement periodic adjustments of individual farmers’ land holdings and to demarcate land for communal use (grazing, residence, local forests, social services).

In marked difference to the derg rules, Proclamation 89 confirms some important provisions of the 1995 constitution, namely the right to receive compensation for labour and capital investments on land that is redistributed (Article 6, 7‐12) and it clarifies the user rights of land holders, including inheritance and leasing (Article 2,3). The compensation right is an important change compared to the derg rule. At the same time, there are serious omissions that leave tenure security unclear, in particular the proclamation does not state the duration of usufruct rights for land holders. Broadly speaking, while the 1995 constitution and the Proclamation 89 largely confirm state ownership of land as a continuation of derg policies, they also provide some specifications that seek to take account of the necessity for a rural land and labour market to emerge. These rules therefore combine some form of efficiency considerations (allowing labour markets and leasing) with the fairness as historical justice principle (the right to obtain land without payment), the latter requiring that the state keeps overall control of distributing access to land (since the market does not necessarily ensure access for all who need it).³

Regional Land Policy Formulations

Several regional governments have made use of the powers vested in them in the 1995 constitution and Proclamation 89 to formulate their land policies, among them Tigray Region (1997, amended 2002), Amhara Region (2000), Oromia Region (2002) and Southern Nations, Nationalities and Peoples Region (2003). Since, according to the constitution, regional land policies need to be in accordance with federal law, all regional policies validate state ownership of land and farmers only receive usufruct rights to plots of land without transfer rights, such as sale or mortgage. All regional proclamations confirm the right to lease out land, although most regions restrict the period of lease. At the same time, there are also marked differences across the regional policies with regard to tenure security (Rahmato, 2004), lease and inheritance rights as well as the right of the regional government to redistribute land. These differences in regional legislation imply that tenure security will markedly differ in the various regions of Ethiopia, in particular on the question of redistribution of land plots, on the definition of user rights and on certification.

Oromia Regional Land Policy

In this section, we discuss the land policy of Oromia region as its legislation has been considered to be the most ‘progressive’ (for neo‐classical economists) in terms of incorporating efficiency considerations. The legislative text is a masterpiece of dialectical reasoning as it translates the federal imperative of state ownership with measures to improve ‘subjective’ tenure security of land cultivators, i.e. the perception of peasants that their use rights are secure. In this sense, the ‘Oromia Rural Land Use and Administration Proclamation 56/2002’ (RGO 2002) grants higher levels of tenure security than the other three regional policies (Rahmato, 2004), because it rules out redistribution of land plots (RGO, 2002, Article 14,1). In addition, the proclamation grants ‘lifelong usufruct rights’ (Article 6,1) to agricultural land ‘free of payment’ to all male and to female residents whose livelihoods depends on agriculture (Article 5,1) and determines payment of compensation in case land is used for major communal purposes. It grants lifelong use rights independent of marital status or locality of residence.

However, there are three important restrictions formulated in the proclamation that allow the state to expropriate land use rights from plot holders and therefore attenuate the lifelong usufruct rights: (1) Article 6,4 grants the right of expropriation if land is required for ‘more important public uses’, but the rights holder needs to be compensated and can remove investments thereon; (2) Article 14,4 specifies that ‘irrigation land’ is excluded from the prohibitions of land redistribution, which is allowed subject to the ‘participation and consensus of the user community’; and (3) Land can be expropriated if needed for irrigation infrastructure. The ‘Oromiyaa Rural Land Administration and Use Regulation No. 39/2003’ (RGO, 2003, hereafter referred to as the regulation) further specifies that if land users fail to use their land in every production season (except in the case of restoring fertility), the land use rights can be terminated (Regulation, Article 3,5). According to the regulation, Article 22, 1, after a period of three years without cultivation, the land will be expropriated, in the case of irrigated land; this can be applied already after two years. The proclamation also imposes a number of obligations on land management practices (cultivation of erosion‐prone areas, planting of specific tree species, and preservation of water conservation structures) as a condition for the usufruct rights.

The proclamation envisages legal provisions for a land titling or certification programme, which would grant lifelong certificates of holding rights (Article 15, 1). This titling is primarily aimed at increasing ‘subjective’ tenure security ‐ it does not provide the plot holder with rights to mortgage or sell the land. Certification, however, becomes an issue when inheritance is concerned. Since land holders are restricted in whom they can bequeath their usufruct rights and in the minimum plot size, heirs of small plots may not receive individual certificates, but may need to cultivate the land in a way to maintain the integrity of the plot. There are also provisions for land leasing, which is a significant change compared to the derg’s land policy, but the leasing period is restricted: rights holders are only allowed to rent out up to half of their land and the leasing period may not exceed three years under non‐mechanised production methods.

The Oromia regional land policy integrates efficiency and fairness principles, although the latter are clearly more dominant (Table 1). Efficiency (understood as productivity per land unit) is expected to rise through measures that seek to increase ‘subjective’ tenure security (certification, lifelong tenure security, rights to compensation for investment on land, minimum plot size), although these are subsequently attenuated with further prohibitions. Fairness as egalitarianism is reflected in the requirements that farmers should enjoy equal access to land, that land redistributed in irrigation schemes should be allocated in equal shares and that heirs should receive equal shares of land. Fairness as historical justice figures prominently through regulations that prohibit the sale of land (to avoid buy out of destitute farmers), that grant everybody who needs agricultural land the principle right for a plot of land and that restrict access rights to residents of the kebele (to avoid urban elites to acquire property rights as absentee landlords).

According to the proclamation, the average male land rights holder in Oromia enjoys a larger bundle of rights than under the derg and can be regarded to be proprietor of the land (see Table 2 opposite), but this only applies to male farmers who hold non‐irrigable land of a size larger than the minimum plot size. Some provisions in the proclamation and in the regulations further attenuate the bundles of rights and effectively undermine the granting of lifelong tenure (e.g. by defining management obligations). These prohibitions allow the land administration an enduring influence on land distribution and land rights.

Multiple Practices

In the following section we provide some insights into the early implementation phase of the Oromia Land Use and Administration Proclamation (RGO, 2002) in Alemaya district (woreda). We use qualitative data from interviews with farmers, bureaucrats and policy makers in the field of land administration at kebele (village), woreda (district), zonal and regional levels.⁴

By the end of 2004, the implementation process, which had started in 2003 was far from being completed. The regional government opted for an implementation in different phases and started with some selected provisions only. This early implementation phase was concerned with a rapid distribution of temporary use rights certificates, which bureaucrats thought to be popular among the peasantry.⁵ It aimed at increasing ‘subjective’ tenure security, i.e. the perceptions of farmers about the security of their property rights to their land plots, and their satisfaction with the ruling regime’s policies. Measures that were likely to attenuate use rights and thereby create dissatisfaction among the electorate (such as management obligations and prohibition of the use of particularly endangered areas for agricultural purposes) were postponed considering the upcoming elections in 2005. By 2006, it was expected that measurement of the agricultural area and the distribution of preliminary certificates to all farming households in Oromia region would be finalised. Regional policy makers justified this differentiated implementation practice of the provision (that was not foreseen in the provision itself) with two arguments. First, implementing those provisions aiming at nature conservation could result in the loss of agricultural areas ‐ a measure that could be unbearable for a destitute rural peasantry facing high levels of food insecurity; second, the lack of accurate data on land holdings was considered to impede the distribution of permanent land use certificates.⁶ The government, therefore, linked the implementation of the full set of provisions to two preconditions: A scientific cadastral survey to improve the database for permanent certification of land plot surveying and the exploration of ‘alternatives’ to farmers to allow them a more ‘profitable’ use of the remaining agricultural area in case of land loss due to conservation measures. The implementation process was also hampered by incomplete rules for implementation of a number of provisions, such as a lack of definition of enforcement mechanisms.⁷

By the end of 2004, the bureaucratic set up was largely in place. The Environmental Protection, Land Administration and Use Authority (EPLAU) responsible for measurement and registration as well as distribution of land holding certificates existed at regional, zonal, woreda, kebele and village level. However, insufficient allocation of financial means lead to understaffing of the woreda authority, and inadequate staff training made knowledge gaps about the regulative provisions pertinent at woreda, kebele and village levels.⁸ At kebele level, Land Administration Committees were formed consisting of (unpaid) village representatives and kebele administrators. These committees are responsible for dissemination of information on the proclamation to the peasants. Although some committee members were trained,⁹ bureaucrats and committee members we interviewed and who had participated in briefings were unable to outline the differences between the previous and current land rights and often had only rough ideas about the content of the new proclamation. Not surprisingly, many farmers were not aware of the forthcoming certification of their holdings and provided very different interpretations of the provisions they had heard of. None of the farmers or administrators was informed about the temporary character of the certificates to be issued, nor did they know about the management obligations implicit and explicit in the new provisions.

Peasant Perceptions: Subjective Tenure Security

In our study of Alemaya woreda, especially the area around Lake Alemaya, subjective tenure security (i.e. the perception of tenure security) among peasants has been high. This can be explained by looking at the process of agrarian policies during and after the derg. In the collectivisation process after 1975, the small number of the district’s tenant cultivators had received use rights. It was the formation of Agricultural Producer Collectives (APCs) that brought about significant changes in land allocation.¹⁰ After the fall of the derg these collective farms, where farmers in each of the kebele had pooled their land, were dissolved and the previously existing farm boundaries were re‐established. In this process, previous ownership was reestablished and all land was given back to previous owners or their heirs.¹¹ This resulted in numerous boundary disputes among neighbours who reclaimed their rights to particular segments of plots. Most of these boundary conflicts were solved by groups of elder men.¹²

At present, male farmers generally acquire use rights to land through inheritance from fathers, a practice which was only abandoned during the existence of the APCs until they were dissolved in 1991. Each son has the right to an equal share of the family land. Recent severe land scarcity (non of the respondents had use rights to more than 0,5 ha of land) and the inability of farmers to support a family by means of working their small plots of land did not change this practice. However, farmers were not (yet) informed about this provision. They, therefore, planned to pass on their land by sharing it among male children. Tenure insecurity was not a major concern among those peasants already in possession of land plots: all (male but also female) respondents expressed that they did not expect any redistribution that could threaten usufruct rights to their specific plots, neither had they heard of other farmers losing land through expropriation or redistribution.¹³ At the same time, bitter competition over access to land was common among the younger generation that will inherit plots too small to support a family. According to village elders the majority of land disputes shifted from conflicts between neighbours after the dissolution of the APC, to previously rare land disputes between fathers and sons, after the fall of the derg. Therefore, land conflict between brothers, but also conflict between brothers and sisters, increased significantly.¹⁴

Senior bureaucrats and policy makers in the region tend to replicate neo‐classical efficiency arguments linked with private property rights to land and often anticipate that an increase in ‘subjective’ tenure security among rural peasants (in a way introducing something that feels like private ownership, but cannot be private ownership because of the constitutional provisions that are politically sacrosanct) would increase productivity in the agricultural areas.¹⁵ In the case of Alemaya woreda, tenure security does not seem to be an issue at all. Peasants considered their use rights to be secure. For them, issuing certificates is primarily useful to reduce boundary conflicts among neighbours rather than increasing overall plot tenure security. Labour intensive agricultural production systems already prevail in Alemaya district and it is therefore very unlikely that farmers will intensify productivity after certification, but investment on boundary protections, e.g. planting of hedges, might increase. Rather, farmers considered the restricted access to adequate irrigation technology and chemical farm inputs such as fertiliser as impediments to increasing plot productivity.

Most farmers had little knowledge about the regulative provisions on inheritance and minimum plot size ‐ the provision aimed at restricting subdivision of land to avoid plot fragmentation (the latter is considered to hamper land productivity). Neither were local bureaucrats aware about the details of the provisions and how to implement them. It is unlikely that farmers will comply with these practices, even when they are known. During current plot measurement exercises, many farmers did not inform authorities about the actual distribution of land among family members¹⁶ and land sharing among heirs is likely to continue in the foreseeable future.

Women’s Rights in Effect

The proclamation explicitly emphasised women’s rights to land. In Alemaya woreda, women are not generally excluded from rights to land, even though in times of land scarcity, their traditional rights had gradually been attenuated. Traditionally, women held the right to a share of land, approximately one‐third of the plot size inherited by sons. However, such practices had been abandoned in the wake of increasing land scarcity and land transfers to daughters have become rare if nonexistent. Only widows without mature sons could hold rights to land, but these rights are only considered temporary until the son has matured.¹⁷

Many farmers emphasised women’s equal legal status, but a number of implementing practices of the regional proclamation foreclosed that this would come into effect. First, the distribution of one certificate per household (which is prescribed by regulative provisions); second, the abandoning of the provision which entitles each woman in polygamous marriages to her own certificate; and third, the prescription that made traditional authorities the first legal level for disputes over land. The role of traditional courts is ambivalent in this process. Many members of traditional local authorities know very little about the current regulative provisions with regard to land property rights and differences in traditional rules and norms and the provisions laid out in the proclamation continue to persist. In addition, bribery is common in courts hampering the neutrality of their rulings. This means that even though traditional norms may entitle some women to specific rights, this does not ensure that courts rule accordingly.¹⁸ Furthermore, costs to access official courts at woreda level are often prohibitively high. This potentially excludes poor people from formal jurisdiction. A kind of legal pluralism has emerged where, depending on availability of financial means claimants have access to different sources of rights to enter into ‘forum shopping’ (Benda‐Beckmann, 1981).

Populist Implementation Strategy

We consider the bureaucracy’s implementation practices as populist, because the hasty distribution of temporary user right certificates based on an estimated size of holdings is unlikely to improve agricultural productivity, incentives for on‐farm investment and sustainable land management, all aims mentioned in the regulative provisions. The certificates currently distributed only provide a rough estimate of the plot size. The size is usually measured with a rope and the position of plots is demarcated by the names of neighbouring landholders. Apparently, the preliminary certification procedure aims at collecting data on land use,¹⁹ and is tailored towards safeguarding peasants’ support to regional (land) policy. At the same time, implementation of restrictive obligations on land use rights and traditional inheritance rules have been postponed, because they were suspected to potentially cause distress and annoyance among the rural peasantry.

At the end of the day, this populist implementation strategy provided additional burdens on the local bureaucracy, in particular the distribution of ‘preliminary’ certificates, which are to be followed by permanent certificates and to be updated on a regular basis. Local administrators did not know how to correctly distribute certificates in case of polygamous marriages, nor were they clear about how to correctly handle the certification of plots of minimum size which were prohibited to be shared between heirs. Due to a lack in adequate legal provisions, the updating of use rights certificates is highly unlikely since practical advice how to certify land rights to parts of plots which fall below the minimum size is unavailable even though at the same time the largest part of land holdings in the research area is already too small for subdividing land plots upon inheritance as would be required according to the provisions in the proclamation.

Revisiting the Battlefield

The currently available empirical evidence of the impact of different tenure regimes on land use and on farmers’ responses to different land policies and land tenure systems in Ethiopia is far from conclusive (Haile Gabriel, 2004; Adal, 2001b; Aredo, 1994; Beyene, 2004; Bogale et al. 2006; Bogale and Korf, 2007; Benin and Pender, 2002; Deininger et al. 2004; EEA/EEPRI, 2002; Gavian and Ehui, 1999; Gbremedhin et al. 2003; Holden and Yohannes, 2001; Nega et al. 2003; Omiti et al. 2000; Rahmato, 1994; Teklu, 2003). Hence, it is premature to develop universal policy prescriptions. Arguably, recent scholarship on land property rights (Bromley, 1991; De Janvry et al. 2001; Larson and Bromley, 1990; Schlager and Ostrom, 1992) acknowledges that privatisation and individualisation is not a priori the most efficient means of achieving tenure security. Several scholars writing on African land tenure systems have argued that land titling and individualisation of ownership rights was found to have little or no effect on investment and farm income (e.g. Carter and Wiebe, 1990; Gray and Kehane, 1999; Place and Otsuka, 2001).

This is not surprising in view of the contradictory bureaucratic practices that go hand in hand with the formulation of policy texts. Our observations from Alemaya woreda indicate how the populist implementation strategies of the Oromia Land Use and Administration Proclamation mainly followed the interest of the ruling regime in improving its reputation and support base among the rural peasantry, but may yield ambiguous impacts on pressing problems in the Ethiopian agricultural highlands, such as resource degradation, land scarcity and limited investment incentives. It appears that while the proclamation document ‐ the policy texts – provide a ‘dialectical’ solution to the efficiency‐fairness dichotomy and thus pleases donor agencies, the local bureaucrats pick those aspects that promise to satisfy the fairness principle and ensure peasant support for the ruling regime in upcoming elections.

For the Ethiopian ruling elite, it is important to safeguard the narratives of the fairness principles that have been shaped by the experiences of the injustices of the imperial land tenure regime where land was concentrated in the hands of absentee landlords, tenure was highly insecure and arbitrary evictions posed serious threats to tenant farmers (Cohen and Weintraub, 1975; Hoben, 1973; Jemberre, 2000; Joireman, 2000; Pausewang, 1983; Rahmato, 1984). State ownership as principle institutes fairness as historical justice ‐ to avoid those debilitating circumstances of the imperial era to come into being again. Fairness as egalitarianism is a further principle in continuity with the derg regime. This principle says that every farmer ought to have equal access to cultivation land without payment. At the same time, efficiency considerations cannot be completely discarded, because only through productivity enhancement can the rural peasantry in the highlands be expected to achieve food security on household level.

But it seems that the ruling elites are not willing to compromise the two fairness principles substantially, even though this may come at the expense of efficiency considerations. Arguably, in the dominant narrative, those fairness principles are to be secured by a benevolent state. The image of the benevolent and paternalistic state implicit in these fairness arguments ‐ the state as the arbiter to ensure equity ‐ is clearly a continuation of the derg regime’s ideological commitment and its fairness as historical justice arguments. If there is no state control on land, then this paternalism is stripped of one of its most formidable power resources.

Agricultural economists have published numerous studies in which they set out to demonstrate the expected positive effects of higher tenure security on farmers’ incentives to invest in land quality thereby increasing land productivity and sustainability of land use in rural Ethiopia (e.g.Nega et al. 2003). They may continue conducting even more studies to demonstrate the investment incentives of privatizing land. Their ambition to change Ethiopian government policy, however, is likely to fail. The Ethiopian government may respond to such pressures from international donors by implementing some kind of land certification as is currently under discussion and preparation at the federal level and is part of most of the regional land policy documents discussed in this article. But it is very unlikely that the Ethiopian government departs from the dependence path in rural politics and the practices to govern the rural populace and gives up its most precious power resource in the rural realm ‐ the power to distribute land.