Review of Landholding Systems and Policies in Ethiopia under the Different Regimes

Review of Landholding Systems and Policies in Ethiopia under the Different Regimes Yigremew Adal December 2002 EEA/Ethiopian Economic Policy Research Institute Working Paper No 5/2002 CONTENTS 1. INTRODUCTION

2. LAND TENURE UNDER THE IMPERIAL REGIME 2.1. Major types of land tenure 2.1.1. Rist /kinship tenure 2.1.2. Private tenure 2.1.3. Church tenure 2.1.4. Government tenure 2.2. Land tenure problems 2.2.1. Tenancy 2.2.2. Tenure insecurity 2.2.3. Diminution and fragmentation of holdings. 2.2.4. Land concentration and underutilization. 2.3. Summary 3. LAND TENURE UNDER THE DERG REGIME 3.1. The 1975 land reform and abolition of diverse tenure 3.2. Land tenure problems 3.2.1. Diminution and fragmentation of holdings. 3.2.2. Tenure insecurity and its impact. 3.2.3. Inefficient allocation of land. 3.2.4. Inadequate land administration. 4. THE CURRENT LAND TENURE SYSTEM AND THE LAND DEBATE 4.1. The current land tenure system 4.2. The current debate on land tenure 4.2.1. The debate 4.2.2. Some queries about the debate 4.3. The case of rural women 4.2. Summary of the current land tenure situation 5. SUMMARY ANALYSIS OF THE LAND TENURE SYSTEMS UNDER THE DIFFERENT REGIMES 5.1. Access to land. 5.2. Tenure insecurity. 5.3. Diminution and fragmentation of holdings. 5.4. Inefficient allocation of land. 5.5. Problems in land administration. 5.6. Rural women. 5.7. The land debate and international experience. 6. CONCLUSION 7. ISSUES FOR INVESTIGATION REFERENCES

1. INTRODUCTION A thorough analysis and evaluation of the past performance of Ethiopian agriculture recently undertaken by the Ethiopian Economic Association (Befekadu and Berhanu 2000) reveals that, despite the numerous commendable initiatives and measures undertaken by the government, the performance of the agriculture sector remained poor. Many elements of the existing land tenure system, such as declining farm size, tenure insecurity, and subsistence farming practices, are identified as part of those causes of the poor performance of the sector. It was also underlined that the sector needs fundamental transformation at the center of which lies the issues in the land tenure system. The Ethiopian Economic Association/ Ethiopian Economic Policy Research Institute considers that the rural land tenure system is an important factor in shaping the socioeconomic structure of the rural sector. It also believes that, one way or another, the land tenure system defines how individuals, communities, and government relate to land and its administration and the social and economic values and benefits associated with it. It is also believed that the land tenure system is more critical among those problems of the agricultural sector because it provides the fundamental ground rule under which all economic agents operate regarding land. As a result, the land tenure issue has attracted widespread attention and debate among policymakers, government and non-government sections, the private sector, researchers and the community at large. It also remained to be a challenge that needs to be addressed based on a comprehensive and thorough research and analysis. It was with such understanding that the EEA/EEPRI decided to launch research on the current land tenure systems and its implications on the overall performance of the agriculture sector in the country. The overall objective of the research is to undertake a thorough and comprehensive study on issues that are related to the land policy and its implications on agricultural performance and thereby contribute towards policy debate, processes and policy formulation in the country. One such study being undertaken is on Review of Theories on Land Tenure and Country Experience, which is published in Working Paper No. 4/2002 of EEA/EEPRI. This study explores land tenure issues- theories, country experiences, and the Ethiopian case during the period pre-1975 to present and raises fundamental issues for further investigation. This part (Part II) deals with the review of landholding systems and policies in Ethiopia under the different regimes and the current debate on rural land tenure. The major objective of this part of the study is to review the existing literature on land tenure and identify the gaps that should be addressed by the proposed study.

The report is organized as follows. Following an Introduction, Ethiopian land tenure under the Imperial regime is discussed. Section Three deals with the land tenure issues during the Derg era. Section Four discusses the present land tenure system. Section Five is about rural women and Section Six discusses the current debate on the prevailing land tenure system. This is followed by Section Seven, which is a summary analysis of the whole land tenure issues relevant to the proposed study and gaps are identified for further research. Section Eight is the conclusion while Section Nine provides a brief list of those outstanding issues of the whole discussion that need further investigation for better understanding and clarity and this is the fundamental objective of the whole exercise in this part of the report. 2. LAND TENURE UNDER THE IMPERIAL REGIME 2.1. Major types of land tenure It is generally acknowledged that the pre-1975 land tenure system in Ethiopia had been one of the most complex tenure systems in the world and had not been thoroughly studied (Dessalegn 1984; Cohen and Weintraub 1975; Gilkes 1975; Dejene 1999) [1]. The country s geographical, ethnic and cultural diversity and its historical background were mentioned among those factors that produced highly differential forms of land utilization and ownership. Such complex nature of that tenure is also noted as playing a major part in hindering any serious progress towards a reform of the system. These difficulties might have also resulted in a variety of classifications and approaches used to describe the then land tenure system of the country. Among others, rist/kinship, communal, diessa/village, private, state, church, and other land tenure designations were used. For reasons of simplicity and to make the discussion congruent to the most common classification system, rist/kinship, private, church, and state holding systems are used here. 2.1.1. Rist /kinship tenure With some differences, the most prevalent tenure in the northern part of the country (Eritrea, Tigre, Begemidir, Gojjam, and some parts of Shoa and Wollo) was the rist/ communal or kinship tenure. Within this tenure system there were two variants of land rights rist and gult. According to Cohen and Weintraub (1975), rist is the right to claim a share of land based on kinship to a historical ancestor held in common with other rist holders. Briefly, the land in rist areas is divided into a multiplicity of geographical units originally held by a founding father. Those who can establish kinship through either parent may enter a claim to a share of the land in a unit from elders controlling the allocation of land held by what is in effect a descent corporation. That is, a person can inherit rist rights from either parent because ambilineal descent principles prevail in rist areas. Moreover, customary laws require that rist rights be honored if proof of kinship can be established. In theory and generally in practice rist rights are use rights and land under rist rights cannot be owned or alienated.

Hoben (1973) explains that gult rights over land were given to members of the ruling elite as a reward to loyal service to their lord and to religious institutions as endowments. The individual or institution that held land as gult had the right to collect taxes from those who tilled it. They also had judicial and administrative authority over those who lived on it. Rist and gult do not refer to different types of land but distinct and complementary types of land rights. Normally they are found over the same land. A single estate of gult land comprising a few square miles includes within its boundaries tracts of fields held as rist by a number of farmers. The gult holder might also hold some fields as rist within his estate of gult land (and outside it as well). Although ristholders were not tenants, their hereditary right to use the land was conditional on meeting taxes and service obligations associated with those rights. Failure to meet such obligations might lead to displacement of the ristholders. Mesfin (1991:79-80) also noted that gult refers to the right to taxes on land and not rights to land per se. Gult right does not extend to ownership of land but is limited only to taxes. When gult right is hereditary it is called riste-gult. Although hereditary on ambilineal basis, the rist system could not guarantee equal access to land to all dwellers of the area. Individuals access to land was highly determined by their social status (Hoben, 1973). Practical issues of politics and status enter into the recognition of the rights keeping the whole process workable. Specifically, they influence the allocation process in such a way that a claimant is likely to succeed only where he has close kinship to the descent corporation and lives near the land he is claiming. The rist system did not also guarantee access to land to some people. Hoben, (1973) for instance, mentioned that in Gojjam this category of people includes smiths, weavers, and tanners. 2.1.2. Private tenure To some extent this landholding pattern was recognized as the most dominant pattern affecting over 60% of Ethiopia s peasants (Kidane, 1990) and prevailing in the area where 65 percent of the Ethiopian population lived (Cohen and Weintraub, 1975). This tenure system was generally found in the southern and southwestern parts of the country. According to (Cohen and Weintraub, 1975: p. 35) the private tenures were created when the crown confiscated land conquered by its armies and granted vast blocks to a wide range of people and institutions. Grants were made to soldiers, northern civil servants who came to administer the new areas, peasants moving south because of land pressure in the north, local tribes that did not resist the conquest, local village and clan chiefs to gain their support, church officials and institutions to facilitate the expansion of the Coptic religion, and a host of central and provincial elites close to the crown. On the whole, these large blocks of territory were located in the provinces of Arussi, Bale, Gemu-Gofa, Illubabor, Keffa, Wellega, and some parts of Shoa and Wollo. Dessalegn (1984) also notes that lands under private tenures were originally expropriated from peasants of the crown. All unoccupied land in these areas was also considered to be state property and it was distributed to men of influence and power in the state apparatus. Much of the land thus acquired was subsequently converted into

private tenure, and Haile Selassie s government accelerated this process by its policy of imperial land grants and by encouraging holders of state tenures to convert them into freehold. According to the above sources, land under private ownership could be sold or exchanged without any restrictions except those provided by law. However, it has been noted that the Ethiopian private tenure was different from what the term generally signifies. According to (Dessalegn, 1984) lands under private tenures were private not in the strictly capitalist but in the specifically Ethiopian sense of the term. What the state had granted (and virtually all land under private tenures was originally state property) the state could take away, and in so far as the authority of the state was concerned, the sanctity of private property was not recognized in principle or in fact. Cohen and Weintraub (1975) also stated that The use of the term freehold is justified by the fact that occupants rights of permanent possession are nearly absolute. But they pointed out that still it was important to note that in general Ethiopians view land in terms of rights in the holder rather than possession of physical territory. 2.1.3. Church tenure The Ethiopian Orthodox Church used to be an important landholding entity during the pre-revolution period. However, the extent of church holdings was not clearly known (see Dessalegn, 1984; Cohen and Weintraub, 1975; Kidane, 1990; Aba Gebre-Egziabher, 1970). Kidane (1990:51) noted, The exact amount of church lands was never really determined. He suggested that the complexity of the forms of church ownership, the decentralized nature of ownership of church lands and also the secrecy of the church's central treasury were factors responsible for the lack of information about the amount of church holdings. Although church holdings were found both in the south and northern parts of the country, much of the lands were in the south. Estimates of the amount of land held by the church varied from 50 % (Aba Gebre-Egziabher, 1970, mentioning foreign sources) to 5% (Cohen and Weintraub, 1975) of the total land area of the country. Dessalegn (1984) doubted those figures and put his estimate as not exceeding 10-12%. For the most part church land had been obtained through grants from the crown. However, aristocrats and provincial gentry had also transferred land to local institutions and officials. In general, land was held by the church, by individuals obligated to provide clerical services for the community, or by peasants who paid tribute to the church rather than taxes to the state. Church lands were held as semon land and gult land. According to Aba Gebre-Egziabher (1970:248), semon lands were those lands whose use was placed at the disposal of the functionaries of the church by the church herself in lieu of giving them cash salaries and in appreciation of their services to the church in their respective region, while gult lands of the church were those lands the use of which was given to the administrators of churches and the high priests. In addition to renting land, the church used to collect tribute or taxes. Regardless of such large possessions, however, what the church had over those holdings was usufruct rights.

2.1.4. Government tenure Like the other holding cases, the amount of government land had never been determined with precision. MLRA (1971:11, 13) estimate shows that government land in ten noncommunal provinces has a share of 28.4% and 22.4% measured and unmeasured lands respectively. Cohen and Weintraub (1975:45) stated that 46.6% of the total land of Ethiopia and 11.8% of its agricultural land was held as government land. It was also stated that (Kidane, 1990) the government used to assume that all nomadic lands belonged to it and had largely appropriated nomadic pasture lands. Kidane pointed out that the local people bitterly opposed the granting of these lands to settlers and commercial farmers from outside the areas and claimed large parts of the government holdings, especially in the nomadic areas of the Awash valley and in the Setit Humera regions. Palace land, gibretel land, maderia land, and mengist land were those commonly cited arrangements under government tenure ( Dessalegn, 1984; Cohen and Weintraub, 1975; Kidane, 1990; Gilkes, 1975). These government lands included arid range lands in the lowlands of eastern and southeastern Ethiopia with grazing potential only and occupied by nomads; arable land to some extent occupied by nomads, generally found in remote areas with poor precipitation and a high incidence of animal and human disease; and those unoccupied arable lands distributed all over the country. 2.2. Land tenure problems The major problems of the pre-1975 land tenure in Ethiopia cited by different studies include exploitative tenancy, land concentration and underutilization, tenure insecurity, diminution and fragmentation of holdings. Fesseha (1970:135-140), for instance, concluded that the presence of large idle lands reinforced by high rate of absentee ownership, high rate of tenancy, insecurity of tenure, operation of sub-economic sizes of farms and lack of relevant supporting institutions constituted principal bottlenecks that kept agriculture in a subsistence and generally retarded state. 2.2.1. Tenancy In a seminar where the land tenure issues were seriously discussed around the eve of the 1974 revolution it was concluded that: It was observed that the problems of agricultural tenancy are so acute that a remedy must be provided promptly. In this connection it was indicated that a large proportion of the farming population works and lives on the land of landlords paying high rents and receiving low yields. Similarly it was pointed out that tenants under the existing landlord-tenant agreements have no security of tenure (Tamirat, 1970:480). It can be argued that thought there were some differences among writers on the scope and nature of tenancy at that time it seems that there is general agreement that tenancy existed in both the south ad the north but it was much more widespread in the south than in the north. As per the findings of the 1974/75 survey by MoA (1975a:60) out of the total holdings 36.4% were totally cultivated by tenants while they partly owned another

15.3%. The report stated that the condition of the tenants was the most prominent problem of the land question in the country. Tenancy was also found more widespread in the central, western and southwestern regions where private tenure was most prevalent. In these provinces, the size of tenant population as percentage of total rural population was 46% with wholly rented land, 9% with partly owned and partly rented land making a total of 55%. Tenancy was not a major problem in the north where communal tenure prevailed. There were, for instance in Gojjam, an area of highest pure tenant holdings, only 13% of the whole holdings were under pure tenancy (Imperial Ethiopian Government (IEG), 1972). Cohen and Weintraub (1975) noted that tenancy was not a major problem in the north because there the peasantry had been associated with kinship and village tenure systems which allowed most peasants access to land, prevented its alienation and hence the emergence of a landless class. Tenancy in this area was associated with submerged caste groups, religious minorities, or young ristholders who sought more land. It was given that out of the Amhara and Tigrean peasants living in the north, fewer than 5% spent their lives as tenant farmers, In fact, many farmers used to begin their farming as tenants with the expectation that they will ultimately hold rist rights over enough land to support their families. The other major pattern of northern tenancy occurred where the major rist holders rented additional land to acquire more wealth. This was interpreted to show that in the north tenancies were not a hindrance but a means for moving toward more production and ultimately higher status. Cohen and Weintraub (1975) also underlined that the view that tenancy was the major constraint to increased peasant production had not been supported by substantial and empirical findings. They argued that assumptions that tenants having no ownership incentives would not innovate or improve the land beyond their minimum subsistence needs and that innovation would not occur where rents were excessive and placed the risk of extra output solely on the tenant were based on comparative international experience. However, some experts noted that the economic impact of the tenancy system was negative. For instance, after results of rigorous studies on the major features of the then tenure system, it was shown (IEG 1972) that tenancy was characterized by such features as: payment of fees as a precondition for leasehold; 90% of all lease agreements were verbal making dispute settlement arbitrary, largely equal division of produces between tenants and landlords; payment of tithe or land tax by tenants; and no compensation for tenants investments in land. It was underlined that Under such conditions the almost 50% of southern Ethiopia s peasants who operate fully rented holdings can scarcely be expected to take part in the technological revolution which Ethiopian agriculture is now moving (p.9). Of course, care could have been taken in showing tenancy figures across the nation, as all those who rented out their plots were not necessarily landlords. For instance it was reported that the number of landlords and tenants in Arussi (1961 E.C) was 105,689 and 128,001 respectively (Jami Mavi, 1970:475). This means that the landlord/tenant ratio was 1:1.2 implying that the general picture does not seem the prevalence of big landlords.

Shiferaw (1992) also cautioned that the whole idea of tenancy was rather slippery. He asked How many of the tenants were tenants of landlords and tenants of other peasants? He noted that there was a difference between being a tenant to a landlord and a tenant to another peasant. With the peasant, eviction and other exactions and services would not be a major problem. But with landlords eviction increasingly became a threat particularly in those provinces where capitalist agriculture was introduced. 2.2.2. Tenure insecurity Tenure insecurity has been cited as one of the limitations of the pre-revolution land tenure system. It had different forms from endless litigations over land rights to complete eviction from holdings. Arbitrary peasant evictions were among such sources of insecurity. For instance, it was reported that only as a result of the CADU project in Arsi, 20 to 25% of all those farmers living in the area in the mid 1960s had been evicted by the end of 1970, most of them leaving Arussi in search of new land (IEG 1972:19). Endless litigations have also been mentioned as one of the feature of the rist tenure. A study report by Alem-Ante (1970:218 ) on legal aspects of agricultural land disputes pointed out that it was generally believed that land in Ethiopia was the subject of numerous disputes and endless litigation. It was mentioned that an estimate made at the number of land cases in relation to the total number of civil cases has put the figure for all land disputes filed in the ordinary courts at 75 %. (This applied for the whole country and for all levels of courts, and it appears quite evident that court cases involving land more than anything else are a basic feature of the Ethiopian legal system). Civil disputes over land included cases such as failure of the tenant to pay rent, mismanagement of farms collection of crops before assessment, eviction, claims of inheritance, trespass, boundary, and ownership. Criminal cases were not uncommon. For instance, it was mentioned that in the study of court files conducted in Kuni wereda (Hararghe) for the period between Tahisas 1, 1958 and Hidar 30, 1959 (E.C.), it was found that 20% of the criminal cases involved land. Tenure insecurity has increased the burdens placed on the tenant by extractive rental arrangements. As mentioned above, (IEG 1972) 90% of all lease agreements were verbal, making dispute settlement arbitrary, and there was no compensation for tenants investments in land when evicted. Dessalegn (1984) has also argued that for all tenants, the major factor for their dependency, and the chief obstacle to improved production was the lack of security of tenure. Each sharecropper was never certain how long he would cultivate his holding, or when he would be told to give it up. He added that evictions were not common but the threat of eviction, rather than the act itself, was the potent weapon in the hands of landlords, and the tenant over whom the danger of unemployment and destitution hung like the sword of Damocles had no alternative but to accommodate all the demands of his landlord, however onerous they were. Shiferaw (1992) also noted the arbitrary nature of eviction by landlords saying that the kind of eviction that was dangerous to the peasant

population was the eviction simply because the landlord wanted to do it for one or another reason. Landlords had full power to evict their tenants though tradition very often militated against it. But the fact remained that the tenant was under certain insecurity as a result of this. This fact was fully recognized by critics as well as by the government. 2.2.3. Diminution and fragmentation of holdings It is quite interesting that diminution of holdings and fragmentation of parcels existed even in the south where there was concentration of land under few landlords. Fesseha (1970) noted that operation of sub-economic holdings and land fragmentation had been among those principal setbacks to agricultural production during the prerevolution period. It was estimated that out of the estimated 4 million farming families about 90% operated farms of less than 5 hectares and of those two-thirds operated farms of less than one and half hectares. This was interpreted as essentially constituting subsistence living in the agricultural sector with a production level limited to home consumption only. In 1972 it was reported that (IEG 1972) in 12 provinces where a survey was conducted two-thirds of all peasant holdings comprised no more than one hectare under crops and forty-two % of all peasant holdings comprised half a hectare or less under crops. Dessalegn (1984) pointed out that rural Ethiopia before the land reform was a land of smallholding peasants and of petty production. The reason for the co-existence of big landlords and smallholding peasantry was that the Ethiopian landed classes had for the most part parcelled up their properties and leased them out to small holding cultivators. The MoA (1975a) survey report shows that in the country as a whole 57.6 % of the holdings had a size of less than one hectare, 38.6 % one up to five hectares, and 3.8% holdings of five and more hectares. Corrected average area of holdings was estimated to be 1.85 hectares. Table 1. Percentage distribution of number and area of holdings by size for the whole country Holdings by size of total area (in ha) Holdings Under 10.01 & 0.11-0.50 0.51-1.00 1.00-2.00 2.01-5.00 5.01-10.00 0.10 above Number Cumulative percentage of total 3.5 31.1 57.6 81.0 96.2 99.0 100 Area 0.2 6.0 18.4 40.7 70.7 82.8 100 Source: MoA (1975a: 54). Cohen and Weintraub (1975:49-50) noted that particular attention of critics in kinship/rist areas was focused on fragmentation and diminution of holdings. It was stated that in many parts of the country, population pressure on cultivated land has become serious. It was also argued that the problem has in part been aggravated by the tendency of

Ethiopian land tenure patterns to promote diminution and fragmentation of plots. In one way or another there has been a general parcelling out of the holdings so that the average unit is small, and the actually cultivated area is smaller still. Such smallholdings were often found fragmented into several plots. Gilkes (1975:121) has also argued that the Ethiopian farmer, whether a tenant or a small freeholder, suffered from the size and fragmentation of his holdings. The MoA (1975a: 58-59) agricultural sample survey (1974/75) puts fragmentation as a very serious problem in some regions of Ethiopia. However, the survey results indicate that for the country, as a whole average parcel of holdings was 2.46. Forty percent of the holdings consist of one parcel while 19.6% of the total holdings consist of 4 and more parcels. The general argument forwarded towards the negative impact of such holding patterns on production was that fragmentation of land had high costs as time is wasted travelling from plot to plot, and it was not possible for the farmer to take advantage of even the smallest economies of scale. And, where total holdings were smaller than the optimum, the already pressing underemployment problem would be increased. It was also argued that smallholdings reduced access to credit that was needed to improve land or to obtain agro-technology. Table 2. Holdings Holdings by number of parcels Total holdings (000 ha) Number of parcels 1 2 3 4-5 6-9 10 and above Holdings 4127 1650 1003 664 545 252 13 % 100 40 24.3 16.1 13.2 6.1 0.3 MoA, 1975a: 59. 2.2.4. Land concentration and underutilization Studies also show that given the existence of many landlords, particularly in the south, there was concentration of land in few hands while there were many land-starved and landless at the same time. It was also found that given the nature of landlordism that existed, many of those big holdings were not utilized. Among the important measure, cited that led to land concentration was government land grants. The maximum size of holdings in gasha for the southern provinces of Illubabor, Sidamo, Shoa, and Keffa were 405, 180, 149, 135 respectively (MLRA, 1971) and in those communal areas average percentage of absentee owners was 26 and their land to total privately-owned area was 36. A study report on such idle lands pointed out many issues associated with the concentration and underutilization of land (Fesseha, 1970). It was found that available estimates on the proportions of actual cultivated to uncultivated arable land showed that by far a large percent of the country s arable land had remained idle. It was mentioned that by far a higher percent (ranging from 74 to 83.4%) of the total arable land of the

country had remained uncultivated. The most important reason cited for such large idle lands was concentration of land under non-farming elites. Moreover it was argued that the presence of a high rate of absentee ownership was one of the reasons for such underutilization of land. This was the case especially in the central, south, and southwestern provinces where absentee owners constitute as high as 41.8% of all land owners in Illubabor ( Fesseha 1970:139). It was noted that in those areas held by absentee landlords, production in most instances, was nil or, at best, less than optimum. Moreover, absentee landlords were found reluctant to undertake capital investment on the lands, their major objective being one sided, mainly acquisition of the land. It was also argued that government land grant system was largely responsible for such concentration of land. Negatu (1970: 215) stated that among the problems associated with government land grants was that land was given as reward for its own sake and that some grantees used it, others sold it, and some others retained it for speculative purposes. He concluded that considerable areas of granted land had remained idle. It was found that land grants were made to patriots, exiles, patriot s wives and their children, ex-civil servants, armed forces and police forces. Grants were made in such amounts as 5 gashas for captains; 5 gashas for lieutenants; 3 gashas for sergeants, and 2 gashas for officers of lower ranks. Cohen and Weintraub (1975) also indicated that out of all grants approximately 95% of the registered land grants and 75% of the unregistered land grants had gone to members of civil service, military and police, and national or provincial elites, whereas only 5% of the registered land grants and 25% of the unregistered land grants had gone to landless or unemployed citizens. Dessalegn (1984:34) explained that the landed aristocracy was one of the important social classes in the pre-revolution Ethiopia. He described members of this class as powerful, often absentee landlords, who held positions of high authority in the state apparatus, who owned vast tracts of land and engaged a vast army of tenants throughout the country. He noted that though a few among them had began to show interest in the more dynamic and more profitable mechanized agriculture towards the end of the 1960s, most of them were content to lease out their property to share-cropping cultivators, or to let them lie unused. At the same time Dessalegn mentioned that the most deprived sector of the rural population were the landless and cited an estimated figure of slightly over 400,000 such households in rural areas. Mesfin (1991) argued that there was inequity in the land tenure system under the Imperial regime. He stated that In the whole country the proportion of peasants who owned some land was slightly more than 49% while nearly 39% were landless peasants (p.82). He added that the fundamental inequity of the land tenure system under the ancient regime was not only a problem of tenancy. That more than 61% of the peasant households cultivated only 26% of the land while only 18% of the landholds (sic) had 53% of the land was painful for most. That is the larger problem (p.83). Fesseha (1970), for instance, suggested the following as the possible reasons as to why private owners (and these were largely those grantees) kept such lands

idle: incompatibility of the individual s financial position with the size of the land, lack of technical capacity, lack of interest to be engaged in agriculture (as they might have other important sources of income), hoarding land for social prestige rather than economic importance, and holding land for speculative purposes. He also added that lack of mechanisms like progressive taxation to encourage full utilization of land had contributed to such concentration and underutilization. As a result, among other measures recommended after the land tenure study of MLRA in the late 1960s was the need for progressive taxation on those lands kept idle. 2.3. Summary The pre-1975 period land tenure had diversified forms throughout the country and this is noted as one if the causes that hindered any serious progress towards reforming the system. Rist/kinship, private, church, and government tenures are used in this study for the purpose of avoiding complications and making the discussion congruent with most literature on the subject. The rist/kinship tenure system was most prevalent in the northern part of the country while the private tenure was found in the rest of the country. At the same time government tenure was most prevalent in the lowland and pastoral areas of the country. Church tenure was associated with the Ethiopian Orthodox Church. The amount of land under each of those tenure systems was never known with a reasonable degree of precision. A good example is that church lands were estimated to constitute from 50 to mere 5 % of all lands. Rist system was characterized by the principle of acknowledging access to land (use rights and transfer rights without land alienation) by all those descendants of people from a common ancestor and in an ambilineal way. Reduced landlessness and tenancy was among the positive attributes of this system compared to the private tenure while diminution of holdings, land fragmentation and persistent litigation over land access were among its serious problems cited in the literature. With its Gult rights over those landholders, rist system had also been an important element mediating the control of the peasantry by the regime through those elites holding such rights. The private tenure was recognized as the most dominant system during the last days of the Imperial regime affecting some 60 % of peasants and 65 % of the country s population. It was largely created by way of land granting by the crown to those members of the army who went from the north and those who were loyal to the regime in the conquered areas. Under this system land was sold and exchanged, but given that all the land was originally state property and that private holders had no absolute rights, it was different from the western concept of freehold system. Serious land concentration, exploitative tenancy, and insecurity have characterized the private tenure system. Government tenure reflected the predatory nature of the regime as it was established on those lands snatched from people in the pastoral and other areas of the country. Though not known with precision, it is indicated that nearly 47 % of all the land of the country and about 12 % of its agricultural land were held under this title. Under this tenure, land was used as a principal spoil of the then regime given as a reward to its political supporters. The land tenure system of the Imperial regime was largely considered as a hindrance to the country s development in general. It was made the most important cause of political

grievances that led to the overthrow of the regime. Institutional inadequacy- lack of necessary legal framework, absolute arbitrary control of the land rights, and lack of land administration organizations- had characterized the Imperial Ethiopian Government s land tenure system. Land concentration for political reasons in the hands of absentee land lords and its underutilization, unchecked and exploitative tenancy, tenure insecurity including arbitrary eviction, diminution and fragmentation of farm holdings, and other problems are noted as features of the Ethiopian land tenure system that hindered the development of agriculture in general and the economy as a whole. 3. LAND TENURE UNDER THE DERG REGIME 3.1. The 1975 land reform and abolition of diverse tenure The 1975 land reform of the Derg has been considered by many as a radical measure that has abolished the tenant-landlord relations in Ethiopia. The rural land reform proclamation ( Proclamation to Provide for the Public Ownership of Rural Lands, No. 31/1975) stated that feudal lords had exploited the peasantry. It aimed at fundamentally altering the then agrarian relations and liberating the peasantry and making it owner of the fruits of its work; increasing agricultural production; creating employment; distributing land and increase rural income; and laying down the basis for the expansion of industry. The basic provisions of the proclamation include: public ownership of all rural lands; distribution of land to the tiller in provinces with privately owned rural land (art 4); prohibitions on transfer of use rights by sale, exchange, succession, mortgage or lease, except upon death and then only to the wife, husband, or minor children of the deceased or in their absence the children of the deceased who have achieved majority; and, in the case of communal lands, possessory rights over the land the peasants till at the time of the reform ( Art. 19). The power of administering land was vested in the Ministry of Land Reform and Administration (MLRA) through Peasant Associations at the grassroots level. The law provided ten hectare of land as maximum a family can possess. No able adult person was allowed to use hired labor to cultivate his holdings. There is a general consensus among many writers and observers that the land reform law was radical in its nature and content and that it was an appropriate measure to take. Teketel (1998) has concluded that there was no doubt that the Ethiopian land reform has basically abolished the legal, political, and economic power of the landed classes as a whole. On the other hand, it is not adequately known to what extent the reform has restructured (say for instance the average farm size peasants worked both as a tenant and as landholder) landholding among the peasantry itself at a country- wide level. Bruce et al (1993) also noted that the reform had a positive side. It abolished landlordism and the exploitation of the peasantry by the landed classes and redistributed available assets to the peasantry on a relatively equitable basis. They added that the reform has given rise to a uniformity of tenure throughout the country (with the exception of the areas in the periphery), thus doing away with the complexity of tenure

arrangements that in the past made rural development work difficult and time-consuming. All peasants were assigned usufruct rights over the land they cultivated. In addition some suggest that the land reform improved the standard of living (level of income) of the former tenants by the amount of the rent and other dues that they had to pay their landlords before the redistribution of land. In the former private land tenure areas, it is also asserted that land redistribution has eliminated rural unemployment by safeguarding the right of the former tenants to a source of income (Kidane, 1990). However many shortcomings of the land reform are also cited, though it seems that they could be related more to the general agrarian policies of the Derg than the land reform per se. Teketel (1998), for instance, argued that unlike in the tenancy areas, the land reform had not brought any immediate material benefit to the majority of the peasantry in the northern areas. Moreover, he added that the rigidities attached to the land reform such as the immobility of land and labor had important bearings on the peasantry in these famineprone areas. Kidane (1990:114) in his analysis of a time series data concluded that no clear improvement of the country's agriculture has materialized as a result of the redistribution of land. He concluded that there was no general improvement of agricultural production attributable to the redistribution of land. Ethiopia's agriculture has continued to stagnate since the land redistribution. Then, regardless of the general assertion on the appropriateness of the initial (1975) land reform, many scholars argue that the land tenure system since then has badly affected the country s agricultural production and sustainable land use (see Yigremew, 1999). According to Bruce, et al. (1993:2) the main shortcomings of the reform were the following: that the land reform forced peasants to engage in periodic redistribution of land to accommodate new claimants and new members of peasant associations with the net effect of leveling down and diminution of individual possessions; repeated redistribution gave rise to and exacerbated tenure insecurity among peasant households, which was perhaps the most damaging aspect of the reform; the socialization policies of the Derg, pursued with great haste and without peasant consent, further heightened peasant insecurity, and led to loss of incentives for land improvements or increased effort; and that the extractive programs of the Derg impoverished peasant households and extended and deepened rural poverty. Similar conclusions were made by the Ministry of Agriculture (1987) Achefer-Shebedino Study. The study noted that the growing population represented a severe pressure on land resources, with serious ecological consequences, including soil erosion and deteriorating standard of living. The explanations associated with the then land tenure problems were that in order to accommodate the ever-growing farming population, redistribution and allocation of farmland were being carried out in a desperate and hazardous way. The fragmentation and diminution of land as a result of redistribution and growing population had created a sense of insecurity among peasants. And as there were no limits to the scale of redistribution, and no minimum landholding size was set, redistributions had led to excessive diminution of plots.

Survey of the literature on the land tenure during the Derg regime (1975-1990) generally shows that diminution and fragmentation of holdings, tenure insecurity and all its consequences, land degradation, and inefficient allocation of land by way of restrictions on land transfer and to some extent lack of appropriate land use and administration are the most commonly cited problems. Many case studies and some large-scale ones illustrate these situations. The most important reason given for such problems was the continuous land redistribution practice of the government. 3.2. Land tenure problems 3.2.1. Diminution and fragmentation of holdings The Ministry of Agriculture (1987) Achefer-Shebedino study, carried out during 1984-1985 in two weredas, in Gojjam and Sidamo tried to demonstrate such a diminution process. The study seriously noted the problem of continuous redistributions which resulted in diminution of holdings. For instance the study mentioned that holding size of respondents from Achefer had dropped by almost 50% over a period of 10 years because of redistributions. The admonition made was that in the foreseeable future redistribution would, if unchecked, reduce peasant plots to uneconomic size. Table 3. Landholding by size Achefer, 1974 and 1984 Holding size in 1974 1984 timad* % of total holdings % of respondents % of total holdings % of respondents 0-10.31 - - 1-5 1.08 5.01 4.61 15.88 6-10 5.24 11.81 27.76 36.70 11-15 9.94 13.48 35.61 30.08 16-20 12.47 12.65 18.78 11.49 21-25 17.00 14.00 8.00 3.90 26-30 17.10 18.48 3.56 1.39 31-35 10.00 4.85 1.68 0.56 36-40 9.82 4.01 - - 40-45 2.40 1.11 - - 0ver 45 14.95 4.29 - - Source: Ministry of Agriculture (MoA) (1987:67). * 1 timad is equivalent to 0.252 hectare. Those figures from the table were taken to clearly show the levelling down of holdings. In the process, the largest holdings and the landless have disappeared. A more exact measurement of the equity of landholding, the Gini coefficient, was noted to be 0.383 in 1974 and 0.288 in 1984 in Achefer showing that the redistribution had been

effective in fostering equity of holdings (a substantial reduction of inequality). The study also showed that in Achefer, both those who owned and those who rented their land in 1974 and earlier, had much less land during the study time than they ploughed before the land reform. It pointed out that there were feelings of insecurity among the majority of peasants. The major reason for insecurity in Achefer was fear of redistribution. In Achefer, since 1980 an average of five redistributions were carried out, and average holdings of the respondents have been reduced by as much as roughly 50% after the land reform. Some similar studies show that the average size of peasant holdings has diminished by far. Teketel (1998) has produced data showing nation-wide changes for a period of fourteen years (1974/75-1988/89). Table 4. Percentage Distribution of Number and Area of Operational Holdings (1974/75 and 1988/890) Size (hectares) 1974/75 1988/89 Changes Number Area Number Area Number Area Up to 0.10 3.50 0.20 10.51 0.65 7.01 0.45 0.11-0.50 27.60 5.80 36.19 13.12 8.59 7.32 0.51-1.00 26.50 12.40 25.40 23.12-1.10 10.72 1.01-2.00 23.40 22.30 20.16 35.22-3.24 12.92 2.01-5.00 15.20 30.00 7.43 25.43-7.77-4.57 5.01-10.00 2.80 12.10 0.29 2.21-2.51-9.89 10.0 & above 1.00 17.20 0.02 0.25-0.98-16.95 Total 100 100 100 100 Source: Adapted from Teketel, 1998:160. The Table clearly shows that the majority of peasant households both in the pre-reform (57%) and in the post-reform period (72%), operated holdings of one hectare and less. The total number of this group of mini-plot holders increased substantially after the reform. On the other hand, the middle category (1.01-5.0) and holdings of 2 to 5 hectares together appear to have been strengthened after the land reform. A substantial change has occurred in the size category of above 5 hectares. In the pre-reform period (1974/75) households operating above 5 hectares constituted 3.8 % of households but held 29.3 percent of the total area. After the reform (1988/89) however this group of households constituted 0.31 percent of households and held only 2.6 percent of the total area. It is in this group therefore that a significant change has occurred indicating the levelling down effect of the land redistribution. It is also interesting to note that there were households in the post-reform period with holdings of above 10 hectares apparently in contravention of the law. (But it is insignificant, 0.02 and 0.25 % of households and operated area). Gizachew (1994) cited some instances that in some areas the average holding size was found much smaller like in the case of the previous Sidamo province where the average holding was reported to be 0.32 hectares.

Dessalegn (1994) stressed that farm plots were becoming too small throughout the country. In his words (P. 3): In the enset zone, for example, many family holdings are now too small to farm, while in parts of Northeast individuals work what may be termed starvation plots. In his other work (1997) he also seriously underlined that Ethiopia s agricultural system since the last three decades is a shift from a smallholder to a microagriculture. In micro-agriculture systems, he explains, households have insufficient basic capital assets and they are trapped in production for survival. Such systems are also not to progress anywhere but deteriorate as they do not produce wealth but survive by consuming their meager assets. These systems are also fragile as they collapse easily under minor pressures such as mild rainfall variability or moderate environmental stress. Dessalegn also argued that this negative transformation has been accelerated since the Revolution and mainly as a result of the post-revolution land policy. The periodic redistribution of holdings was mentioned to be the main reason for such problems. Many researchers have indicated that diminution of holdings has many implications for agricultural development. One serious concern with this phenomenon is that given the level of farming technology in this country, peasants with such small holdings will not be able to produce enough food surplus for the market or even for their families. For example, Gizachew (1994) pointed out that a rough crude estimation of viable minimum economic holding size per household under low and moderate input levels of technologies was estimated to range from 6.50 hectares to 5.37 hectares respectively. Another group of six researchers (Mulat et al. 1998:5), after analyzing data from different surveys and field research, concluded that the average farm size of about one hectare for a family with approximately 5 persons was, at best, barely adequate for feeding household members. Another impact of smallholdings is related to the adoption of new agricultural technology. Low level of agricultural technology is one of the reasons frequently mentioned for low level of agricultural production and productivity in this country. Although sometimes reluctance of peasants is given as an excuse for not adopting new technologies, it is more associated with resource endowment than attitude. Here are some examples: Mulat et al. (1998) in their study on the determinants of fertilizer use concluded the most important factor explaining the quantity of fertilizer used per hectare is average farm size. As farm size increases, so does the intensity of fertilizer use. Small farms need help in intensifying their use of fertilizer in order to make broad based improvements in farm productivity. Frew and Shilema (1997) in their study on agricultural technology argued that tractors are uneconomical in fields that are less than two hectares and a field length of less than 200 meters and animal draught technology is recommended for a plot size of up to three hectares. They noted that animal power technology would be more efficient provided that it is supplemented with improved and efficient implements. The researchers analysis of results in experimenting with an improved plough shows that the main reason for peasants not accepting the improved plough was the small size of land which could easily be handled with the traditional maresha.