UNDER THREE FLAGS:
THE POLICY ENVIRONMENTS FOR PASTORALISTS
IN ETHIOPIA AND KENYA
 
 
 
 
SR/GL-CRSP Pastoral Risk Management Project (PRMP)
Technical Report 04/99
July, 1999
 
 
 
 
 
 

This publication was made possible through support provided by the Office of Agriculture and Food Security, Global Bureau, United States Agency for International Development, under International Development Grants No. DAN-1328-G-00-0046-00 and PCE-G-98-00036-00. The opinions expressed herein are those of the authors and do not necessarily reflect the views of the U.S. Agency for International Development.
 
 

PRESENTED TO THE SR/GL-CRSP PROJECT FOR IMPROVING PASTORAL RISK MANAGEMENT ON EAST AFRICAN RANGELANDS
 

PREPARED BY JON R. MORIS
UTAH STATE UNIVERSITY
 
 

Proper citation: Moris, J.R. 1998. Under Three Flags: The Policy Environments for Pastoralism in Ethiopia and Kenya. SR/GL-CRSP Pastoral Risk Management Project Technical Report No. 04/99. Utah State University, Logan. 119 pp.

 

TABLE OF CONTENTS

Executive summary

Acronyms

Tables

1. Introduction

2. Comparing Ethiopia and Kenya
            Similarities
            Differences

3. Past Policies Towards Pastoralists
            Differences
            Similarities

4. National Policy Formulation and Documentation

5. Ethiopia and Kenya's Current Policies

6. Some Reflections about Ethiopia's and Kenya's Policies

7. Possible Consequences of Ethiopia's Unstated Policies

8. The Limits of Kenya's Technocratic Approach

9. Five Partially Successful Policy Initiatives

10. Unresolved Policy Issues
            Major Unresolved Policy Issues
            Some Implications
 

Appendices:

1. Themes for Analyzing the Policy Portfolio

2. Policy Documents Reviewed

3. Sources of Special Interest on East African Pastoral Development

4. Sources of Special Interest on African Agricultural Policy Choice

5. References
 
 

LIST OF TABLES

1. Policy Related Events in Kenya, 1978-98

2. Historical Events in Ethiopia, 1928-

3. Ethiopia's Third Livestock Project Achievements

4. Policies Affecting Ethiopia Food Security

 

EXECUTIVE SUMMARY
 

This study focuses on understanding the evolving policy environment for pastoralism in southern Ethiopia and northern Kenya. It is important because donors put great stress upon policy analysis as the starting point for addressing development issues, and because the pastoral environments pose unusual difficulties to national policy-makers. Our approach involved an in-depth review of documentation supplemented with interviews with decision makers in Ethopia and Kenya during March to June of 1998. While the report gives a detailed historical account (in chapter 3) of how policies towards the pastoral sector were implemented in different decades, this summary focuses on the lessons from the study for current policy-making.
 

Compared to highlanders, pastoral lowlanders in both countries have been marginalized in terms of development. They occupy borderlands which are often politically sensitive. Despite social and ecological stresses, human population growth in the rangelands has been significant. During the past 20 years a network of settlements has emerged based on continual distribution of food relief and has contributed to a "desert urbanization of the poor." General differences between the two countries: Both countries have attempted (Kenya) or are attempting (Ethiopia) to transfer more power of governance to local levels, but this has taken different forms-- Kenya began with a "District-level focus" in 1984 and Ethiopia is now implementing a federal system. The Kenya effort at decentralization has waned somewhat over time, however, and most power is still found in Nairobi. Kenya has an advantage over Ethiopia in that pastoralists in Kenya have been more exposed to rural education-- in some cases for over 20 years-- and Kenyan pastoralists appear more aware of the commercial economy. Ethiopia is more pre-occupied with building schools and getting pastoralists to send their children to school, while Kenya's problem appears more to be finding a sustainable way to finance the existing educational operations and infrastructure and getting better performance out of pastoral children who attend school. More provision of pastoral services are provided by NGOs in Kenya compared to that for Ethiopia. Kenya has encouraged and used NGOs to help capture more external funding for development, while Ethiopia has recently been more strict with NGOs and has compelled NGOs to conform and contribute more to governmental priorities for development. In Kenya there is somewhat less diversity in official languages used by government (i.e., Kiswahili, English) compared to that for Ethiopia (i.e., Amhariña, Oromigña, Tigrigña, English). Government in Ethiopia has been subjected to huge upheavals and change during the past 30 years, while Kenya government has been relatively stable, especially at the district level where continuity has been maintained in some rural areas for up to 70 years.
 

In terms of past policy interventions, there are many similarities between the two countries including: A commonly unsatisfactory experience with technical interventions in pastoral regions, which has contributed to disillusionment among pastoralists and agency personnel; overly optimistic development projections by planners; and unhappy donors. From the 1950s to 1980s development programs in both countries have commonly included grazing blocks, veterinary campaigns, range planning, water development, marketing infrastructure improvements, breeding ranches, and various forms of pastoral associations. Ethiopia has continued with large-scale range projects under the auspices of World Bank and African Development Bank financing, while Kenya has tended to refrain from such continued involvements on a broad scale, except for the Arid Lands Resource Management Project financed by World Bank. In the 1990s more typical interventions are localized, small-scale, and participatory including drought mitigation, re-stocking, women's programs, improvement of goats and camels, etc.
 

There is variation between Kenya and Ethiopia in the process of how policies are forwarded, debated, processed, and approved. Major policies in Ethiopia appear to be crafted in general and idealistic forms, while those in Kenya appear to be more detailed and technocratic. In Ethiopia policies appear to be used more for mobilizing a population that is still trying to learn about policy implications. In Kenya, proclamations are often highly specific about investment commitments to various sectors. Investments may not always be delivered, but the Kenyan population appears generally better informed and knowledgeable about policy implications. Ethiopia at present is more sensitive to intrusion by donors or other outsiders in debates over key policy issues, while in Kenya many draft policies and background papers are viewed as not particularly sensitive and are even prepared with donor assistance. Government structure in Ethiopia is dynamic at state and federal levels-- given the massive changes in Ethiopia over the past 30 years this is not surprising. Government structure in Kenya, however, is relatively stable and even outlined in the Nairobi phone book.
 

Other features of Ethiopia include: A national early warning system is run by the federal Disaster Prevention and Preparedness Commission (DPPC); agricultural field bureaus runs by the various states tend to be weak and have insufficient resources; NGOs are experimenting with local paravets, but drug supplies are a problem; the status of livestock marketing infrastructure is unclear-- some resources are in the process of privatization; revenue generation occurs through annual taxation at the household level as administered through peasant associations; pastoral lands are still held by the state, but can be leased to commercial operators; government allows pastoralists the freedom to own guns; bush encroachment is a threat to carrying capacity for range livestock, and privatization is an issue commonly manifested at a local level. Other features of Kenya include: District-based, early warning systems are coordinated from the Office of the President; the Ministry of Agriculture and Livestock Development still carries out a reduced set of technical and participatory activities at the District level; NGOs are also experimenting with paravets; status of marketing infrastructure is unclear-- much has deterioriated and is no longer used; revenue generation solely by fees raised by County Councils on commercial activity; pastoral lands are still held by the state, with some group ranch legislation still in effect-- uncertainty for the future of land tenure in some areas prevails; government only officially condones that pastoralists have guns if they reside near an international border; bush encroachment is a threat to carrying capacity for range livestock; and privatization is an issue more at the national level.

The lengthy process of forming new policies in Ethiopia is caused by several factors, and a "policy gap" can be problematic if it allows poorly thought-out initiatives to be approved in the interim. The lack of policy and relevant technical interventions is particularly evident for the rangelands in Ethiopia. For example, a new move towards forced sedentarization of pastoral populations or initiation of irrigation schemes may be poor choices that could be pursued in the absence of a well thought-out policy framework. In the absence of new policies, old policies of previous Ethiopian regimes would remain as defacto rules and regulations, even if they are not appropriate today. Despite a "policy gap" and a lack of operating resources, the federal and state governments in Ethiopia appear committed to redressing problems in the rangeland areas by virtue of their commitments to creating new development agency offices in the south and paying attention to problems of infrastructure and supporting institutions. Kenya's problem, in contrast to Ethiopia, is a rather poor record of policy implementation in the rangelands despite having more governmental stability and a generally greater access to operating resources and well-trained agency manpower. The few larger range projects that remain in Kenya may still not be making the most effective use of a vast amount of technical information at their disposal. Problems with banditry, infrastructural damage from El Niño, and a decline in tourist revenues have recently hurt Kenya and have probably contributed to more isolation and poverty stress in the northern rangelands. The once impressive governmental services in the range sector still appear well-staffed, but operating resources to make good use of trained manpower are increasingly scarce.
 

Some issues that merit more attention as avenues for new initiatives in both southern Ethiopia and northern Kenya include consideration of how to sustain and strengthen local-level, decentralized development agencies, user associations for water supplies, other forms of grass-roots community mobilization, locally supported paravets, and alternative approaches to existing disaster early-warning systems. One fruitful area of applied research could involve study of how economic policies influence pastoral behavior at international borders, especially when those borders are like the one between southern Ethiopia and northern Kenya that bisects a functioning market space. How government can best achieve its goals and improve the welfare of locals with instruments such as exchange rates, duties, taxes, etc, could be illuminating.

Major unresolved policy issues for Ethiopia may include: (1) How to improve food security in the lowlands; (2) how to mitigate resource degradation in semi-arid range sites, especially in the form of bush encroachment; (3) how to train and retain more talented personnel in the public sector, and encourage more talent to reside in rural locations to better serve the rural population; (4) how to engage more pastoralists in obtaining formal education and making rural education sustainable; and (5) consider other equitable forms of revenue generation for rangeland areas. For Kenya, major unresolved policy issues may include: (1) Dealing with how to improved the security of land claims, rural people, and rural commerce; (2) rehabilitation of infrastructure; (3) reconsidering how formal education can best meet changing needs of pastoralists; (4) how to limit and redress environmental degradation and growing poverty associated with rangeland towns and settlements; (5) how to retain talented public sector personnel, as above; and (6) consider other forms of revenue generation, as above.
 

Readers interested in more details of summarized findings should see the section on Unresolved Policy Issues. A table of contents follows.
 
 

ACRONYMS
 

ADB African Development Bank

ALDEV African Land Development Board schemes, colonial Kenya

ALRMP Arid Lands Resource Management Project, Kenya (World Bank)

AMC Agricultural Marketing Corporation, Ethiopia (Derg years)

ARDU Arsi Rural Development Unit, incorporating former CADU, Ethiopia

ASAL Arid and Semi-Arid Lands, Kenya

CADU Chilalo Agricultural Development Unit, Ethiopia

CBE Commercial Bank of Ethiopia

CBP Contagious bovine plauropneumonia

CGIAR Consultatative Group for International Agricultural Research

CIDA Canadian International Development Agency

CRSP Collaborative Research Support program, USAID

DIDPS District Integrated Development Programs, Kenya

DPIRP Drought Preparedness Interventions and Recovery Program, Kenya

DPPC Disaster Prevention & Preparedness Commission, Ethiopia

DRSRS Dept. of Resource Surveys and Remote Sensing, Kenya

DVO District Veterinary Officer

EDDC Ethiopian Domestic Distribution Corporation (Derg years)

EARO Ethiopian Agricultural Research Organization

ECU European Currency Unit

EEC European Economic Community

EFDR Ethiopian Federal Democratic Republic

EPA Ethiopian Privatization Agency

EPRDF Ethiopian Peoples Revolutionary Democratic Front

EU European Union

EWS Early Warning System

FAO Food and Agriculture Organization of the United Nations

FEWS Famine Early Warning System

FFW Food-for-work activities, linked to World Food Program

GDP Gross domestic product

GHAI Greater Horn of Africa Initiative, USAID

GOK Government of Kenya

GTZ Gesellschaft fur Technische Zusammenarbeit, German aid

IADP Integrated agricultural development program

IAR Institute of Agricultural Research, Ethiopia (now EARO)

IBRD International Bank for Reconstructions & Development (World Bank)

IDA International Development Association

IGADD Inter-governmental Authority on Drought & Development

ILCA International Livestock Center for Africa (now become ILRI)

ILRAD International Laboratory for Research in Animal Diseases (now ILRI)

ILRI International Livestock Research Institute

IMF International Monetary Fund

IPAL Integrated Project for Arid Lands, Marsabit, Kenya

IRDPS Integrated rural development programs

JEPSS Joint Ethiopian Pastoral Systems Study

JIRDU Jijiga Rangelands Development Unit, Ethiopia

KALRES Kenya Arid Lands Research Station, Marsabit (under KARI)

KANU Kenya African National Union

KARI Kenya Agricultural Research Institute

KLDP Kenya Livestock Development Project (1996-81)

KMC Kenya Meat Commission

KREMU Kenya Rangeland Ecological Monitoring unit (now DRSRS)

KVAPS Kenya Veterinary Association Privatization Scheme

LMB Livestock and Meat Board, Ethiopia

LMD Livestock Marketing Division, Kenya

MAB Man and the Biosphere, world-wide UNESCO research program

MOA Ministry of Agriculture, Kenya

MOALD Ministry of Agriculture & Livestock Development Kenya (also MALDA)

NCA Norwegian Church Aid, Ethiopia

NE Northeast (NE Province, Kenya; NE rangelands, Ethiopia)

NEAP Kenya's National Environment Action Plan

NERDU North-East Rangelands Development Unit, Ethiopia

NGO Non-governmental organization

NLDP National Livestock Development Program, Ethiopia

NORAD Norwegian Agency for International Development

OAU Organization for African unity, Addis Ababa

ODA Overseas Development Administration, United Kingdom

OLF Oromo Liberation Front

OP Office of the President, Kenya

PARC Pan-African Rinderpest Campaign

PDRE Peoples Democratic Republic of Ethiopia

PENHA Pastoral and Environmental Network for the Horn of Africa

PIDA Pastoralist Integrated Development Project, Kenya

PMAC Provisional Military Administrative Council, Ethiopia (or Derg)

PMU Project management unit

PRA Participatory rural appraisal

RCC Relief and Rehabilitation Commission, Ethiopia (now become DPPC)

SAP Structural adjustment program

SEPHA United Nations Special Emergency Program for the Horn of Africa

SIDA Swedish International Development Authority

SLDP Second Livestock Development Project, Ethiopia

SORDU Southern Rangelands Development Unit, Ethiopia

SR/GL- Small Ruminant/Global Livestock Collaborative

CRSP Research Support Program, USAID

SSA Sub-Saharan Africa

TGE Transitional Government of Ethiopia

TLDP Third Livestock Development Project, Ethiopia

TRP Turkana Rehabilitation Project, Kenya

UNDP United Nations Development Program

UNESCO United Nations Educational, Scientific & Cultural Organization

UNHCR United Nations High Commissioner for Refugees

UNICEF United Nations Children's Fund

UNEP United National Environment Program

USAID United States Agency for International Development

WADU Walyaita Agricultural Development Unit, Ethiopia

WFP World Food Program of the United Nations

WPE Workers' Party of Ethiopia
 

1. INTRODUCTION
 

This report describes the evolving policy environments within adjoining pastoral areas of southern Ethiopia and northern Kenya. Pastoralists here come under three flags: the Kenya flag, for those like the Gabbra, Rendille, and Samburu who live south of the border at Moyale; the flag of the Oromiya state for those living north of Moyale in the Borana and Guji areas, and also the Ethiopian national flag. A major theme in this study will be the complexities of policy formulation and implementation under these different flags, which stand for three major sources of initiative affecting pastoral development.

The need to synthesize a broad overview of pastoral situations in both countries arises for three reasons. In the rangeland areas, some communities have experienced considerable economic turbulence (related perhaps to El Niño weather) and insecurity. Any statements made about particular areas are likely to be rapidly overtaken by further changes - droughts following floods, for example. A second major reason is the weak performance of many past technical interventions in both countries. When so many promising initiatives have made no lasting difference, it is time to look beyond specific technical interventions to larger contexts which might explain poor policy implementation. And, finally, in both Ethiopia and Kenya sweeping changes were under active consideration during the March to July, 1998 period. Neither government had announced its new policies, whose specific form was still under negotation (with the ABD and World Bank, in Ethiopia's case, and with the World Bank and EU in Kenya's case). Furthermore, USAID's own Greater Horn of Africa initiative was itself being translated into specific proposals which were also not known. In such circumstances, the best that outside analysts can do is to examine underlying structural features which constrain policy choice and delineate policies which show greatest promise.

Fieldwork was undertaken in Ethiopia during March and April of 1998, and in Kenya during May and June of 1998. While brief field visits were undertaken in each country, the author was directed to focus on larger issues at the macro level: the government's policies towards pastoral areas, the institutions involved in implementing pastoral policies, and the issues which such policies must address. Special thanks are owed to the Ethiopian Ministry of Agriculture, to the Oromiya State Government, and to field officials within Ngelle, Yavello, and Moyale; and, in Kenya, to the Office of the President, the Ministry of Agriculture, and district offices in Moyale, Marsabit, Maralal, and Isiolo. The author is especially indebted to fellow collaborative members Dr. Abdulahi Aboud of Egerton University and Ato Solomon Desta. Also, valuable assistance was received from the libraries of ILRI in Addis Ababa and from the World Bank's regional office in Nairobi. Christine Cornelius of the World Bank and George Käsler of the GTZ's Maralal project were especially helpful. Finally, this study could not have been done without the support of the International Livestock Research Institute's Livestock Policy and Planning section, under Dr. Simeon Ehui. The conclusions presented here are, however, the author's own. Since this is only a preliminary report meant to raise issues for discussion, the author (and the CRSP project) would welcome corrections and clarifications from Ethiopians and Kenyans involved in livestock development.

2. COMPARING ETHIOPIA AND KENYA

Similarities

1) Ethiopia and Kenya have an intertwined history. They were "opened" to Europe by explorers and elephant hunters in the late 1890s (see Monty Brown's Where Giants Trod, which maps the specific routes explorers took in northern Kenya and southern Ethiopia). Ethiopia and Kenya have also shared:

• A common experience of fighting the Italian regime in World War II

• A joint experience of major droughts and local impacts

• A partially integrated market across their common border, especially for live animal movements
 

• Repercussions across borders from events, such as the fall of Somalia in the early 1990s and similar ripple effects from Southern Sudan.
 

2) In both Ethiopia and Kenya, pastoralists occupy the lower, more marginal lands which nevertheless constitute the largest share of each country's land area. However, because each country's capital city and main population concentrations occur in the highlands, the needs of the pastoral sector have often been neglected in the past. A major purpose of this CRSP is to provide information on pastoral systems to assist national leaders in Ethiopia and Kenya in finding more effective development strategies to address pastoralists' needs.
 

3) In both Ethiopia and Kenya, USAID launched the first serious efforts to create "modern" range management in the early 1960s, with a pilot project in Borana and a similar NE Rangelands Project in Kenya. Both employed large, 5 block grazing units designed by Frank Abercrombie (Abercrombie 1974). Though largely forgotten by USAID today, these two major projects still survive "on the ground" as a set of rectilinear roads and associated water supplies.
 

4) In both countries, the people living in pastoral areas are the poorest in their relative countries, and face a far more difficult development situation than do the highlanders. Pastoralists tend to be seen by highlanders as "backward" and uneducated. While there are some pastoral groups with powerful national connections - the Borana in Ethiopia, for example, and the Maasai in Kenya - other groups reside as minorities on contested lands.
 

5) In both countries, the borderlands occupied by pastoralists are militarily sensitive. At earlier periods these areas were thought to be either disloyal or potentially disloyal to the regimes in power. As elsewhere, African leaders feel outsiders should not have major say over policies that have major security implications. There is consequently an overlaid political dimension to discussions of development policy for these pastoral areas.
 

6) In both countries, local population growth in drylands has been high despite the huge problems. Much of this growth comes from net immigration, which in turn is composed of either refugees and returnees on the one hand or from incoming highlanders on the other. Immigration of farmers into dry lands has increased the human populations vulnerable to drought.
 

7) In both countries, the pastoral lands are mostly drier and more stressed than the highlands, and have a distinctive set of ecological challenges including the large impact of animal diseases upon the economy. How to privatize veterinary services to assist poor pastoralists in very remote places is a major question facing both regimes.
 

8) In both countries, since 1976 there have emerged a network of rural centers supported by famine relief where the poorest people live. The potential for "relief urbanization" is a new feature which complicates future policy choices greatly. The people in these "desert towns" may depend heavily upon the food aid economy, and usually lack claim to independent local resources. Some are either refugees or returnees, whose rights to remain may be contested locally. The constitute a new type of rural "underclass" often missed in formal statements of development policy towards the pastoral areas.
 

9) In both countries, governments have tried to create agricultural research organizations to promote livestock development. As it happens, in each instance the country's main agricultural research institutions (EARO in Addis, and KARI near Nairobi) have parastatal status and their main focus upon improving crops. Both, too, exist at a long distance from the major pastoral areas. As a consequence, in Ethiopia the Oromia Agricultural Research Coordination Office was involved in mid-1998 in establishing a livestock-oriented research ranch near Yavello. In Kenya, KARI has a long established program for range research at its Kiboko station far to the south, but there is a substation at Marsabit (the former IPAL headquarters) now receiving EU assistance (and thus a prime candidate for collaborative liaison). Egerton University also maintains its own field station near Baringo at the edge of the survey area. It would seem that creating an effective and on-going linkage between pastoral communities and livestock researchers should take high priority in the CRSP plans. Internationally, ILRI with two major complexes of facilities and staff in Addis Ababa and near Nairobi is ideally situated to assist the CRSP team in achieving this objective.
 

10) Both Ethiopia and Kenya are experiencing a widening gap between public and private sector salary levels, with the effect that many top scientists are seeking NGO or private employment. On the positive side, this gives the Ethiopian and Kenyan regimes large pools of experienced, national staff to serve as consultants and planners for new initiatives. On the negative side, it makes it increasingly difficult to retain good research staff in remote, publically financed field assignments. (This also explains the reluctance of some national experts to participate in thinly-funded CRSP activities.)
 

11) Ethiopia and Kenya have broadly similar structures for degree and postgraduate training in the agricultural fields (and in fact cooperate within a common M.Sc. training program in economics). Ethiopia's agricultural degree training began in the then Agricultural College of Alemaya (a USAID project) in 1952, while Kenya's started with a diploma college at Egerton (also receiving subsequent USAID support). Egerton eventually became a University, as did Alemaya in 1985/86. Each institution struggles for its share of funds with an older and well-established national university - for Ethiopians, the Addis Ababa University (with an excellent M.Sc. program in economics), and for Kenyans, with the University of Nairobi (with its own Faculty of Agriculture and Department of Range Management).
 

In both countries, too, the government has established additional institutions whose training overlaps Alemaya's and Egerton's. For Alemaya, the competition comes from Asmara University in the north - an institution to which Alemaya's M.Sc. training in agricultural economics was moved - and from the Awassa College of Agriculture in the south (being upgraded to University status in mid-1998). For Egerton, there is nearby Moi University active in environmental and natural resource fields, and the Kenyatta University of Agriculture (a Japanese aid project just outside Nairobi).
 

12) A paper by Habtemarian Kassa (1996:163) points out that the facilties USAID designed at Alemaya for 200 students now accommodate over 1,200 "using roughly the same facilities." At Egerton, a large block of offices, classrooms, and laboratories stands uncompleted because of a past dispute between USAID and GOK. It would seem a good time, then, to commence equipping Ethiopia's and Kenya's agricultural graduates to deal intelligently with the problems of pastoral development.
 

Differences

The similarities itemized above make it productive to treat Ethiopia's and Kenya's pastoralists as comparable cases. The differences, however, are what make these cases especially interesting from a policy perspective. They also constitute an important reason for conducting parallel research studies, as in the CRSP.
 

1) The boundary between Ethiopia and northern Kenya roughly coincides with an ecological division between better watered perennial grassland which becomes bush thicket when misused or protected from annual fires, and the drier arid lowlands characteristic of Kenya's Chalbi, Koroli and Kaisut deserts. For many years Ethiopian officials have tried to prohibit annual burning, thereby intensifying the problems of bush encroachment within Borana rangelands. In Kenya, the vast lava fields provide insufficient vegetation for either cattle or annual fires, and the people are more likely to emphasize camels and goats as the basis of their livelihood.
 

2) The road network from Addis Ababa to Moyale is paved, and there were regular Ethiopian airways flights to Ngelle (in early 1998). In contrast, the El Niño rains of late 1997 had destroyed large sections of the unpaved highway from Isiolo northwards within Kenya, greatly restricting the movement of people, stock and goods from Moyale and Marsabit southwards towards Nairobi. As of mid-1998, then, physical movement within livestock areas was considerably easier on the Ethiopian side of the border - at least along the north-south corridor within reach of the main roads. Kenyan officials were negotiating donor assistance to rebuild their entire north-south road system (all the way from the port at Mombasa).
 

3) The two countries have adopted major policies to give local areas a significant say in their own governance. Kenya did this from 1984 onwards with a "district focus" approach, a deconcentration of finance and power to the district level. An excellent policy, this initiative faltered when Scandinavian donor funding was terminated and it became clear that little real power had been relinquished at the center. Ethiopia following the 1991-93 change in regime has instituted a federal structure, which devolves national powers to the state governments (here Oromiya State occupying the former Region Four). As of mid-1998, this transition is still underway, making it too early to assess how it will operate once fully implemented. (Of course the outbreak of hostilities between Eritrea and Ethiopia added a further complication just as this report was being written). Nevertheless, in the abstract "deconcentration" and "federalism" constitute the two main alternatives for devolving power to local units, and so make the contrast between Kenya's more mature policy and Ethiopia's especially interesting.
 

4) In line with these two policies, Ethiopia's new structure for its Ministry of Agriculture puts the federal Ministry into a purely advisory role. The Kenya MOALD (Ministry of Agriculture and Livestock Development) still retains executive control over its field agencies, but lacks funding to do very much operationally. If Ethiopia's policies continue their trend up to early 1998, the MOA will have very little operational power over the extension services, which are now assigned to the State Governments for policy and implementation.
 

5) Another major difference is the presence in Ethiopia of a large parastatal banking system, the Commercial Bank of Ethiopia (CBE). This Bank is more than 50 years old, and is one of the top ten banks in all sub-Saharan Africa. It employs 5,000 staff located in 170 branches, including at Ngelle, Yavello, and Moyale - the three main towns serving the CRSP's Ethiopian area. From interviews, it was clear that CBE has established these remote branches in hopes of promoting future growth rather than on the basis of their present miniscule deposits. There is also (see Ziedy, 1996) some question whether the CBE's precarious balance between assets and liabilities will allow it to continue to subsidize its unprofitable rural branches. In Kenya, by way of contrast, the long established commercial banks in Moyale, Marsabit, and Isiolo have contracted their coverage over recent decades, discontinuing the mobile banking they once offered (perhaps because of recent insecurities). They also apply commercial criteria when evaluating profitability. By and large, it seems Ethiopia's rural banking is heavily subsidized whereas Kenya's banks are withdrawing from the peripheral regions they once attempted to serve. (This suggests a need to learn why.) Other reports funded by this CRSP describe the banking situation in detail.
 

6) Ethiopia's previous Derg regime established "peasant associations" (PAs) as the basic unit within which people were registered and to which they paid taxes. In the highlands, people were forced to move into consolidated villages - a source of much of the Derg's unpopularity. In the lowlands like Borana, however, the PA served mainly as an organizing device, identifying a person's registered place of residence. Under the present Ethiopian Federal Democratic Republic (EFDR), PAs continue but have been amalgamated to form larger units with about 5,000 members. Each PA assembly elects a PA Chairman and an executive committee, and is also represented in the Wereda (or district) council. (See the NCA 1997 report by Helland, Olafsson, Ato Teshome Woldesemayat, and Gufu Oba, pp. 37-42.) Some PAs in the Borana also formed into service co-operatives, a major focus of SORDU's activities in its final phase of World Bank support (here see Hogg, 1990). Further discussion of these developments occurs in Chapter 7.
 

While Kenya is also organized into sub-locations and locations (administered respectively by sub-chiefs and chiefs), these are mainly territorial units and do not have the intensive internal organization characteristic of Ethiopian PAs. Kenya's cooperatives are associated with cash crops like coffee, but have not tried to organize pastoralists.
 

7) Ethiopia's pastoralists are for the most part still illiterate and uneducated. Education took root in Kenya one generation earlier, and Kenya's pastoralists have a sizable number of school leavers at the Form IV level and even local communities where most children now attend school. Similarly, Kenya's pastoralists seem much more dependent on the commercial economy, itself in part dictated by the necessity of earning school fees. Consequently, in Ethiopia the stress is on building schools and then getting pastoralists to send their children, while in Kenya it is on how to pay for education and also how to get better performance (many pastoral children do very poorly when in school).
 

The issue of differential education participation and then performance has not yet been adequately studied for either southern Ethiopia or northern Kenya. An important source which looks at this question for the Maasai of Narok is Killian Holland's (1996) book, The Maasai: On the Horns of a Dilemma.
 

8) Associated with the difference in educational participation is the relatively large role of NGOs in provision of pastoral services in Kenya, and the smaller role accorded NGOs in Ethiopia. Currently, Kenya has encouraged local NGOs as a device for assisting remote communities and as a means to capture increased donor funding. (These days, international donors often prefer to work with "private sector" NGOs rather than through governmental agencies.) Ethiopia's view (as expressed to us) is that NGOs should fit their programs into the government's development priorities, should focus upon expanding local capacities, and should have a definite plan for their eventual withdrawal. Thus early 1998 was a time of considerable uncertainty among NGOs working within Ethiopia as they reorganized their efforts to fit within the government's new guidelines.
 

9) There is also a major difference with regard to languages. Kenya's officials are almost all fluent in both Kiswahili and English, while using their "tribal" languages with those from the same area. In Kenya major policies are given in English - witness the 1998 Budget Speech broadcast live in English nationally and listened to avidly by Kenyan citizens.
 

In Ethiopia, while university education is in English many officials are not fully at ease with English and official documents are issued both in Amharic and English. The various state governments have their own languages, which are especially numerous in Sidamo bordering on Borana. Within Borana itself, the Oromifa language is used for government documents and for most field communication. In general, language differences make it difficult for Ethiopia to offer mass education and standardize extension packages suited to all areas and groups.
 

10) Ethiopia's civil regime has been subjected to huge changes over the past thirty years: first, the fall of the Imperial regime in the 1970s and then the collpase of the socialist Derg regime in early 1991. Adults have thus been under a traditional monarchy, a radical socialist regime imposing villagization, a civil war, and now a Federal Republic.
 

Kenya's policies and administration have been characterized by greater continuity. In northern Kenya, the District Commissioner's offices still list the names going back to the colonial period (in Isiolo, to 1929!). Officials use the same uniforms and the same designations as did their predecessors. However, this stability of the Kenya administration--a major objective of the Kenya Government (see Moi 1986:136-151)--contrasts with considerable behind-the-scenes political turbulence within Kenya as the government moved to and then away from one-party rule. (In this author's experience, Kenyan officials tend to view politicians as a disruptive force in Kenya's development.)
 

11) The fact that so many documents are published in English, and the large role of the donor community in Kenya, plus the large number of resident expatriates all combine to make the Kenyan system seem relatively transparent. In contrast to Kenya, Ethiopia is still undergoing a major reorganization of its administration. Policy responsibilities between regional and central authorities are still being devised and negotiated. Officials may not be fluent in English, and were (in 1998) reluctant to discuss policies that had not yet received final approval or were still ambiguous.
 

3. PAST POLICIES TOWARDS PASTORALISTS
 

Differences

We turn now to the core of this paper, a comparative review of the different policies Ethiopia and Kenya have pursued since the 1960s in their attempts to develop the traditional pastoral sectors. While in both countries the policies cluster into three, overlapping sets, the basic differences can be related to fundamental differences in the pastoral sectors between Ethiopia and Kenya. Ethiopia's large population and great poverty meant that the Imperial Government was starting almost from scratch in the late 1960s when establishing a commercial beef industry among its pastoralists. Then, after a military coup and revolution in 1974, its socialist policies were directed towards creating peasant associations compatible with public lands and public entrepreneurship. Kenya, on the other hand, achieved self-rule from the British in 1963 and full independence in 1964. It inherited well developed livestock services oriented towards minority owned mixed farms and ranches in the Rift Valley highlands. Its policies attempted to extend its already robust livestock industry into the remote, pastoral areas which were however lower in population and potential than Ethiopia's. In both countries, policy-making in the third, most recent phases occurs against a backdrop of sometimes dramatic political events, summarized in Tables 1 and 2. Let us begin with a review of Kenya's policies, used by donors as a model for other national programs, before going on to examine Ethiopia's evolving livestock policies and programs.

KENYA(1)

The Harambee Policies (1963 - 78)

Because Kenya achieved "home rule" under Jomo Kenyatta as Prime Minister in mid-1963, its initial policies predate its independence as a republic at the end of 1964. Kenyatta's espousal of "Harambee" politics (or "let-us-pull-together") were aimed at retaining the confidence of Kenya's minority farmers and ranches living in the Rift Valley, while bringing Africans into control over Kenya's well developed services and institutions. Towards pastoralists, however, the colonial regime had done little beyond a few African Land Development (or ALDEV) schemes aimed at localized land rehabilitation (Mwangata 1986). Kenya's veterinary services were substantial, but mainly protected its European livestock raisers (Grey 1963). While in addition a few ALDEV grazing blocks had been established in more remote places like Masailand, Baringo, and West Pakot, it remains generally true that "the colonial period was one of almost total neglect of the pastoral areas" (Liveingstone 1986:251). Indeed, the northern districts lay within Kenya's Northern Frontier District, closed to outsiders for security reasons.

The main task facing Jomo Kenyatta's government in regard to the livestock sector was consequently to try to bring Africans into an already well developed system (von Kaufmann 1976). UNDP/FAO assistance (1964-73) was crucial in allowing the reappointment of expatriate staff who had already worked in Kenya (notably David Pratt, author of a text on range management in East Africa). The FAO/UNDP project established a Range Management Division in Kenya's Ministry of Agriculture (MOA) in 1963; Pratt became the first head of this Division (Keya 1998:5). From his base within MOA, Pratt and his team then mobilized World Bank/IDA support for an ambitious new Kenya Livestock Development Project (KLDP), implemented in two main phases from 1969 onwards with multiple donor assistance (USAID, CIDA and ODA). This was the Bank's first large investment project directed towards African livestock producers. It was meant as a pilot venture to be replicated elsewhere. As described by Livingstone (1986;252-53), the KLDP had by 1974 five main components:

• grazing block developments in the NE Province (under USAID)

• finance for 148 commercial ranches

• group ranches for pastoralists in Kajiado, West Pakot and Kwale

• co-operative society ranches to take over some European ranches

• stock routes and holding grounds to link pastoralists to urban markets

In practice, the KLDP did not meet its ambitious targets. Ole Sadera (1986:19) says that from KLDP's phase one only 15 group ranches and 41 private ranches were developed, adding that a "large portion" of Phase II loan was used to reschedule loans in default under Phase I because of the 1975-76 drought. The Bank, however, portrayed KLDP Phase I as a success even though its own audit report did not support this conclusion. Phase II was expected to last five years and added US $21.5 million, but because of numerous delays was extended through 1981 and expended only $12.4 million (World Bank 1985). It continued Phase one's focus on ranch development, grazing schemes, and livestock marketing while adding special funds to assist Kenya's national parks (Amboseli, Maasai Mara, and Nairobi) and to create the Kenya Rangeland Ecological Monitoring Unit (KREMU, recently become the DRSS). It also brought in additional districts (Lamu, Narok, Samburu, and Machakos) and expanded professional training for Kenya's African staff.

Only three of the KLDP's components overlapped Kenya's northern pastoral districts of concern here. The project included an ambitions experiment in granting group tenure to ranches within pastoral areas under Kenya's Land (Group Representatives) Act of 1968. We should note that after 20 years of implementation, the bulk of Kenya's 159 registered group ranches were found in Rift Valley Province (129), where Samburu District's 27 group ranches averaged 12,315 ha in size (ole Sadera 1986:27).

Then there was a very different USAID designed sub-project to construct immense grazing blocks within Kenya's wild (and desert) Northeast. Using heavy equipment, USAID constructed water points and firebreaks in a pilot area half way between Wajir and Garissa, while also training some 100 Kenyans on livestock topics. The KLDP II goal had been to cover 7.5 million acres within these blocks (World Bank 1985, Annex 4). However, the whole venture encountered huge problems including the fact that the local Somali were camel and not beef producers and the disruption of the 1977 Somali/Ethiopian war, when Somali forces actually invaded Ethiopia from the south coming through Kenya's Northeast. USAID eventually concluded its part of the KLDP "fell far short of meeting original objectives" and terminated further assistance (World Bank 1985, based on USAID's 1983 completion report).

And, finally, the KLDP provided funds to facilitate the flow of surplus animals from "cow/calf" operations by pastoralists into Kenya's developed urban economy. In the North, four major stock routes were planned, linking up at Wajir and passing to Isiolo (where a major holding ground would hold stock under quarantine), supplemented by two subsidiary routes from northern Isiolo amd Marsabit. To manage this vast network of stock routes and holding grounds, the KLDP established a Livestock Marketing Division (LMD), within the MOA. However, the Kenya Government refused to lift controlled meat prices, the very lengthy northern stock routes never became fully operational, and there were long delays in constructing water points by a newly created Ministry of Water Development. There was also no recognized, district-level body (similar to Ethiopia's SORDU) which could coordinate these interventions in the field.

Under the colonial regime, Kenya had established the Kenya Meat Commission (KMC) with a large slaughter facility outside Nairobi to serve as a "last resort" market for cattle (Raikes 1981:191-303). After its establishment in 1970, the LMD was meant to supply about 50,000 cattle annually rising to 100,000 by KLDP's end. Instead Kenya's rising budgetary deficits from the late 1970s cut off funds to the LMD. By 1982, the LMD handled less than 500 animals and in 1983/84 lost Treasury support entirely (Mweya 1986:329). Thus ended Kenya's attempt to set up a parastatal marketing organization which would buy and then resell stock from Kenya's distant northern pastoralists.

The Nyayo Policies (1979-98)

Daniel T. arap Moi assumed office in August of 1978, following the death of President Kenyatta. Among the Kalenjin peoples (a minority in Kenya's population) from whom Moi comes, "nyayo" means roughly "follow-in-steps." Moi's regime did, in fact, continue most of Kenyatta's policies, but added a distinct focus upon Kenya's Arid and Semi-arid Lands (or ASALs)--those lands with between 200 and 800 mm of annual rainfall (Wiggins 1985:94). This should not be surprising: Moi grew up near Kabarnet, at the edge of Kenya's northern Rift Valley, where its wandering pastoralists are a frequent sight.

Moi's policies towards the ASAL districts were set out officially in a 1979 document, revised in 1992, and also figured prominently in Kenya's 1989-93 National Development Plan. Their major long term goal was "to improve the standard of living of the ASAL population by integrating ASAL into the main-stream of the national economy and social development in an environmentally sustainable manner" (GOK 1992:6). Ten subsidiary goals were listed:

• developing human resources and institutional capacities to carry out projects and programs

• strengthening community participation

• encouraging adoption of low-cost, appropriate technologies

• reclaiming and protecting stressed environments

• introducing risk-minimizing agro-pastoral packages

• opening new channels for off-farm employment in micro-enterprises

• improving delivery of education, health and extension services

• upgrading infrastructure for communities

Equally important were the guidelines telling how policies should be implemented: In their stress upon popular participation, NGO involvement, support to the environment and to micro-enterprises, achievement of food security, and reliance upon the district administration, these policies represented a consistent and attractive package closely suited to the actual needs of Kenya's disadvantaged, marginal pastoralists. However, there was an implicit assumption that within ASAL districts the public and private sectors would work together as partners under the leadership of the district and provincial administration - making the Office of the President a key player and initiator of any major initiatives. While consistent with Kenya's past, this is not how western donors see the state's role within a "free market" economy (World Bank 1997). More to the point, Moi's Nyayo policies failed to address the crucial issue of how revenue would be generated to pay for the improved services ASAL districts would enjoy. (Moi's 1986 book remains an essential source outlining in some detail how he hoped Kenya would develop.)

For donors, then, the "problem" with Kenya's post-1978 policies is their continued reliance upon what some call Kenya's "managed open economy" rather than the "market led" policies which Western donors insist all African nations should adopt (see our discussion below of Kenya's "structural adjustment" phase). For Kenya's pastoralists, the problem has been a continued gap between attractive policies and what actually happens "in the field" far across the deserts from Kenya's Nairobi-based politicians.

Moi's government did many of the things which could be addressed by central initiatives. A new university was created outside Kitale in western Kenya, and given environmental management as its special mandate (Moi 1986). A Ministry of Environment and Natural Resources was created, which in 1994 issued The Kenya National Environment Action Plan, or NEAP (see particularly chapter 5, "desertification and drought"). Intensified, area-based planning was undertaken with donor support under World Bank-style "integrated agricultural development programmes" (IADP) with a focus on expanded small holder credit, notably the EEC assisted Machakos Integrated Development Programme (Livingstone 1985:335). Special responsibility for the ASAL districts was vested in a newly created Ministry of Reclamation and Developmen of Arid, Semi-arid Areas and Wastelands which, however, was eventually merged into the Ministry of Rural Development. (Today ASAL responsibilities insofar as donor support is concerned occur in the Office of the President.)

As we have just seen, the Nyayo policies put the territorial administration firmly in command of initiatives to address ASAL needs. This central role was reinforced by the government's need to respond to increasingly frequent droughts from 1979/80 onwards, and by the nearly total reliance of the Kenya government on external financing for any new ASAL investment projects.

Kenya's 1979/80 drought was centered along its northwest border, where Turkana pastoralists were heavily affected (though more so in the north than in the south). NORAD's initial attempts to assist the Turkana only showed that narrow-focus, sectoral interventions were of little benefit under emergency conditions. Instead, a district-wide Turkana Rehabilitation Project (TRP) emerged to coordinate relief and rehabilitation activities throughout Turkana between 1980 and 1987, led by Dutch assistance but involving also many other donors. Turkana pioneered Kenya's famine early warning system, restocking projects, community-based revegetation, field livestock centers, and community-based inoculation campaigns. When in 1984 a national drought struck, it was the Kenya administration's recent Turkana experience which guided its responses. Similarly, for NGOs their greater local involvement was a crucial testing ground for subsequent activities elsewhere. OXFAM, Farm Africa (founded in 1985), and Actionaid became major players assisting the district administration in Samburu and Isiolo, while the Dutch Government and GTZ became lead donors addressing respectively famine early warning and local technical initiatives in Samburu and Marsabit. And, finally, religiously based NGOs became the accepted GOK conduit for World Food Programme assistance in much of northern Kenya. Generally, Kenya's 1980s pastoral development policies were disaster and relief driven, rather than focusing on improved technical interventions (Wiggins 1985; Downing et al. 1990; Buchanan-Smith & Davies 1995).

The one major exception within our study districts was Hugh Lamprey's 1976 "Integrated Project on Arid Lands" (IPAL) which became a UNESCO supported Man in Biosphere (MAB) site for studying local processes of desertification within an ecosystem context. A major UN Conference on Desertification was held at UNEP in Nairobi in 1977. Kenya's isolated Marsabit District seemed an ideal place where the impact of pastoralists on an overstressed local environment could be documented by a multidisciplinary scientific team. Ultimately, over 28 major technical reports were produced describing present land use and husbandry practices in a 22,500 km square area of desert west of Marsabit town.

IPAL went through three phases and two revisions between 1976 and 1984 (Lusigi 1986). Under its Kenyan director, Dr. Walter Lusigi, it became the Kenya Arid Lands Research Station (KALRES) under KARI. It obtained EEC and GTZ funds for applied extension research involving everything from stocking rates to demonstration camel herds, pastoral associations, mobile "camel back" retailers, and even organized stock buying (Oba 1992, 1994; Timberlake 1987:61-72). However, Fratkin (1991) judges IPAL and contemporaneous mission activites as a disaster, akin for local Ariaal Rendille to a drought in respect to the stresses these interventions placed upon pastoralists.

With major droughts returning to Kenya in 1984 and then 1991/92 (in the northern ASAL districts), the Kenya Government drew upon the Turkana experience with TRP to negotiate several additional emergency projects aimed at the northern ASAL zone which had been largely neglected earlier. Key interventions included:

Kenya's larger multidistrict and multidonor projects (like the ALRMP) implemented within ASALs in the 1990s are extraordinarily difficult to evaluate as policy interventions. Their official objectives are only loosely connected to a wide range of components. They work within several districts, involving different NGOs as local partners in each area. Their staff are located at different levels, from the Office of the President at the top to divisional level NGO offices at the bottom. They are mainly financial in nature, supplying a key ingredient - external funds - but one whose impacts depend on the quality of accompanying activities. The field activities in turn have been heavily affected by droughts, flooding and the fluctuating security situation. The whole set of activities is highly dependent on adequate communications, which in recent years almost disappeared in some areas because of El Niño linked flooding. Any attempt to evaluate a project as a whole becomes arbitrary and artefactual. However, on what other basis can project outputs be assessed?

Two other policies, towards veterinary privatization and wildlife, require mention. Kenya's veterinary profession has long seemed an obvious candidate for privatization (Leonard 1993). In 1988, the GOK stopped automatic employment of the University of Nairobi's annual output of 60 veterinarians. From 1994, the policy has been that public veterinarians are withdrawn from districts with private service. In late 1994 the Kenya Veterinary Association launched an EU supported privatization scheme (KVAPS) which advances loan guarantees for the start-up costs of entering private practice. By 1994, there were 154 private veterinary practices in Kenya, though none within the 17 ASAL districts (1995 figures cited in a draft Consortium BCEOM/SATEC report, 1998:53-54). Meanwhile, on a pilot basis Kenya's Director of Veterinary Services has also allowed some exemptions from the legal requirement that only licensed veterinarians can handle and supervise scheduled drugs. USAID, UNICEF, and EU funded programs have supported private paravets, called "community-based animal health workers" (and equivalent to the GOK's "Junior Animal Health Assistants") to administer injections and conduct inoculations, usually with profits returning to a community-supervised revolving fund. As of mid-1998, such activities run under the Pan-African Rinderpest Campaign (or PARC-VAC) through the OAU were considered quite successful. (A USAID field evaluation of such activities in Turkana was just completed in April of 1998.)

Finally, there is the problem that Kenya's major wildlife resources overlap its pastoral areas and could not survive without pastoralists' protection (Western 1997:94-110). Until recently, tourism was Kenya's leading foreign exchange earner, 28% of such revenues in the early 1990s (Ndulu & Mwega 1994:111). To protect the wildlife tourists come to see, the GOK has experimented with game reserves, national parks, licenced hunting blocks, shared park revenues, an outright ban on hunting, and even military-style wars on poachers. Kenya's wildlife officials spear-headed a world-wide ban on ivory sales (Western 1997:233-254). The appointment of Richard Leakey and then David Western to head Kenya's Wildlife Service by President Moi is significant. However, Western's gripping account shows that many senior Kenyan wildlife officials have seen their careers destroyed by opposition from well-organized poaching interests (a carry-over from the Kenyatta years). Today, in northern Kenya an influx of raiders and illegal arms could have the same impact as yesterday's mass poaching. The loss of tourism revenues from internationally reported incidents was having a devastating impact on Samburu and Isiolo County Council finances in mid-1998.
 

The Structural Adjustment Policies (1982-99)

By the late 1970s, Kenya began to experience persistent budgetary deficits nationally, despite a temporary boom in coffee prices from 1976-78. By 1980/81, the deficit reached a record level of over 10 per cent of GDP (Bevan et al. 1990:253). In response, the GOK issues four policy papers between 1980 and 1984 setting out what would be termed today a "structural adjustment program" (or SAP). While these policies did restore self-sufficiency in maize production (through higher controlled prices to farmers), the overall macro situation remained grim.

Prodded by international donors, the GOK finally issued its 1982 Sessional Paper (no. 4) Development Prospects and Policies which advocated limited privatization and austerity. Maxon and Ndege (1995:154-56) describe the changes as a gradual elimination of protective tariffs, remission of price controls, devaluation of the Kenya shilling, and a better system of import licensing. President Moi is quoted saying: "In view of the substantial government deficits and serious balance-of-payments problems, the government had decided to reduce its subsidies to parastatals and may even withdraw its participation from businesses which have failed to perform well" (Aseto & Okello 1997:122). The cautious tenor of these remarks is significant. While Moi did set up a Task Force on divestiture (one of several successive bodies), the pace of government disengagement from its "managed open economy" remained deliberate and slow. Finally, in Sessional Paper No. 1 of 1986, the GOK committed itself to more complete liberalization and privatization in a document which laid out an indicative plan for Kenya's economy up to the year 2000.

Meanwhile, these were especially tumultuous years in Kenya's politics, commencing with an abortive coup by Air Force officers in 1982 (see Table 1). The World Bank had switched a large share of its Third World lending from development projects like the KLDP to structural adjustment, typically accompanied by even more stringent IMF-imposed "stabilization" conditionalities. Other African countries had been forced to accept the IMF package, so why not now Kenya? The US, through its ambassador Smith Hempstone (1989-93), sought to force Moi to reverse his regime's 1982 adoption of a de jure one-party state. There were more frequent droughts, in 1979/80 for northern Kenya, nationally in 1984, and again in 1992 for Kenya's north. Somalia's government fell apart, sending a flood of refugees into eastern Kenya. Ethiopia's DERG regime also collapsed, followed by more refugees into Kenya. In Kenya itself, there was open corruption, assassinations, and student riots. In November of 1991 the international donors as a group froze all discretionary disbursement of external aid.

How, in retrospect, are we to explain this turn of events, and what impact did it have on the distant pastoral areas? Clearly, as Aseto and Okello document at many points (1997), the Kenya Government has moved very slowly when implementing its announced goal of privatization. As late as 1991, Kenya's parastatals employed 170,000 of Kenya's 715,000 formal jobs, contributing to 11% to GDP and 16% of gross capital formation (1997:121). As Nelson (1984), Leonard (1987), and others had argued, Kenya's administration needed the support of its public sector workers and the patronage which the Government's control of parastals provided (see also Bevan et al. 1990:253). Kenya's "controlled, open economy" had earlier achieved a strong economic record, which Moi's Nyayo policies had vowed to build upon (1986).
 
TABLE 1. POLICY RELATED EVENTS IN KENYA, 1978-98
1978 August. Daniel T. arap Moi succeeds to the Presidency, following the death of President Jomo Kenyatta
1979 Kenya Ministry of Finance and Planning publishes a strategy document, Arid and Semi-arid Lands Development in Kenya
1980 Drought in Turkana, increase in famine relief leads to Turkana Rehabilitation Project (TRP)
1982 Unsuccessful coup attempt to oust President Moi; Kenya Parliament amends the constitution to make Kenya a de jure one-party state
1980s District Integrated Development Programs (DIDPs) organized as donor-assisted, area-based rural development efforts. Donors allowed to make direct disbursements within assisted areas.
1983 July. District Focus Strategy implemented as national policy
1984 Severe drought causes Kenya's GDP growth to fall to 1 per cent per annum
1986 Sessional Paper No. 1, Economic Management for Renewed Growth, expresses GOK commitment to restructuring, charts indicative plan to year 2000
1988 GOK stops automatically employing University of Nairobi's veterinary graduates
1989 New Ministry of Reclamation and Development of Arid, Semi-arid areas and Wastelands (MRDASW) declared
1988 National Plan (1988-93) gives explicit recognition to ASAL needs
1989 Smith Hempstone serves as controversial US Ambassador until 1993
1990 Murder of Kenya's Foreign Affairs Minister triggers riots
1990 World Bank publishes its view of Kenya's policy situation, Kenya's Stabilization and Adjustment
1990 December 4. Special KANU delegates Conference under President Moi makes dominant party more representative and democratic
1991 July. Interest rates fully liberalized.
1991 Kenya's first private export-processing zone (EPZ) commissioned in Nairobi
1991 November. Donors suspend disbursement of aid until Nov. 1993
1991 December 10. Kenya's Parliament repeals Section 2A of the Constitution, restoring right to have multiple party politics
1992 Drought returns to northern Kenya
1992 GDP growth slides to 0.4% per annum, lowest since Independence, while inflation reaches 27%, Kenya's highest to that time
1992 December 29. Kenya holds its first multiparty general election since 1963, accompanied by renewed ethnic violence
1994 May. Kenya shilling becomes full convertible at market rates for transactions up to $500,000
1998  World Bank and EU involved in reorganization of Kenya's agricultural services jointly with Kenya Government
 
 
 

And there was the fact, well understood among Kenya's senior economists, that Kenya depends heavily on export crop earnings and tourism for its investment and growth. This makes the economy vulnerable to external shocks (like the 1984 drought), and reduces the attractiveness of laissez-faire policies which generally accentuate price swings, speeding investment in boom years but yielding (these days) nearly instantaneous flights of capital in bad times. As Ndulu and Mwega (1994:102) put it,

Countercyclical management becomes extremely important for economic stability. Net foreign resource inflows have tended to be procyclical ... which puts even more stress on economic management.
 

We face here an unacknowledged war between paradigms. At the very time when the Kenya Government's restrained management of food purchases to offset the 1984 drought was being praised (Cohen & Lewis 1987), the World Banks's economists and less sympathetic observers (Bates 1989:93-122) were condemning it for gross patronage and inefficiency. Buchanan-Smith and Davies (1995) argue that whereas the Kenya Government's response to Turkana's 1990 drought was well managed, by the time of the 1991/92 drought relations to donors had deteriorated so badly that the response was far less effective.

Other than in regard to food relief, most of the "structural adjustment" battles in Kenya concerned national, macro policies. In 1992, the World Bank published its country study of Kenya, calling for a reduced public sector role. The Kenya Government did eventually lift price controls, and free its currency exchanges to depend on market rates (in 1994 the Kenya shilling became fully convertible). Multiparty elections were restored in 1992, and in November of 1993 donors recommenced aid disbursements. In November of 1994, the Bank and GOK embarked on a major project to slim the civil service, drawing on an IDA credit for US $25 million in a total program estimated to cost US $156 million (World Bank 1994). And, finally, in mid-1998 the EU, the World Bank, and GOK officials were in intense negotiations to reorganize the Ministry of Agriculture, sending many of its senior officials into early retirement -- some of the very officials USAID had trained for their present public sector roles.
 

ETHIOPIA

Ethiopia's livestock policies overlap three distinct regimes: in the 1960s under the Imperial Government of Haile Selassie; a post-1974 socialist period under Mengistu's Derg regime; and a post-1991 shift towards liberalization under the Transitional Government now confirmed in office by national elections. These three regimes have had very different policies' outlined in Table 2. Yet, paradoxically, linked pastoral developments were slow to emerge and then were carried over into the next regime. The Imperial Government's growing reliance on the World Bank was continued and intensified under the Derg government's initial years. Similar, the Derg's major livestock projects still survive as possible models for today's liberalized regime. The disjuncture occurs because, as Hogg (1993) points out, successive Ethiopian regimes have been unsure how to adapt policies designed for highlands to the vastly different circumstances of Regions 4 and 5, the Borana and Somali lowlands of greatest interest here. This uncertainty protected the Borana area from the worst excesses of the Derg regime, and continues into the present. Only one account has been located which describes how Ethiopia's pastoral policies evolved, Desta's (1993) paper, used heavily for much that follows. On Ethiopia's agriculture policies, Dejene's 1987 book is also an excellent source.
 
Table 2. Historical Events in Ethiopia, 1916 - 1995
1916 Ras Tafari Makonnan emerges as head of the Ethiopian government
1928 A rebellion and struggle leads to Tafari's victory
1930 Ras Tafari crowned as Emperor Haile Selassi
1935 Italian invasion of Ethiopia
1936 Italy annexes Ethiopia, rules until 1941
1941 Haile Selassi returns from exile in England
1952 Ethiopia takes over Eritrea under a UN mandate
1952 An Agricultural College at Alemaya founded (USAID project)
1962 Ethiopia annexes Eritrea
1963 Organization of African Unity (OAU) founded, headquarters in Addis Ababa
1964 Livestock and Meat Board established
1966 Institute of Agricultural Research established
1967 Chilalo Agricultural Development Unit (CADU) formed with SIDA support
1971  Walyaita Agricultural Development Unit (WADU) formed with World Bank
1973 Major drought spreads over northern Ethiopia
1974 Unofficial council, or "Derg" formed of 108 military men
1974 Sept. 12, Haile Selassie arrested by Derg officers
1975 Rural Land Proclamation (no. 31) nationalizes land-holdings
1976 Government establishes Agricultural Marketing Corporation
1977 War breaks out when Somali forces invade Ogaden, also fighting in Eritrea
1978 Major defeat of combined Soviet/Cuban/Ethiopian forces in Asmara
1985 Massive drought commences, lasting into 1986, across Horn of Africa
1985 Mengistu regime launches compulsory villagization over much of the country
1986 Alemaya upgraded into an independent agricultural university
1987 Mengistu has himself elected to head the Peoples Democratic Republic of Ethiopia (PDRE) under a Marxist constitution
1987 Territorial administration of Ethiopia reorganized into 30 regions
1988 Mengistu is refused further arms purchases in USSR
1990 The Ethiopian People's Revolutionary Democratic Front (EPRDF) formed 
1991 Defeat of Ethiopian army in Eritrea, Mengistu flees to Zimbabwe
1991 Transitional Government of Ethiopia established under Meles Zenawi
1993 Eritrea secedes after referendum on Eritrean independence
1994 New constitution approved
1995 National elections confirm establishment of Federal Democratic Republic of Ethiopia
 
 
 

The Imperial Government's Policies (1960s - 1974)

By the 1960's when our review commences, Ethiopia's Emperor Haile Selassie had been in power for over thirty years. His Imperial Government began to take a more active role vis-a-vis the countryside in the post-World War II period. An Agricultural College was built with US assistance in 1952 at Alemaya, within Selassie's home area of Harar. It meant that senior MOA officials were trained at Oklahoma State, which held the Alemaya contract, and on their return began building a "land grant" style extension system with a focus on "self-help" community development (Kassa 1996:168).

Nationally, Haile Selassie took steps to position Ethiopia at the forefront of Africa's newly emerging nations. In 1963, the Organization of African Unity was created with its headquarters in Addis Ababa. A UN Economic Commission for Africa was located nearby. In 1964, a Livestock and Meat Board (LMD) was established and given the role of promoting the emergence of a modern meat industry within Ethiopia. It helped develop veterinary services to control animal diseases in the pastoral areas, and also worked with the World Bank and FAO in negotiating the World Bank funded Second Livestock Development Project, initiated in 1973 (Desta 1993:45-46).

The US role in Ethiopia's livestock sector preceded the SLDP, however. In the mid-1960s, it implemented the Yavello Rangelands Pilot Project covering 2,400 square kilometers of Borana rangelands (Desta 1963:46). The main technical idea was that by shifting to a deferred grazing, rotational block system, the Boran could substantially increase beef output. The Project used heavy equipment to cut fire-breaks and perimeter roads around huge, rectilinear blocks (still visible in the countryside near Yavello). Large reservoirs, termed "ponds", were built to serve each rotation. What actually happed was that spontaneous settlements grew up around each of the reservoirs, while Boran elders failed to observe the planned grazing rotations and failed to punish trespassers (Desta 1993:46). The "pilot project" became seriously overgrazed and degraded - a condition still evident today, over 30 years later.

The main thrust of the Imperial Government was not toward the distant pastoral areas, but rather focused on introducing Ethiopia to modern, high input farming. In 1966, the Institute of Agricultural Research (IAR) was set up on land near Addis Ababa alongside the CGIAR's new International Livestock Centre for Africa (ILCA), also negotiated at this time. Among donors, SIDA took the lead in promoting a "package" approach to high yielding varieties. It funded the Chilalo Agricultural Development Unit (CADU) within the Arsi highlands (north of Borana) from 1967 onwards. CADU was Ethiopia's first integrated rural development project, a fore-runner of many World Bank Projects of a similar nature all across the Third World. CADU combined the functions of several ministries within its project area: adaptive research on crop varieties and livestock production, input supply (with a strong emphasis on fertilizer adoption), credit and marketing assistance (to pay for increased use of purchased inputs), and generalized soil and water conservation practices. The rationale was that high input farming could not show its potential unless improved practices were adopted as a package, so that modern varieties were matched to better plant nutrition and were in turn supported by on-farm credit.

Indeed, CADU was a success, leading to rapid yield increases. Then the unanticipated occurred. When landlords saw the profits tenants realized from their high-yielding varieties, they promptly evicted the tenants and expanded their own farming (Dejene 1987:17, Cohen 1975). It became clear that without land reform, the outsiders to whom Menelik's Imperial government had given Arsi lands would be the main beneficiaries of improved local farming (Stahl 1974, Gebregziabher 1975).

The CADU "success" was nevertheless influential as a model for other donors, such as the World Bank's Walyaita Agricultural Development Unit (WADU) established in 1971 and USAID's Ada District Development Project (ADDP). CADU lasted nearly twenty years, and went through three phases to become the Arsi Rural Development Unit (ARDU) applied to its entire region. It set the mold for the MOA's approach to extension throughout Ethiopia. For areas not under CADU, WADU, or ADDP, the MOA developed a similar "comprehensive package program" (CPP) which stressed geographically delimited areas stratified by ecological zones and within which complementary inputs would be supplied. The MOA developed seven "packages" for use nationally, matched to differing degrees of intensification within the farming system (Kassa 1996:168 ff).

Unfortunately, extensive pastoral production of the types employed by Ethiopia's 3 million Boran, Afar and Somali nomads has no part within such a scheme. Dejene admits that for his farmers, shortages of oxen were "the most serious problem they faced in farming operations" (1987:102). Goyder & Goyder (1988) suggest this was a general problem throughout the highlands during the traumatic 1985-86 famine years. Yet in the early 1972, neither MOA nor ILCA saw the crucial role which the lowlands play when supplying replacement stock to adjacent highland plough farmers. Scientists looked instead at pastoral animals as potential beef slaughter stock. The LMP's Second Livestock Development Project neglected actual linkages between the lowland and highland agricultural systems, while promoting a hypothetical national beef industry which existed nowhere outside of the planner's idealized maps.

The Imperial Government's failure to comprehend the realities average Ethiopians were encountering came back to haunt Haile Selassie himself, leading ultimately to his death in 1975. The key event which triggered Selassie's downfall was his administration's tardy and completely inadequate response to Ethiopia's 1973 famine. As Goyder & Goyder note, the MOA's report for crop production in 1972 had warned of impending famine (1988:81):

Not only were the drought areas accurately identified down to the district level but the telling signs of disaster - rising grain prices, an acute shortage of purchasing power, and the falling value of livestock in exchange for grain were reported.
 

The Imperial Administration did nothing. It ignored requests from regional governors for grain to feed a total population of nearly two million. No food reserves were held commensurate with the scale of the unfolding disaster, because the famine struck different areas in different seasons. It began in the north, progressed into Harerge on the east, and from 1973 to 1975 affected the Ogaden (Goyder & Goyder 1988:81). Finally, public outrage at the government's failure to act brought about the resignation of Ethiopia's Prime Minister, and in mid 1974 the creation of a new national agency to deal with drought. The Relief and Rehabilitation Commission (RCC) was to play a crucial role in Ethiopia's next major drought during 1982-86 (Jansson, Harris & Penrose 1987).
 

The Socialist Derg Policies (1974-1991)

The events of 1974 sealed Emperor Haile Selassie's fate, though things might easily have taken a different turn had people understood what was happening (Georgis 1989:5-53). In June a shadowy committee of 108 military men - representing all branches at different levels - was constituted taking from Ethiopia's ancient Géez the name dergue or "committee of equals". More commonly, it was called the Coordinating Committee. Initially, its members had no ideology and no intention of deposing the Emperor. They were soon joined by radical students and teachers from Haile Selassie I University, several thousand mostly Marxists. The Emperor was by then in his eighties. He seems to have believed the Derg would carry out a general "house cleaning," but would leave him in power. However, within the Derg was a smaller group of planners. Their leader, Major Mengistu Haile Mariam, was a non-Amhara and still comparatively junior. Mengistu had been to the US for a nine month Army training course. He returned bitterly anti-US because of racist incidents; he had equal hatred for the old order in Ethiopia. His inner group joined forces with the radical students to move the Derg steadily leftwards, culminating with the arrest of the Emperor himself on 12 September 1974.

The details of Mengistu's consolidation in power do not concern us here. We note simply that the Derg, or Coordinating Committee, became the Provisional Military Administrative Council, which in turn came under the Worker's Party of Ethiopia in 1979. Already by 1977, Mengistu had established himself as an absolute ruler in the Stalinist mode. In September of 1987 he finally dissolved the Military Council, becoming instead the General Secretary of the WPE and President of the Democratic Republic of Ethiopia (Giorgis 1989:56-68). For simplicity's sake, I shall use the "Derg" label to refer to Mengistu's regime in all its successively more authoritarian transformations. We should remember that Mengistu's leadership was ruthless and bloody from the start. In person, however, Mengistu was soft-spoken and charming (Giorgis 1989:56-68).

The Derg regime under Mengistu moved rapidly to consolidate its mandate. On March 4, 1975, the regime nationalized all land and industry in Ethiopia (proclamation no. 31). It vested control of rural lands in peasant associations, or PAs. On July 26, the Derg outlawed the ownership of more than one home, and confiscated all extra houses for the nation.

Nevertheless, it was the land reforms which proved most popular. The ideal size for a peasant association was stated to be 800 ha (2,000 acres), with 275 households. As Dejene points out, pre-revolutionary land tenure in Ethiopia was extraordinarily complex, with some the greatest inequities in the south where large landlords predominated in the grain-growing regions (1987:30). Dejene found 61% of his sampled Arsi farmers had been tenants, and only 30% former owner-cultivators. (Earlier studies showed approximately 50% as tenants.) Of Dejene's former tenants, about one third had been evicted from their land prior to the revolution; a majority of them regained the same land under the 1975 proclamation (1987:31). This explains why the Agrarian Reform was implemented with only localized resistance. As Dejene explains (1987:33):

Peasants in the tenancy-ridden south were more enthusiastic about the Reform that those with a communal tenure in the north... Since tenants in the north are landowners themselves, officials of the military government also assured them that the Reform would not confiscate their land or under take major land redistribution in their region.
 

Indeed, Article 19 amended the 1975 proclamation to grant peasants possessory right over the land they till.

Even so, the 1975 Agrarian Reform represented a huge change in how rural lands were owned and managed. The Derg simply did not have sufficient administrative strength to implement changes. Accordingly, it turned to university students who were sent into the countryside in a Maoist-style zemecha "Development Through Cooperation Campaign" (Dejene 1987:33). Students wanted to establish communal farms "for the land-hungry peasants in the south" but also "hoped to mobilize the countryside against the military regime which they considered illegitimate" (Dejene 1987:33). Conflicts soon erupted, forcing the military government to intervene in places like Jimma (serving region 5) and in Dilla, Yirgalem and Wollamo Soddo within Sidamo (adjacent to Borana). Within a year, most students had returned to the cities and their campuses, but by mid-1976 they had achieved organization of 20,000 PAs with some 5 million households - a total which eventually grew to 23,500 PAs with perhaps 7.2 million members (Dejene 1987:34-35).

When the students came to Borana, they encountered a pastoral system where private land-holdings had not yet developed and where the Imperial administrations had ruled through a system of indirect representation, drawing elders from the two main moieties into which Boran clans are organized (Helland 1977:73). The students made the Borana indigenous madda (literally "well" or "spring") units in the basis for peasant associations, within a new hierarchy for territorial administration (Hogg 1990). Among the Borana, a madda unit "includes all people and all the animals" who used a given well complex (Helland 1977:73). Helland warns, however, that there was "no evidence" that the madda functioned to manage resources territorially, and that while the people did have a sophisticated system for managing their deep wells, madda groups "cannot deny access to pastures to any Borana .... limit the size of the Borana herds or do whatever else may be deemed necessary to manage the rangelands" (1977:74). Such restrictions are reserved for the adda, the term Borana use to describe the traditional customs all good Borana follow (Helland 1977:59).

The Derg regime did not attempt to relocate the Boran into these kebelle units (as PAs are called within Ethiopia). Instead, the people were simply registered into whichever PA was in their vicinity. PA members did get access to subsidized economic goods, such as grain and sugar, and clans began to use the PAs to buttress claims to land in the Somali areas of Region 5 (Hogg 1977:116). Among the Borana, the madda continued to operate as they usually had, but would now also contain overlapping PA officers (some of whom might be traditional elders) who dealt with the territorial administration as representatives for their area. Thus the indigenous systems for resource use co-existed with a more formalized PA system with its list of registered members and its committees. Peasant associations also became very useful as a basis for local defense during the war with the Somalis in 1977/78.

In 1979, the Derg regime issued a further directive calling for a stage-by-stage evolution of PAs into eventual producers co-operatives. As a first step, the PAs were instructed to pool their land resources; in a second step, they would combine using oxen and tools; and in a third and final step, smaller units would join to form a mechanized farm of some 4,000 ha (10,000 acres) containing about 2,000 households. At this final stage, the PA's service cooperatives would no longer be necessary.

In highland Arsi (north of Borana), Dejene found that producer co-operatives made up about 7% of rural households and cultivated 11% of the land (1987:74). Nevertheless, they got priority in the supply of inputs or services. Those visited by Dejene exhibited many problems and few benefits. Arsi's weak producers co-ops remained heavily dependent on the MOA's support. Their bad reputation became a major source of the peasant's negative attitudes towards the entire co-operative movement being promoted by Derg Officials (1987:138). Once again, since Borana was not a grain growing area, it escaped the ill-will which collectivization generated within Ethiopia's highland areas.

Instead, within Ethiopia's lowland areas, the Derg focused on the formation of service co-operatives. These were announced in 1978 as intermediaries to tie PAs to the government's recently established Agricultural Marketing Corporation (AMC) which traded in grain and agricultural inputs under supervision of the MOA (in 1979 the AMC was transferred to the Ministry of Domestic Trade). Each service cooperative combined the economic activities of between 3 and 10 PAs; Dejene says the average was 5 (1987:66). Each PA would send three-representatives to constitute the Service Co-op's annual general assembly, which in turn elects a 10 person executive committee every two years. Under the executive, come various specialized subcommittees (for finance, purchasing, and education). Peasants would thus become indirect members of any service co-op formed to which their PA was attached. In addition to some mandated grain trading (in the grain producing areas), the service co-op supplied agricultural inputs like fertilizers, improved seeds, and pesticides, and ran its own shop carrying salt, sugar, cooking oil, soap and simple consumer goods for its members (Dejene 1987:66-67). For these latter items, co-ops dealt with the Ethiopian Domestic Distribution Corporation (EDDC), which supplied consumer goods to cities as well as to the rural sector. Obviously, then, the Derg was creating a parallel system for trading which would have eliminated private traders entirely had the regime achieved its policy objectives.

With regard specifically to Ethiopia's lowland pastoralists (mainly the Afars, Somalis, and Boran peoples), the Derg regime inherited two major technical assistance projects planned by the FAO and LMD staff and implemented under World Bank funding. The first, known as the Second Livestock Development Project, provided US $5 million towards improvements in a proposed national network of stock routes. (Conceptually, it was very similar to the marketing component in Kenya's KLDP phase 2 project under implementation at this same time.) Three main stock routes were identified, by map rather than through local consultations (Desta 1993:46). One led from the NE rangelands to Asmara, another from the Jijiga lowlands to the railhead at Dire Dawa, and a third from the southern rangelands linking northwards to Shewa Province. Along these routes, which were meant to tie existing centers of traditional pastoralism to emerging markets for beef cattle, would be constructed stock-route facilities and market-places. The 1977 Ethio-Somali War destroyed most of what had been built, and, in any case, the mapped routes were purely hypothetical without much connection to where existing animals moved between lowland and highland areas. As Desta concludes, the Project (1973-81) "was never completed" because of poor management and cost overruns, and yielded "no significant impact" (1963:46).

While the SLDP was already launched at the time of the 1974 revolution, a much more ambitious and larger Third Livestock Development Project (TLDP) was also in final stages of preparation drawing upon FAO and World Bank staff working once again with the Livestock and Marketing Board. This time, the Project had as its specific focus Ethiopia's pastoral areas, each given its "development unit" on the ground to act as an executing agency. The Southern Rangelands Development Unit (or SORDU) is the one of most direct interest here, since it covered the 90,000 square kilometer area occupied by Borana pastoralists (from Mega southwards to the Kenya border). Then there was the Jijiga Rangelands Development Unit (JIRDU), covering 33,000 square kilometers within the much larger Somali lowlands of the southeast, and the North-East Rangelands Development Unit (NERDU) covering 75,000 square kilometers among Afar pastoralists to the northeast. The total project cost was set at US $43 million, with $27 million from IDA funds, $5.5 million from the African Development fund, and roughly $10 million from Ethiopia's own contributions (Desta 1993:46). The project, begun in 1976, was supposed to have ended in 1981 but because of the Ethio-Somali War was extended, to mid-1983 for NERDU and JIRDU and to mid-1984 for SORDU.

SORDU's field headquarters at Yavello served six districts (Woreda), and grew into a substantial complex of offices, workshops and staff housing. Its structure was divided into three tiers: six staff functions (audit, finance, etc.) which jointly supervised five main technical sections (for co-operative development, animal health, range development, marketing, and infrastructure) each with between 2-3 units (or programs). (Today the Project Manager has 295 staff but hardly any field budget, see our following section on the post-Derg policies.) A quick review of SORDU's main services offered within its project area includes: veterinary treatment (nearly 8 million animals treated); vaccinations (16 million innoculated against CBP or Rinderpest); 9 dipping vats serving 5 woreda; pond construction (880 stock ponds up to 1996); deep well improvement (43 dug out); experiments on bush control (none as successful as fire-using, banned by the Government); road maintenance over a network of bush tracks linking the six Woredas; three fattening ranches, 2 of which failed and were turned back to local communities; and the creation of functioning service co-operatives with livestock drugs, grain supplies, and consumer shops (5 of 6 in operation). All in all, the SORDU manager tries to support thirteen field programs. (These figures come from the1998 SORDU Project Manager.)

A formal listing of the Third Livestock Development Projects total achievements between 1976 and 1986 is given in Table 3 overleaf. At least in comparison to its equivalent units, JIRDU and NERDU, the SORDU field team appears to have been moderately successful. Only one component of the larger project was never implemented, the Awassa feedlot programme - because of a failure of its associated company (Desta 1993:47). Hogg (1997:15 fn) cites the World Bank's own (1991) completion report to the effect the Ethiopia "received little benefit in terms of increased livestock productivity, off-take or improved range resources" from these three, interlinked range development projects making up the TLDP. Desta notes a failure of ILCA to disseminate its linked research studies on SORDU, and the underuse of the $1.4 million allocated for staff development and training (1993:47). Compared with East Africa's other projects of the period, it seems almost incredible that only 5 project staff attended post graduate training, and only 14 participated in short courses or study tours (1993:47). He concludes that overall the project was overly ambitious: a complete change in production attitudes and practices was expected in too short a time.
 
Table 3. Ethiopia's Third Livestock Project Achievements.
(July 1976 to July 1986)
Achievements: SORDU JIRDU NERDU
Water Development
- stock ponds 95 26 9
- shallow wells - 77 -
- cisterns - 3 2
Range Management
- grazing reserve identification - 2 8000 km2
- formation of grazier associations - 74
Veterinary services
- vaccinations 11,977,000 9,239,000 1,897,000
Road Construction
- admin./trade road 1,137 km 1004 km 55 km
- access tracks 1,829 km 515 km 311 km
- maintenance 323 km 86 km -
Ranch Program
- establishment 3 - -
- steers purchased 6,071 - -
- steers sold 4,550 - -
Smallholder Fattening Program
- cattle purchased 3,804 5,197 -
- cattle distributed 3,706 4,956 -
Training and Information
- senior staff 6 2 1
- veterinary scouts 76 134 20
- range guards 50 54 60
Irrigation
- irrigation - - 1,380 ha
- water spreading - - 820 ha
Trials and studies
- trials in progress in progress in progress
- weather stations 10 9 6
 

SOURCE: Girma Bisrat (1987) Range Development and Research Proposal. In Proceedings of the First National Livestock Improvement Conference, 11-13 February 1987. Addis Ababa, Institute of Agricultural Research, p. 181.
 
 

When SORDU's official World Bank support ended in June, 1984, Ethiopia was in the midst of another major drought afflicting its northern lands. While Mengistu had consolidated his power within the Derg, his opponents now established liberation movements with support from outside Ethiopia's borders. In 1984-85, the Derg mounted a major resettlement and villagization program aimed at resettling drought-afflicted families from the north into better watered southern lands. Perhaps half a million people were shifted to live in compact and geometrically laid out villages (Goyder & Goyder 1988:103). In such settlements, government control became much easier, farmers could not conceal grain from the AMC, and liberation movements could be kept out. In fact, Mengistu had very specific and largely political objectives in mind for what was an immensely ambitious program (Giorgis 1989:288-292). By September 1986 Mengistu announced that 10% of Ethiopians, over four million people, were living in villages. Because of the many private costs associated with mass relocations, people resented the loss of their former independence even if their new homes were in less stressed areas.

It is an oversimplification to say that similar tactics were not considered for the pastoral areas further south, where the Oromo Liberation Front (OLF) became increasingly active. Dawit Wolde Giorgis tells of one attempt by the RCC to get destitute Somali nomads to settle along the Shebelle River (1989:282-84). At that time, the RCC was responsible for some 40,000 Somali ex-nomads who had spent two years in temporary camps. Dawitt decided the Wabi Shebelli's valley could allow the whole group to be resettled into irrigation-based farming communities. He admits, "I assumed that they would readily accept the idea" (p. 282), but found instead: "They asked lots of questions, some of them very odd..." Finally he took representatives by air to view the proposed site - a horrifying experience for the mainly old refugees selected. After a two day site visit, Dawit spent three days preparing the Somalis for their transfer. One hundred trucks were readied for the first day's move. However (1989:284):

Early in the morning, some of the drivers who had been sleeping in their trucks near the shelters came rushing up to me in absolute shock. The camp was deserted. The tukuls had been quickly and silently dismantled and over 10,500 people had vanished... We later learned they had crossed the border into Somalia.
 

Desta's reference cited above to the poor communication of ILCA's SORDU-linked research studies requires further comment. ILCA had wanted to locate its arid and semi-arid "systems study" within Ethiopia. It began a field program with this objective in 1975. The initial study involved 17 researchers, focusing on land use near Jijiga above Ethiopia's eastern Somali lowlands. The 1977 Ethio-Somali War forced a relocation of effort to the Afar's deserts in the NERDU territory. (It, too, had been abandoned by 1982 for security reasons.) Finally, in 1981, agreement was reached with the World Bank for a jointly funded research program looking at SORDU's activities in the Southern Rangelands, but also carrying out ILCA's main research effort on Boran pastoral productivity. Called the Joint Ethiopian Pastoral Systems Study (or JEPSS), this 1981-84 effort resulted in 12 research reports and 13 monographs, but no synthesis (a task eventually completed for ILCA by Coppock).

Meanwhile, Ethiopia became convulsed in another major drought which brought international attention and the inevitable questions why a large CGIAR center had nothing positive to offer to assist Ethiopia's suffering pastoralists. (Webb & von Braun estimate Ethiopia had over three million pastoralists, upon whom ILCA had already spent US $3.3 million by 1981.) Indeed, the International Livestock Centre for Africa (ILCA) was itself an anamoly within socialist Ethiopia. Agreed with the previous Imperial Govenment, ILCA became the CGIAR's major complex for studying African livestock, matching a sister institution, ILRAD, to study African livestock diseases. By the time ILCA was completed, however, it found itself conducting its two Ethiopian-based systems studies, on the highlands and on the arid and semi-arid zone, within a radical Marxist regime not noted for attention to scientific inquiry. ILCA's Director General was determined his new institution could contribute more than mere basic research to Ethiopia's rapidly evolving production systems.

The JEPSS studies showed calf nutrition was a key constraint limiting Boran productivity. In 1985, ILCA persuaded CARE to embark on a follow-up program carrying ILCA's scientific work into actual application. Thus was born CARE's Southern Sidamo Rangeland Development Project, headquartered in Yavello within what had been a food relief compound. Hogg (1992) gives us a fairly detailed description of CARE's Borana project (supplemented by field interviews in early 1998 at site in Yavello).

CARE's Borana program had the usual ambitious objectives: 1) to facilitate pastoralists' abilities to identify problems and solutions; 2) to strengthen the linkages between communities through pastoral associations and service co-operatives; and 3) to provide technical advice, materials and training in support of local solutions to the food security problem (Hogg 1992 citing CARE documents). What it did in practice was to search out small-scale incremental changes people could adopt if given small amounts of external assistance.

It tried a long list of potential innovations: well upgrading, pond desilting using animal-drawn scoops, cement cisterns and water catchments, haymaking, camel transport of water, below ground grain stores, credit for hand tools, even handicraft sales during times of emergencies. Some proved valuable and spread quickly. Others evidenced unsuspected technical problems and had to be abondoned. Obviously, in restrospect, CARE provided an essential "adaptive research" function to ILCA's distant scientists, and one which private service providers could not duplicate without going bankrupt. If the program had a major weakness, it was its dependence for funds on donor supplied grain sales, whose profits in Birr were then passed on to NGOs like CARE.

Hogg (1992:128-130) criticizes CARE for adopting a hierarchical project organization, complete with in expatriate project manager and a small cluster of offices and stores. He also argues CARE showed little direct cooperation with the government, and did not document its impacts through on-going monitoring. His larger criticisms, that CARE gave no attention to the problem of bush encroachment facing pastoralists and that it disregarded pastoral systemic linkages are accurate but unfair. A prohibition on burning has been an Ethiopian government policy since perhaps Imperial days. And no organization, not even ILCA, has paid attention to the socio-economic systems within which the Boran, Gabbra, and Garri households must operate when trying to safeguard sustainable livelihoods.

Alongside CARE's expanded activities in Borana came a major reorganization of Ethiopia's territorial administration following the establishment of Ethiopia as a "Peoples' Democratic Republic" in 1987. Ethiopia was divided into 30 regions, in which now Borana became a single unit (combining two former Awradjas) coupled with an upgrading of the quality of administrative staff. For the first time, ethnic Boranas assumed leadership positions in both the administration and the party (Helland et al. 1997:39). (This major territorial reorganization was still being implemented when the EPRDF victory in Addis Ababa brought the Derg government to an end.

The regime kept SORDU in operation using Ethiopian funds pending a possible new project by World Bank. However, in World Bank the old-style range development projects were being phased out while Bank officials debated the merits of projects based on pastoral associations (de Haan 1994). Finally, in 1988 the Bank with some reluctance incorporated a small (US $5 million) "pilot" component for SORDU within its mainly highlands oriented Fourth Livestock Development Project (Desta 1993:58). SORDU's administrative structure was reorganized into basically its present form (described above), while incorporating a monitoring and evaluation section into its range management activities and changing its main focus to be upon establishing effective service co-operatives run by pastoral communities themselves. Richard Hogg became SORDU's main external advisor, taking up residence in Yavello. There was much more focus on involving both SORDU staff and local beneficiaries within a collaborative mode of work. The Pilot Project's main official objective became "sustainability, participation of beneficiaries, and utilization of pastoralist social organization" (Desta 1993:48).

At the international level, the popular outcry in the UK and USA over a failure to assist Ethiopia's northern regions during the long 1982-86 drought had forced donors to deal with the Mengistu regime despite strong disapproval of Mengistu's policies. It also made apparent the futility of reacting to such a major catastrophe on a country-by-country basis. The drought afflicted the entire region: Sudan, Eritrea (then at war for its independence from Ethiopia), Ethiopia, Djibouti, Somalia, and Kenya. In early 1986, these countries joined by Uganda (Eritrea did not officially enter until 1993) formed the Inter-Governmental Authority on Drought and Development (IGADD), with its headquarters in Djibouti.

These seven states form a contiguous block where pastoralists often move across borders and whose lowlands often experience famine simultaneously. In 1991, it was estimated by the UN Special Emergency Programme for the Horn of Africa (SEPHA) that the region contained 9 million refugees and displaced people (Markakis 1998:16). Its poorest states were also highly dependent on food imports to meet staple food needs. In 1989, the dependency ratios of food imports to total food production was for Sudan 28%, for Djibouti 100%, for Eritrea 70%, for Somalia 28%, and for Ethiopia 26%. This contrasts with 6% for Kenya (Markakis 1998:85).

IGADD's executive is constituted of the Heads of State from its member countries. IGADD's Executive Secretary serves a four year term (once renewable) and reports to its council of Ministers. The poverty of IGADD's member states gives it virtually no reliable funds of its own. Instead, it looks to donors for support of its many infrastructural projects, a combined "wish list" from all its member states. In 1990, just prior to Mengistu's fall, IGADD approved a food strategy for the region, giving priority to four areas: 1) increasing local food production, 2) better food marketing, 3) increased food consumption, and 4) better food policy management. It also established with donor support a region-wide famine early warning sy