“ Food and Agriculture in Ethiopia P R O G R E S S A N D P O L I C Y C H A L L E N G E S Paul Dorosh and Shahidur Rashid Editors Dorosh • Rashid Editors Food and Agriculture in Ethiopia Food and Agriculture in Ethiopia This book is published by the University of Pennsylvania Press (UPP) on behalf of the International Food Policy Research Institute (IFPRI) as part of a joint-publication series. Books in the series pre- sent research on food security and economic development with the aim of reducing poverty and eliminating hunger and malnutrition in developing nations. They are the product of peer-reviewed IFPRI research and are selected by mutual agreement between the parties for publication under the joint IFPRI-UPP imprint. Food and Agriculture in Ethiopia Progress and Policy Challenges EDITED BY PAUL A. DOROSH AND SHAHIDUR RASHID Published for the International Food Policy Research Institute University of Pennsylvania Press Philadelphia Copyright © 2012 International Food Policy Research Institute All rights reserved. Except for brief quotations used for purposes of review or scholarly citation, none of this book may be reproduced in any form by any means without written permission from the publisher. Published by University of Pennsylvania Press Philadelphia, Pennsylvania 19104-4112 www.upenn.edu/pennpress Library of Congress Cataloging-in-Publication Data CIP DATA TO COME Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 v List of Figures vii List of Tables ix List of Boxes xv Foreword xvii Acknowledgments xix Acronyms and Abbreviations xxiii Glossary xxvii 1 Introduction 1 paul dorosh and shahidur rashid PART I Overview and Analysis of Ethiopia’s Food Economy 2 Ethiopian Agriculture: A Dynamic Geographic Perspective 21 jordan chamberlin and emily schmidt 3 Crop Production in Ethiopia: Regional Patterns and Trends 53 alemayehu seyoum taffesse, paul dorosh, and sinafikeh asrat gemessa 4 Seed, Fertilizer, and Agricultural Extension in Ethiopia 84 david j. spielman, dawit kelemework mekonnen, and dawit alemu 5 Policies and Performance of Ethiopian Cereal Markets 123 shahidur rashid and asfaw negassa 6 Livestock Production and Marketing 159 asfaw negassa, shahidur rashid, berhanu gebremedhin, and adam kennedy Contents vi Contents 7 Patterns in Foodgrain Consumption and Calorie Intake 190 guush berhane, linden mcbride, kibrom tafere hirfrfot, and seneshaw tamiru PART II Major Agricultural and Food Policy Interventions in Ethiopia 8 Implications of Accelerated Agricultural Growth for Household Incomes and Poverty in Ethiopia: A General Equilibrium Analysis 219 paul dorosh and james thurlow 9 Disaster Response and Emergency Risk Management in Ethiopia 256 john graham, shahidur rashid, and mehrab malek 10 Targeting Food Security Interventions in Ethiopia: The Productive Safety Net Programme 280 sarah coll-black, daniel o. gilligan, john hoddinott, neha kumar, alemayehu seyoum taffesse, and william wiseman 11 The Evolving Role of Agriculture in Ethiopia’s Economic Development 318 paul dorosh Contributors 329 Index 335 vii 2.1 Characteristics of rainfall and altitude 23 2.2 Dominant cereals in Ethiopia 25 2.3 Map of woreda domain assignments, 1994 30 2.4 Map of woreda domain assignments, 2007 31 2.5 Travel time to a city, 1994 and 2007 35 2.6 Fixed-line and cellular telephones, 2003, 2008, 2009, and 2010 39 3.1 Shares of area cultivated, by farm size and agroecology, 2007/08 57 3.2 Per capita cereal consumption and imports, 1961/62–2008/09 64 4.1 Total and per capita grain production and grain yields, 1991/92–2007/08 89 4.2 Area under improved seed application and quantity of improved seed distributed, cereals only, 1993/94–2007/08 93 4.3 Area under improved seed application, main cereal crops, 1995/96–2007/08 95 4.4 A schematic of the Ethiopian seed system 97 4.5 Hybrid maize seed distribution, by type of supplier, 1993–2008 97 4.6 Basic seed demand and supply for maize hybrid multiplication, 2006–08 99 4.7 Raw seed production, Ethiopian Seed Enterprise, 2000–08 99 4.8 Hybrid maize seed–grain price ratios and real seed prices, 1991/92–2007/08 101 4.9 Fertilizer imports, 1996–2008 105 4.10 Fertilizer use intensity, 1996–2008 105 4.11 Fertilizer value–cost ratios, 1992, 1997, 2004, and 2008 107 4.12 Fertilizer import shares, by type of importer, 1996–2008 110 5.1 Trends in road development in Ethiopia, 1951–2007 131 5.2 Cellular phone ownership per 100 people in Ethiopia and its neighbors, 2000–08 133 5.3 Cereal value chain map involving traditional market channels in Ethiopia 135   6.1  Outflow of cattle, sheep, and goats, 2009  165 Figures viii Figures 7.1 Nonparametric regression estimates of per capita food expenditure, 2004/05 193 7.2 Nonparametric Engel curve, 2004/05 193 7.3 Share of major cereals in total food expenditures in Ethiopia, by income group and rural–urban place of residence, 2004/05 199 7.4 Distribution of rural population, by region and livelihood, 2008 206 7.5 Ethiopia’s rural agropastoral population, by region, 2008 206 7.6 Ethiopia’s rural pastoral population, by region, 2008 207 7.7 Share of major cereals in total food expenditures, 1994/95, 1999/2000, and 2004/05 211 7.8 Share of major cereals in total food expenditures across income quintiles, 1994/95, 1999/2000, and 2004/05 211 7.9 Ethiopia Regional Hunger Index, 2005 214 8.1 Zones in the Dynamic Regional Economywide Model of Ethiopia (DREME), with city sizes, road types, and agroecological zones indicated, 2009 222 8.2 National poverty headcount results from the model scenarios, 2005–15 230 8.3 Current, expected, and targeted crop yields, 2005 and 2015 232 8.4 Changes in real market prices from the baseline under the “all agriculture” scenario, 2005–15 239 10.1 Incidence of PSNP, public works, and direct support transfers, by region and per capita or total consumption decile, 2007 306 ix 1.1 Summary of major historical events in Ethiopia, 1890–2010 3 1.2 Ethiopia: Selected economic variables, 1981–2009 10 1.3 Ethiopia: Selected economic and social indicators, 1981–2009 11 1.4 East African countries: Selected indicators, 2009 13 2.1 Traditional ecological zones and altitudinal ranges of selected crops 24 2.2 Domain summary statistics: Share of national totals by domain factor, 1994 and 2007 (percent) 31 2.3 Domain summary statistics: Domain share of national totals, 1994 and 2007 (percent) 32 2.4 Percentage of regions’ populations connected to a city of at least 50,000 people, by access time, 2007 36 2.5 Percentage change in regions’ connection to a city of at least 50,000 people, by access time, 1994–2007 36 2.6 Electricity-generating capacity, 1958–2011 (annual averages) 38 2.7 Fixed-line and cellular telephones in Ethiopia compared to all of Africa, 2003, 2008, 2009, and 2010 39 2.8 Agglomeration index: Percentage of people considered urban by region, 1984, 1994, and 2007 40 2.9 Alternative urbanization estimates, 1984, 1994, and 2007 41 3.1 Crop area and production (smallholder farms, meher season), 2004/05–2007/08 55 3.2 Total area cultivated, by farm size and agroecology, 2007/08 56 3.3 Crop area, by season and size of farm, 2007/08 58 3.4 Crop production, by season and size of farm, 2007/08 59 3.5 Crop yields, by season and size of farm, 2007/08 60 3.6 Cereal area, yield, and production, decade averages and growth rates, 1961/62–2008/09 63 3.7 Average cereal production, area covered, and yield, by crop, 2004/05–2007/08 (1997–2000 E.C.) 66 3.8 Cereal area cultivated, decade averages, 1961/62–2008/09 67 Tables x Tables 3.9 Median contributions to changes in quantity of cereal production, by crop and year, 1998/99–2007/08 (percent) 70 3.10 Median contributions to changes in cereal revenue, by crop and development domain, all periods, 1998/99–2007/08 (percent) 71 3.11 Median contributions to changes in cereal revenue, by crop and period, 1998/99–2007/08 (percent) 72 3.12 Median contributions to changes in cereal revenue, by development domain and period, 1998/99–2007/08 (percent) 73 3.13 Area cultivated, share of all crops, and growth rate (smallholders, meher season), 1994/95–2008/09 75 3.14 Fertilizer application, by crop, private holdings, 1997/98, 2001/02, 2007/08 78 3.15 Area under improved farm management, by practice and crop, private holdings, 1997/98, 2001/02, 2007/08 78 3.16 Cereal yield, by country and year, 2006–08 80 3.17 Yields of maize, wheat, and sorghum, by country and year, 2006–08 (kilograms per hectare) 81 4.1 Policy regimes and development programs in agricultural input systems and markets, 1957–95 86 4.2 Policy regimes and development programs in agricultural input systems and markets, 1995–present 90   4.3   Yields in on-farm field trials using improved seed and fertilizer  versus farmers’ yields using traditional cultivation practices, 1993–2008 (metric tons per hectare) 91 4.4 Seed supply shortfalls in Ethiopia, 2005–08 95 4.5 Fertilizer value–cost ratios, 1992–2008 108 4.6 Major problems of the fertilizer supply system, 2004 and 2009 109 4.7 Development agents and farmer training centers, 2008 112 4.8 Agricultural Technical and Vocational Education and Training (ATVET) college graduates, 2003–04 to 2007–08 115 4.9 Frequency of visits by an extension agent during the previous main growing season, 2004 and 2009 115 5.1 Number of trucks and telephone subscriptions and kilometers of road network, by type (thousands) 130 5.2 Size and distribution of licensed commercial trucks, by operator groups, 2006 133 5.3 Broad structural changes in Ethiopian cereal markets since the 1960s 138   5.4   Processing capacity and regional distribution of flour mills in  Ethiopia, 2007–08 141 5.5 Cooperative membership and use of cooperatives for cereal sales, 2005 and 2008 142 Tables xi 5.6 Summary of cereal market integration studies in Ethiopia, 1992–2007 144 5.7 Summary of wholesale price seasonality indexes of staple cereals over time, 1980–2010 146 5.8 Cereal price variability over time, 1983–2008 148 5.9 Changes in the real costs and margins of the grain trade since 1996 150 5A.1 Chronology of government grain market interventions in Ethiopia, 1950–2007 152 6.1 Trends in average number of animals, by livestock species, 1970–2008 162 6.2 Regional distribution of cattle, by gender and age structure, 2008/09 (percent) 163 6.3 Distribution of livestock ownership, by region and species, 1999–2000 and 2004–05 (percent) 168 6.4 Livestock productivity in African countries versus the world, 1999–2008 170 6.5 Livestock production and productivity, 2000, 2004, and 2008 171 6.6 Increases in livestock feed prices, by feed type, 2004–08 173 6.7 Distribution of livestock producers, by market participation regime and species, 2003 or 2004–05 (percent) 174 6.8 Export volume and value of major livestock exports, 1970–2008 177 6.9 Value and share of exports of live animals from Ethiopia, by destination, 2007–08 179 6.10 Livestock off-takes and valuation, by species, 2007–08 182 7.1 Per capita calorie contributions of food items, by rural–urban place of residence, 2004/05 191 7.2 Budget shares and own price and income (expenditure) elasticities of demand for selected food items, by rural–urban place of residence, 2004/05 195 7.3 Per capita budget shares, consumption quantities, costs, and calorie shares of major staples, by rural–urban place of residence, 2004/05 196 7.4 Share of major cereals in total food expenditures, by income group and rural–urban place of residence, 2004/05 (percent) 199 7.5 Per capita share of the total quantity of cereals and enset consumed, by agroecological zone and rural–urban place of residence, 2004/05 (percent) 200 7.6 Annual per capita total expenditure, shares of major cereals and enset in total food expenditures, by region and rural–urban place of residence, 2004/05 (percent) 202 xii Tables 7.7 Calories consumed and cost of calories, by region, 2004/05 203 7.8 Calories consumed and cost of calories, by income quintile, agroecological zone, and rural–urban place of residence, 2004/05 204 7.9 Source of rural food consumption, by livelihood zone and region, 2008 208 7.10 Source of rural food consumption by the very poor and poor, 2008 209 7.11 Consumption of calories per adult equivalent per day, by region and rural–urban place of residence, 1994/95, 1999/2000, and 2004/05 212 7.12 Ethiopia Regional Hunger Index and underlying components, by region and rural–urban place of residence, 2000 and 2005 213 8.1 Sectors in the Dynamic Regional Economywide Model of Ethiopia 221 8.2 Selected income elasticities of goods and services in the Dynamic Regional Economywide Model of Ethiopia, by rural–urban place of residence, 2004/05 224 8.3 Sector growth results from the model scenarios, 2009–15 (percent) 225 8.4 Production targets for various crops in the baseline and agricultural growth scenarios, 2005–15 227 8.5 Household poverty results for rural and urban regions from the model scenarios, 2005, 2008, and 2015 (percent) 234 8.6 Regional growth results for various crops under the model scenarios, 2009–15 (percent) 235 8.7 Household average consumption shares of various goods and services, by rural–urban place of residence and poverty status, 2009 (percent) 238 8.8 Poverty–growth elasticities from the model scenarios: Percentage change in the poverty headcount rate from a 1 percent increase in national agricultural GDP, by rural–urban place of residence and sector, 2009–15 241 8A.1 Computable general equilibrium model sets, parameters, and variables 247 8A.2 Computable general equilibrium model equations 251 9.1 A historical account of Ethiopian famines and major food shortages, geographic locations, and attributed causes, 1888–2009 258 9.2 Quarterly average cereal stocks of the Emergency Food Security Reserve Administration (EFSRA), 2004–08 (metric tons) 269 9.3 Major droughts and drought-related consequences, 1965–2008 272 9.4 Food production and food aid in Ethiopia, 1996–2008 273 Tables xiii 10.1 Targeting criteria for public works, by priority, 2006 (percentage of communities reporting criteria used) 284 10.2 Targeting criteria for direct support, by priority, 2006 (percentage of communities reporting criteria used) 285 10.3 Criteria used to select public works participants, by region and year, 2006 and 2008 (percent) 286   10.4   Criteria used to select direct support beneficiaries, by region and  year, 2006 and 2008 (percent) 287 10.5 Regional participation of households in the public works component of the PSNP, 2006–08 (percent) 288 10.6 Correlates of access to public works, 2006–08 291 10.7 Household participation in the direct support component of the PSNP, by region, 2006–08 (percent) 293 10.8 Household characteristics, by participation in the direct support component of the PSNP and region, 2006–08 (percent) 294 10.9 Household characteristics correlated with receipt of direct support, by year, 2006–08 298 10.10 Household characteristics correlated with per capita expenditures in the Household Income, Consumption, and Expenditure Survey, by region, 2004–05 302 10.11 Coady–Grosh–Hoddinott index values by region, consumption decile, and transfer type, 2007 304 10.12a Size of PSNP transfers relative to mean transfers, by region and consumption decile, 2007 (percent) 308 10.12b Size of public works transfers relative to mean transfers, by region and consumption decile, 2007 (percent) 308 10.12c Size of direct support transfers relative to mean transfers, by region and consumption decile, 2007 (percent) 309 10.13 Local-level targeting-related outcomes: Distributions and means, by region, 2008 310 10.14 Local-level targeting-related outcomes: Characteristics of correlates, 2008 311 10.15 Correlates of locality-level targeting-related outcomes, 2008 312 10.16 Correlates of targeting performance, public works transfers to the poorest two quintiles as measured by predicted per capita household consumption, 2008 315 11.1 Average sectoral growth rates under the Plan for Accelerated and Sustained Development to End Poverty (PASDEP) and the Growth and Transformation Plan (GTP), 2005/06–2014/15 (percent) 319 xv 11.1 Ethiopia’s Growth and Transformation Plan, 2010/11–2014/15 320 Box xvii Food security issues in Ethiopia are extremely complex because of large varia- tions across space and over time related to agroecologies, weather shocks, gov- ernment policies, and other factors. In this context, Ethiopia’s agricultural and food policies are crucially important, having profound effects on tens of mil- lions of low-income people throughout the country. Following major famines in the 1970s and 1980s, the country has made huge strides in the past two decades—increasing food production, promoting market development, building an effective safety net for millions of food- insecure households, improving its disaster food emergency response capabili- ties, and laying a foundation for future economic growth. Yet there is much to be done to reduce the still-high levels of food insecurity. This book describes these past developments and the policies likely to shape future trends. The analysis it contains builds on and extends a long history of IFPRI work in Ethiopia in collaboration with the Ethiopian Development Research Institute and researchers at Addis Ababa University, including work on causes of and responses to major famines in Ethiopia in the 1970s and 1980s, rural poverty dynamics, analyses of cereal markets that helped lead to the cre- ation of the Ethiopian Commodity Exchange, and the impact of the Productive Safety Net Programme. The analysis presented in this book shows the importance of continued investments in increasing agricultural productivity in order to promote both economic growth and poverty reduction. The book also highlights five major  challenges Ethiopia must face to achieve high economic growth rates, reduce poverty, and enhance household food security: sustaining growth in crop and livestock production, increasing market efficiency, providing effective safety  nets, maintaining macroeconomic incentives and stability, and managing the rural–urban transformation. If these challenges are met, there is good reason to expect that Ethiopia will enjoy further progress in reducing poverty and enhanc- ing food security. Shenggen Fan Director General, International Food Policy Research Institute Foreword xix We owe a debt of gratitude to many people and organizations that contributed to the content of this book and the work of the Ethiopia Strategy Support Pro- gram (ESSP). ESSP is a collaborative program of research, capacity building, policy analysis, and knowledge dissemination by IFPRI and the Ethiopian Development Research Institute (EDRI). We thank Ato Newai Gebre-ab, chief economic adviser to the prime min- ister of Ethiopia, and director of EDRI, for his leadership and active support as chairman of ESSP’s National Advisory Committee (NAC). His encouragement of rigorous and objective research and analysis has set the tone for all the work of the project and the material presented in this book. Other members of the NAC also provided valuable overall guidance on research themes and insights into key aspects of policy, including Dr. Abera Deressa (former state minister of agriculture), Ato Mekonnen Manyazewal (former state minister of finance  and current minister of industry), Samia Zekaria (director general, Central Sta- tistics Agency), Dr. Solomon Assefa (director general, Ethiopian Institute for Agricultural Research), and Ato Bulbula Tulle (Hawas Agri-business). Eleni Gabre-Madhin, program leader of the first phase of ESSP from 2003  to 2007, established a firm foundation for the research conducted under phase  2 from 2008 to the present. She also conducted important research on cereal markets cited in this book, which contributed to the establishment of the Ethio- pian Commodity Exchange. Hashim Ahmed (director of the Macro Economics Research and Trade Policy Directorate at EDRI) and the team of researchers that produced the 2004–05 Ethiopian Social Accounting Matrix—including Ayanaw Amogne, Tewodros Tebekew, Birouke Teferra, and Eyasu Tsehaye of EDRI as well as Sherman Robinson and Dirk Willenbockel—made possible the computable general equilibrium analysis presented in Chapter 8. Many other researchers provided comments and suggestions on various parts of this book and related background papers that have significantly strength- ened the work, including Gezahegn Ayele, Abdur Rahman Beshir, Yoseph Geta- chew, Kindie Getnet, Derek Byerlee, Kristin Davis, Dennis Friesen, Steven Acknowledgments xx Acknowledgments Haggblade, Ruth Vargas Hill, M. A. Jabbar, Thomas Jayne, Nicholas Minot, Robert Tripp, and Elias Zerfu. We are likewise grateful to participants in vari- ous sessions of the Ethiopian Economics Association annual conferences for their feedback on drafts of some of the work presented in this book. We also owe a great debt of gratitude to two anonymous reviewers whose careful cri- tiques  of  the first  draft  of  this  book were  extremely  helpful  in  producing  a  much-improved revised manuscript. Alemayehu Seyoum Taffesse, coauthor of two book chapters, also made major contributions to the research and analysis in many other chapters. More- over, his guidance and insights have been invaluable to most aspects of ESSP research and policy analysis. Bart Minten, current program leader of ESSP, gave detailed suggestions on early drafts of the chapters that were instrumental in moving the manuscript of the book forward at a crucial stage in the process. In addition, numerous research assistants and analysts contributed to the research, and many are coauthors of various background papers. These include Meron Assefa, Befekadu Behute, Ermias Engida, Belay Fekadu, Mekamu Kedir, Solomon Lemma, Wondimsiamregn Mekasha, Zelekawork Paulos, Nigussie Tefera, and Yisehac Yohannes. Valuable administrative support for the editing and production of this book and various research papers was pro- vided by the ESSP staff, particularly Tigist Mamo, Nahume Yadene, Mahalet Mekuria, and Yetnayet Begashaw. We are also grateful for the editorial and secretarial support provided by Indra Lamoot, Lisa Moorman, Martha Negash, and Etenesh Yitna. Mehrab Malek very ably undertook the roles of production editor, research assistant, and coauthor at various stages of the work. His detailed and careful work has made a major contribution to the quality of every chapter. As usual, any remaining errors and omissions in the manuscript are the responsibilities of the chapter authors and the volume editors. The World Bank, the UK Department for International Development (DFID), and the United States Agency for International Development (USAID) provided funds for the analysis that appears in Chapter 9. The authors thank Berhanu Wolde-Michael and Wout Soer for their guidance and the government and donor staff supporting the Productive Safety Net Programme for their help- ful comments and analytical input into this work. The results presented in Chap- ters 5, 6, and 9 are partly based on research grants provided by the Bill and Melinda Gates Foundation and the World Food Programme of the United Nations. We are thankful to Alemayehu Assefa, GebreEgziabher Abey, and Tadesse Bekele for sharing valuable insights about the food policies and func- tioning of markets. Finally, we thank the donors who have made the work of the ESSP pos- sible: the Canadian International Development Agency, DFID, USAID, and the Development Cooperation of Ireland. Acknowledgments xxi It is our hope that this book, produced with the support and efforts of all those mentioned here, along with those of the chapter authors, will prove useful to students, researchers, development specialists, and government officials and  will ultimately contribute in some way to increased agricultural growth and enhanced food security in Ethiopia. xxiii AAU Addis Ababa University ADLI Agricultural Development–Led Industrialization AISCO Agricultural Input Supply Corporation AISE Agricultural Input Supply Enterprise AMC Agricultural Marketing Corporation BBC British Broadcasting Corporation BoARD Bureau of Agriculture and Rural Development CAADP Comprehensive Africa Agriculture Development Programme CES constant elasticity of substitution CET constant elasticity of transformation CFSTF Community Food Security Task Force CGE computable general equilibrium CGH Coady–Grosh–Hoddinott CLVI Cuddy La Valle index CPI consumer price index CRED Centre for Research on Epidemiology and Disaster CSA Central Statistical Agency CTF community therapeutic feeding CV  coefficient of variation DA development agent DAP diammonium phosphate DFID Department for International Development (UK) DPPC Disaster Prevention and Preparedness Commission DREME Dynamic Regional Economywide Model of Ethiopia DRMFSS Disaster Risk Management and Food Security Service DS direct support EA enumeration area EARO Ethiopian Agricultural Research Organization E.C. Ethiopian calendar ECX Ethiopian Commodity Exchange EDRI Ethiopian Development Research Institute Acronyms and Abbreviations xxiv Acronyms and Abbreviations EFSRA Emergency Food Security Reserve Administration EFSS Ethiopian Food Security Survey EGB Ethiopian Grain Board EGC Ethiopian Grain Council EGS Employment Generation Scheme EGTE Ethiopian Grain Trade Enterprise EIAR Ethiopian Institute of Agricultural Research EOS Enhanced Outreach Strategy EPRDF Ethiopian People’s Revolutionary Democratic Front ERHI Ethiopia Regional Hunger Index ERHS Ethiopian Rural Household Survey ERSS Ethiopia Rural Smallholder Survey ESE Ethiopian Seed Enterprise ESRC Economic and Social Research Council ESSP Ethiopia Strategy Support Program ETB Ethiopian birr EU European Union EW early warning FAO Food and Agricultural Organization FAOSTAT FAO Statistical Database FCU foreign currency unit FTC farmer training center GDP gross domestic product GIS geographic information system GoE Government of Ethiopia GTP Growth and Transformation Plan ha hectare HABP Household Assets Building Program HAP high agricultural potential HICES Household Income, Consumption, and Expenditure Survey HMA high market access IFPRI International Food Policy Research Institute KFSTF Kebele Food Security Task Force LAP low agricultural potential LIU Livelihoods Integration Unit LMA low market access masl meters above sea level mm millimeter MoA Ministry of Agriculture MoARD Ministry of Agriculture and Rural Development MoFED Ministry of Finance and Economic Development MoH Ministry of Health MPP-I Minimum Package Program I Acronyms and Abbreviations xxv MPP-II Minimum Package Program II n.a. not available NAC National Advisory Committee (ESSP) NAEIP National Agricultural Extension Intervention Program NEPAD New Economic Partnership for Africa’s Development NGO nongovernmental organization NPDPM National Policy on Disaster Prevention and Management OCHA  Office for the Coordination of Human Affairs OFSP Other Food Security Program OTC outpatient therapeutic care PA peasant association PADEP Peasant Agricultural Development Program PADETES Participatory Demonstration and Training Extension System PASDEP Plan for Accelerated and Sustained Development to End Poverty PBM parity-bound model PIM Program Implementation Manual PPP purchasing power parity PSNP Productive Safety Net Programme PW public works R&D research and development RRC Relief and Rehabilitation Commission SAM social accounting matrix SG2000 Sasakawa Global 2000 SIDA Swedish International Development Agency SNNPR Southern Nations, Nationalities, and People’s Region SWC soil and water conservation T&V training and visit TFP total factor productivity TLU tropical livestock unit TPLF Tigray Peoples’ Liberation Front UK United Kingdom UN United Nations US United States USAID United States Agency for International Development US$ US dollar VCR value–cost ratio WFP World Food Programme WFSTF Woreda Food Security Task Force WMS Welfare Monitoring Survey xxvii aja Amharic for Emmer wheat ayib traditional cheese belg secondary rainy season (February to May) bereha desert Borana an administrative zone in Southern Ethiopia, located on the border with Kenya bulla highest-quality processed enset chat a stimulant product dega highland agroecological zone (2,300 to 3,200 meters above sea level) Derg literally “committee”; Ethiopia’s ruling government, 1974–91 enset “false banana” (Ensete ventricosum), a perennial tuber grown as a staple foodcrop, especially in southern Ethiopian highlands ergo sour milk gebbar private land ownership contracts gult grants of land given to individuals injera Ethiopia’s main national dish: a fermented pancake traditionally made with teff (although often made with cheaper grains) kebele the smallest administrative unit in the local government system kocho a bread made out of processed enset kolla lowland agroecological zone (less than 1,500 meters above sea level) Megabit March meher the primary rainy season and therefore the primary crop season mengist government Meskerem September Nehase August neug niger seed, a common oilseed crop in Ethiopia ras king (literally “head” in Amharic) rist a communal land tenure system Glossary xxviii Glossary samon land owned by the church teff a cereal crop related to millet (Eragrostis tef ) tsimdi an amount of land equivalent to 0.25 hectare weyna dega midelevation agroecological zone (1,500 to 2,300 meters above sea level) woreda administrative level under the zone level wurch frost, or extremely highland area (3,200 to 3,700 meters above sea level) Yekatit February Food and Agriculture in Ethiopia 1 Ethiopia is often perceived as a country of droughts, widespread poverty, and economic stagnation. Indeed the country experienced severe famines in both the 1970s and the 1980s, which resulted in a depletion of household assets and savings and caused excess mortality, estimated at 250,000 in 1972–74 and 590,000 in 1984–85 (Africa Watch 1991; de Waal 1997).1 More localized food shortages have often been less documented, such as the one in Somali region during 1999–2000 whereby an estimated 100,000 people died following three consecutive years of drought (Hammond and Maxwell 2002). Other serious production  shortfalls  related  to  droughts,  such  as  that  in  2003,  significantly  reduced the food production and consumption of millions of households. More- over, even in normal years, the level of food insecurity is high, with an esti- mated 44 percent of the population undernourished, 35 percent of children under five years of age underweight, and 11 percent of children dying before  the age of five (von Grebmer et al. 2010). Yet much of Ethiopia differs sharply from the grim view suggested by these figures. Most of the rural population resides in rainfall-sufficient areas in  which harvests are normal in most years. Nationally, the available data suggest that food production and availability are increasing due to increases in area cultivated and, in 2005–09, increases in yield. Outside the agricultural sector itself, massive investments in roads and the spread of cell phone technology have greatly increased access to markets, urban centers, services, and informa- tion for tens of millions of rural Ethiopians. And per capita incomes increased by over 50 percent from 2001 to 2009 (World Bank 2010). Indeed, the reality of Ethiopia’s agriculture and food security situation is complex because of variations across space within Ethiopia as well as varia- tions over time due to changes in policies, weather shocks, and other factors. A complete picture of Ethiopia’s agriculture and food security must include both 1. Estimates of famine deaths vary widely. The Centre for Research on Epidemiology and Disaster (CRED 2011) reports that 100,000 people died in the 1972–74 famine and 300,000 people died in 1984–85. 1 Introduction PAUL DOROSH AND SHAHIDUR RASHID 2 Paul Dorosh and Shahidur Rashid the very serious acute and chronic food insecurity problems faced by many Ethiopians and the progress achieved in other areas and at the national level. Moreover, the country faces major strategic questions regarding the role of agriculture in its overall development strategy. Agricultural Development– Led Industrialization has been a foundation of government economic policy since 1992. Up until the early 2000s, however, the relative stagnation of cereal yields suggested to many that significant expansion of smallholder production  (particularly for cereals) was severely limited by relative land shortage in the Ethiopian highlands, limited potential for irrigation, inadequate marketing infrastructure, and a weak seed sector. In addition, because of Ethiopia’s high costs for transport and marketing both internally and to external international markets, neither export nor import trade is profitable for most cereals (though  imported wheat has been an exception in some years). Thus, there has been concern that agricultural growth alone (even with sizable growth linkages to the nonagricultural sector) cannot generate sufficient domestic demand to keep  agricultural prices from falling substantially and putting a brake on further growth (World Bank 2006; Dercon, Hill, and Zeitlin 2009). The official Central  Statistical Agency  (CSA)  data  indicating  large  in- creases in both cereal area cultivated and yields from 2003–04 to 2009–10 (Ethiopia, CSA 2005–10) have not settled the debate, however. The manner in which this gain in cereal productivity has been achieved remains a puzzle given land constraints, the uncertain effectiveness of agricultural extension, and the slow dissemination of improved seed varieties to complement increased fertil- izer use (Dercon, Hill, and Zeitlin 2009). Equally important for future policy are whether further gains in productivity are realistic and the extent to which future agricultural growth can generate sufficient income growth (in both the  agricultural and the nonagricultural sectors) to ensure there is sufficient demand  to prevent a price collapse. This book is designed to inform this policy debate by documenting the state of the agricultural and food economy of Ethiopia through the first decade  of the 2000s and to highlight major structural features that will greatly influence  the outcome of future development strategies in Ethiopia. In showing the com- plexity of Ethiopian agriculture, ongoing constraints to growth, the substantial progress achieved, and the scale of the food security challenge that remains, the book can provide a solid foundation for students, researchers, policy analysts, and decisionmakers. It builds on previous research documenting Ethiopia’s political, economic, and agricultural transformation while aiming to provide a realistic and balanced foundation for the key aspects of major food policy issues in Ethiopia. It extends the discussion presented in People of the Plow (McCann 1995) to assess the dramatic investments in infrastructure and agricultural development (that have occurred since 1990) on current economic growth. Overall, the tone of the book is cautiously optimistic, bearing in mind the trends Introduction 3 and manifestations of previous food shortages and mismanaged food policy decisions documented in Famine and Food Security in Ethiopia (Webb and von Braun 1994). The Long Path from Famine to Food Security Although Ethiopia has experienced major food production shortfalls over the past two decades, it has managed to avoid large-scale, countrywide famines such as those in 1972–74 and 1984–85 (Table 1.1). In part, this can be explained using the lens of the basic components of food security: availability, access, and utilization (nutrition). The occurrence of a famine—a complete collapse in food security on a large scale—however, involves much more than the proxi- mate cause of a drought or other disruption in food supply. A complex inter- action of short- and long-term policies related to agricultural investments and markets, capacities of the government to respond, household coping strategies, TABLE 1.1 Summary of major historical events in Ethiopia, 1890–2010 Time frame Event March 1896 Ethiopian army under Menelik II defeats Italian troops at the Battle of Adwa March 1929 Ras (King) Tafari proclaimed as Emperor Haile Selassie 1936–41 Ethiopia occupied by Italian troops; Ethiopian imperial family exiled 1972–74 Droughts and famine in northern Ethiopia and parts of southa cause quarter million deaths September 1974 Haile Selassie deposed by the military 1974–91 Mengistu Haile Mariam leads Derg government; state takes ownership of private land and enterprises 1984–85 Droughts and famine cause 1 million deaths May 1991   Mengistu flees Ethiopia; Ethiopian People’s Revolutionary  Democratic Front (EPRDF) troops enter Addis Ababa 1991–present EPRDF government led by Meles Zenawi May 1993 Eritrea declares its independence following a referendum May 1998–June 2000 War between Ethiopia and Eritrea 2005 Violence erupts after national elections 2005–10 Five-year plan: Plan for Accelerated and Sustained Development to End Poverty 2010   EPRDF wins national elections; new five-year Growth and  Transformation Plan announced sources: Authors; Webb and von Braun (1994); Marcus (2002). aHarerge, Bale, Sidamo, and Gamo Gofa regions. 4 Paul Dorosh and Shahidur Rashid and other factors were key to Ethiopia’s previous famines. Moreover, chronic malnutrition and periodic localized severe food insecurity continue to affect tens of millions of Ethiopians.2 Adequate availability  (supply) of  food  is a necessary but not  sufficient  condition for food security. Production shortfalls related to drought directly threaten food supply, but increases in public interventions (including distribu- tion of food aid) and private trade can prevent a major decline in food supply. Even when there is adequate availability at the national or local level, poor households often do not have adequate access to food because they lack suffi- cient entitlements to food—that is, legal means to acquire sufficient food—such  as their own production of food, other earned incomes, gifts, and government transfers.3 In addition to availability and access, food security of individuals depends on their utilization of food (the amounts and types of food they eat) and various factors affecting the body’s capacity to absorb nutrients, which in turn affect overall health. Thus, adequate nutritional outcomes for individuals depend not only on consumption of macronutrients (calories, proteins, and fats and oils) or micronutrients (for example, iron, vitamin A, and iodine) but on whether diarrheal disease or other health problems inhibit effective absorption of nutrients.4 The large-scale famines in Ethiopia in 1972–74 and 1984–85 involved collapses of all three components of food security. In both cases, droughts played a key role in reducing production and availability, and food aid and other imports were insufficient to offset the losses. But other broad factors were also  major causes of these famines, including key government policies. In the case of the 1984–85 famine, military conflict, government policies on land reform  and investment, and market failures were instrumental (Webb and von Braun 1994). Ongoing wars against the federal government by regional armies resulted in loss of life and serious injuries.5 These conflicts also reduced the labor avail- able for crop cultivation and diverted scarce public resources away from needed investments in agriculture, roads, telecommunications, and emergency response programs. Government land reform policies abolished private landownership in 1975 and placed ceilings on rights of access to private land, imposing a limit 2. Chapter 9 of this book provides an overview and chronology of food security and famine in Ethiopia since 1888, as well as the medium- and long-term policy forces behind food shortages and emergencies. 3. The concept of entitlements is due to Amartya Sen and his work on the Great Bengal Famine, which argued that the famine was due largely to a loss of purchasing power (entitlements) of many poor households rather than to a shortfall in supply of food. See Sen (1981). 4. See Devereux and Maxwell (2001) for a more detailed summary of the various aspects of food security. 5. The armies involved were those of the Eritrean People’s Liberation Front and the Tigray Peoples’ Liberation Front. Introduction 5 of 10 hectares per farm. Public investment in agriculture was allocated mainly to state farms and producer cooperatives instead of independent small farmers. Finally, market restrictions (regulation of and bans on interregional movement of grain and labor) forced sales of private traders’ grain to the Agricultural Marketing Corporation at low fixed prices, and poor market infrastructure also  contributed to a lack of market integration. Since the 1984–85 famine, and particularly following the fall of the Derg regime in May 1991, government policy in Ethiopia has included substantial liberalization of agricultural markets, investment in agricultural research and extension, building of key transport infrastructure, and establishment of the Productive Safety Net Programme. These policy developments have contrib- uted to major increases in national food production and enhanced food security. Production of the major cereals increased substantially in the 1990s and the first  decade of the 2000s. Although increases in area cultivated accounted for most of the production increase from 1991–92 through 2003–04, yield increases have accounted for about half of the more recent surge in cereal production from 10 million metric tons in 2003–04 to more than 14 million metric tons in 2008–09.6 Growth in production during the 1990s was nearly 2 percent lower than produc- tion gains during the 2000s. Growth in the area cultivated was near 6 percent in the 1990s, with insignificant yield increases, whereas in the 2000s improved  intensification and yield growth of 3.5 percent were recorded, with an average  area increase of 3 percent. Nationally, the proportion of people below the pov- erty line decreased from 45 percent in the mid-1990s to 39 percent in 2005. Rural poverty rates dropped from 48 to 39 percent, although the urban poverty rate rose slightly, from 33 to 35 percent, in the same period (Ethiopia, MoFED 2008). Nonetheless, according to estimates from 2005 national household sur- vey data, the level of malnutrition in Ethiopia remains very high: an estimated 44 percent of the population in the country consumes fewer calories than the minimum dietary requirement (Schmidt and Dorosh 2009; von Grebmer et al. 2010). The Historical and Geographical Context Ethiopia’s current food security and prospects for the future are very much shaped by the geography and history of the country. Located in East Africa, with borders connecting it to Sudan, Eritrea, Djibouti, Somalia, and Kenya, Ethiopia has been a landlocked country since the independence of Eritrea in May 1993. Geographically, Ethiopia is commonly described as “Three Ethio- pias”: the dry, semiarid lowlands that dominate the eastern third of the country; the rainfall-sufficient (mainly highland) areas in the western third of the coun- 6. All tons in this book are metric tons. 6 Paul Dorosh and Shahidur Rashid try; and the drought-prone highlands in northern and central Ethiopia. Most of the population resides in the latter two broad areas.7 Throughout the country, the population is more concentrated along major road networks, but much of Ethiopia’s population remains remote: 45 percent of the population lives more than five hours from a city of 50,000 (Schmidt and Kedir 2009).  These variations in geography and agroclimatic zones have important implications for the food economy of Ethiopia. The rugged terrain in much of the highlands makes transport and communication difficult. Rainfall varies  significantly between mountains and valleys, even across short distances. With  these large variations in altitude, rainfall, and connectivity, cropping patterns also vary sharply, and no single crop dominates Ethiopia’s food consumption, as does rice in much of Asia, maize in Latin America, or wheat in many cooler climates. Instead, five cereals are cultivated on a wide scale: teff (an indige- nous crop widely grown only in Ethiopia and Eritrea), wheat, maize, sorghum, and barley. Coffee, cultivated in the rainfall-sufficient southern highlands, is  Ethiopia’s major export crop. Livestock—mainly cattle, sheep, and goats—are the major sources of meat and livelihood of the pastoralist and agropastoralist populations. Ethiopia’s recorded history dates to a biblical account of a visit by the Queen of Sheba to King Solomon of Israel (10th century BC). According to the Kebre Negast (The glory of the kings), a history written in Ge’ez in the 14th century AD, King Solomon and Queen Makeda’s son became Emperor Mene- lik I, Ethiopia’s first emperor.8 From the 1st to the 6th century AD, the kingdom of Axum in northern Ethiopia (in today’s Amhara region) was a major eco- nomic and military power in the region, with significant interactions with Egypt  and the Middle East. Until the 19th century, however, what is today Ethiopia was divided into small subnational groups, which in the highlands were each ruled by their own ras (literally, “head”). A series of emperors in the 19th century—Tewodros II (1855–68), Yohannes IV (1872–89), and finally Mene- lik II (1889–1913)—gradually united more and more groups, extending their control to most of the Ethiopian highlands by the end of the 20th century. Ethiopia avoided colonization in the 1900s through a combination of the military advantages of defensive positions in the highlands, defeat of an Italian army at the Battle of Adwa in March 1896, and deft diplomacy under Menelik II that played British, French, and Italian interests against one another. Ethiopia retained its independence under Haile Selassie (crowned emperor in 1929), 7. Administratively, Ethiopia is divided into 11 regions largely based on ethnic groups, in- cluding three urban administrative areas (Addis Ababa, Dire Dawa, and Harari); the major regions are further subdivided into zones, woredas, and kebeles. To avoid confusion, this book uses the terms regions or zones only when referring to actual administrative areas. 8. Marcus (2002). See also Zewde (2002). Introduction 7 though it was occupied by the Italian army during World War II (from 1936 to 1941). Under the long reign of Haile Selassie Ethiopia began to modernize, but rural Ethiopia, in general, remained isolated and very poor. Most land was formally owned by the state, churches, and the rural elite, with significant rents  due from private small farmers for use of the land. There were few paved roads in the country. The electricity-generating capacity was only 4.7 watts per per- son nationally in 1972, and almost no electricity was available outside of a few major cities. A series of droughts contributed to famines in much of Ethiopia from 1972 to 1974, and as the economic and social situation worsened, a group of officers  later known as the Derg (“committee” in Amharic) gradually gained more power. Haile Selassie was increasingly seen as being out of touch with the suf- fering of the Ethiopian people, particularly after a television report by the Brit- ish Broadcasting Corporation, originally aired in Britain in October 1973, was re-edited and shown in Ethiopia, showing scenes of the emperor feeding his dogs along with horrific scenes of famine in Wollo (northern Ethiopia).9 The aging emperor was finally deposed on September 12, 1974, and ultimately suc- ceeded by Mengistu Haile Mariam, who took power formally in 1977. Under the Derg, Ethiopia adopted socialist economic policies, national- izing land and private businesses and placing greater controls on markets and prices. Agricultural investments focused on large state farms. When a severe drought hit much of Ethiopia in 1984, a famine ensued. Although the inter- national community responded with millions of dollars of food aid and other relief through various charity events and programs,10 the famine led to more than half a million deaths. The Derg regime nonetheless survived until 1991, when it was overthrown by forces of the Ethiopian People’s Revolutionary Democratic Front (EPRDF), which marched on Addis Ababa from the north. The new government, led by Meles Zenawi, subsequently adopted a more liberal economic policy, allowing private-sector markets to operate more freely. Eritrea, up until that point the northernmost part of Ethiopia, seceded in May 1993 following a referendum on independence. Initially, relationships between the two countries were peaceful, but a war confined largely to Eritrean  territory broke out and lasted from May 1998 to June 2000. Since that time, economic and diplomatic ties have been disrupted, depriving Ethiopia of its only ports (Mitsiwa and Aseb on the Red Sea coast). 9. Gill (2010, 34–35). 10. These charities and events included Band Aid, formed by Bob Geldof in 1984; USAID (United States Agency for International Development) for Africa, for which Michael Jackson and Lionel Ritchie wrote “We Are the World”; and the Live Aid concerts held in 1985 in London and Philadelphia (Gill 2010, 12–13). 8 Paul Dorosh and Shahidur Rashid After the downfall of the Derg, Meles and the EPRDF won national elec- tions in 1995 and 2000 by wide margins. The 2005 elections were very close, however, and violence erupted following announcement of the EPRDF victory. The EPRDF won handily in 2010, but Prime Minster Meles died soon thereaf- ter, in July 2012. Economic Reforms and Outcomes in the 1990s and 2000s The immediate post-Derg period of the mid-1990s was marked by administra- tive decentralization in 1992 and substantial liberalization of the economy as restrictions on domestic food markets were removed, private enterprise was encouraged, export taxes were eliminated, and the currency was devalued by more than 100 percent in 1993 (from 2.5 to 5.5 birr/US$). Moreover, a new development strategy, Agricultural Development–Led Industrialization (ADLI), announced in 1992, dramatically shifted the focus of government investments away from industry and large farms to support broad-based growth of small- holder agriculture. This emphasis on smallholder agriculture based on the ADLI approach continued with the Sustainable Poverty Reduction Strategy of 2002 (Ethiopia, MoFED 2002) and the Plan for Accelerated and Sustained Develop- ment to End Poverty in 2006 (Ethiopia, MoFED 2006). In further support for private markets, the Ethiopian Commodity Exchange was launched in 2008, designed to provide a transparent and efficient market for agricultural products. Economic liberalization in Ethiopia did not mean a complete withdrawal of the government from markets. Although initially the government liberalized the  fertilizer  sector,  the  market  shares  of  individually  owned  private  firms  dropped from 30 percent in 1995 to zero in 1999. According to Jayne et al. (2003),  these firms were  replaced  by  “private”  holding  companies  in  1999,  which continued to dominate markets until 2007. Since 2008, the import of fertilizer has been controlled by the Agricultural Input Supply Enterprise, and the distribution is carried out by the cooperative unions.11 In addition, land markets were not liberalized but rather remained under state control, with farm- ers receiving usufruct rights to agricultural land. The Ethiopian Grain Trading Enterprise, a public marketing enterprise, was established in 1992 to stabilize grain prices, though the volumes of its commercial imports, domestic pur- chases, and sales have been small relative to the overall cereal markets. More important have been public imports of food aid, generally distributed as direct transfers to households until the advent of the Productive Safety Net Programme in 2005, which introduced a work requirement for able-bodied individuals as a condition of food or cash transfers. In addition, a dramatic rise in domestic inflation  in  2007  and 2008  led  to  occasional  crackdowns on private  traders  deemed to be hoarding commodities or charging excessive prices. 11. Chapter 4 discusses these issues in greater detail and examines the systems and markets for seed, fertilizer, and extension in Ethiopia. Introduction 9 Overall, Ethiopia’s economic performance in the past two decades has far surpassed its performance in the 1970s and 1980s. Real gross domestic product (GDP) grew only 1.2 percent between 1981 and 1991 but has since accelerated to 4.3 percent from 1991 to 2001 and to 8.2 percent from 2001 to 2009 ( 1.2). Over these three decades, the total population has more than doubled, from 36.3 million in 1981 to 82.8 million in 2009, though population growth slowed from 3.3 percent per year in the 1980s to 3.0 percent in the 1990s and 2.6 percent in the 2000s. Thus, per capita GDP, which fell by 2.0 percent per year in the 1980s, rose by 1.3 percent per year in the 1990s and 5.4 percent from 2001 to 2009. In constant (2000) US dollars, per capita GDP in 2009 is esti- mated at $201, which is almost 75 percent higher than the per capita GDP of $116 in 1991.12 Agricultural growth has made a major contribution to Ethiopia’s impres- sive overall growth performance, accelerating from 1.3 percent per year in the 1980s to 2.9 percent in the 1990s to 6.2 percent in the 2000s (see Table 1.2). Nonetheless, the industrial sector (which includes processing of agricultural goods) and services grew even more rapidly than did agriculture. In fact, the share of agriculture in national GDP fell steeply between 1991 and 2001, from 64.1 percent to 47.7 percent (and to 47.3 percent in 2009). Substantial investment, which increased from 11.1 percent of GDP in 1991 to over 20 percent in 2001, has been a major source of growth. Most of this  investment,  however,  has  been financed  by  foreign  capital  inflows  and  transfers (workers’ remittances and private transfers). Gross domestic savings was only 9.7 percent of GDP in 2001 and just 2.3 percent of GDP in 2009. Up until 2007, Ethiopia enjoyed considerable macroeconomic stability in the post-Derg period. Inflation, which averaged 7.0 percent per year in the 1980s,  fell to 3.0 percent per year in the 1990s. Inflation surged to an annual average  of 44 percent in 2008, though, as Ethiopia’s money supply and international energy prices increased sharply. Tight monetary policy in late 2008 contributed to a sharp decline in inflation in 2009 and 2010, and a gradual depreciation of  the birr relative to the US dollar and other currencies restored the real exchange rate to its levels of the early 2000s by mid-2010 (Ahmed and Dorosh 2009). The subsequent 19 percent devaluation of the birr from 14.1 to 16.8 birr/US$ between August and September 2010 further improved the incentives for pro- duction of tradable goods in Ethiopia’s economy and may have signaled a greater emphasis on incentives for long-term economic growth. Other economic and social indicators generally paint a picture of sub- stantial progress in Ethiopia over the past two decades as well (Table 1.3). The country’s electricity-generating capacity and the number of kilometers of paved roads increased by 123 and 52 percent, respectively, between 1991 and 2009. Most of the population still resides in rural areas, though the rate of urbanization (as measured using the official administrative definition) increased from only 12. All dollar amounts in this book are US dollars. T A B L E 1 .2 E th io pi a: S el ec te d ec on om ic v ar ia bl es , 1 98 1– 20 09 A nn ua liz ed g ro w th r at e (p er ce nt ) V ar ia bl e 19 81 19 91 20 01 20 09 19 81 –9 1 19 91 –2 00 1 20 01 –0 9 Po pu la tio n (m ill io ns ) 36 .3 50 .0 67 .3 82 .8 3. 3 3. 0 2. 6 G D P (c on st an t 1 98 0 U S$ , m ill io ns ) 5, 14 7 5, 78 9 8, 85 9 16 ,6 23 1. 2 4. 3 8. 2 G D P pe r ca pi ta ( co ns ta nt 2 00 0 U S$ ) 14 1. 9 11 5. 9 13 1. 7 20 0. 7 –2 .0 1. 3 5. 4 G D P (c on st an t 1 98 0 bi rr , b ill io ns ) 41 .9 47 .2 72 .2 13 5. 5 1. 2 4. 3 8. 2 A gr ic ul tu ra l G D P (c on st an t 1 98 0 bi rr , b ill io ns ) 22 .4 2 5. 6 3 4. 1 5 5. 0 1. 3 2. 9 6. 2 Sh ar e of a gr ic ul tu re ( pe rc en t o f G D P) 60 .7 6 4. 1 4 7. 7 4 7. 3 0. 5 –2 .9 –0 .1 Sh ar e of in du st ry ( pe rc en t o f G D P) 1 0. 6 8. 7 13 .0 1 4. 1 –2 .0 4. 1 1. 0 Sh ar e of m an uf ac tu re s (p er ce nt o f G D P) 4. 9 3. 0 5. 7 5. 2 –4 .7 6. 5 –1 .3 Sh ar e of s er vi ce s (p er ce nt o f G D P) 2 8. 8 27 .3 39 .3 3 8. 6 –0 .5 3. 7 –0 .2 G ro ss d om es tic s av in gs ( pe rc en t o f G D P) 10 .2 6. 0 9 .7 2 .3 –5 .2 5. 0 –1 6. 3 G ro ss c ap ita l f or m at io n (p er ce nt o f G D P) 14 .5 11 .1 2 1. 5 2 0. 5 –2 .7 6. 8 –0 .6 E xp or ts o f go od s an d se rv ic es ( pe rc en t o f G D P) 7 .6 4 .1 1 2. 0 9 .9 –6 .1 11 .4 –2 .3 Im po rt s of g oo ds a nd s er vi ce s (p er ce nt o f G D P) 1 1. 9 9 .2 2 3. 7 2 8. 1 –2 .6 10 .0 2. 1 O ffi ci al  e xc ha ng e  ra te  (b irr /U S$ )   2 .1     2 .1     8 .5     1 0. 7    0. 0  15 .1   3. 0 C on su m er p ri ce in de x (2 00 5 = 1 00 ) 27 .3 53 .8 72 .5 20 6. 2 7. 0 3. 0 14 .0 so u rc e: W or ld B an k (2 01 0) . n o te : G D P = g ro ss d om es tic p ro du ct . T A B L E 1 .3 E th io pi a: S el ec te d ec on om ic a nd s oc ia l i nd ic at or s, 1 98 1– 20 09 A nn ua liz ed g ro w th r at e (p er ce nt ) In di ca to r 19 81 19 91 20 01 20 09 19 81 –9 1 19 91 –2 00 1 20 01 –0 9 Pa ve d ro ad s (t ho us an ds o f ki lo m et er s) n .a . 2 8. 0 3 1. 4 4 2. 4 n .a . 1. 2 3. 8 E le ct ri ci ty -g en er at in g ca pa ci ty ( th ou sa nd s of k ilo w at ts ) 1 75 .5 3 31 .5 4 17 .0 7 37 .7 6. 6 2. 3 7. 4 U rb an p op ul at io n (m ill io ns ) 3 .9 6 .4 1 0. 2 1 3. 7 5. 3 4. 7 3. 8 U rb an iz at io n ra te ( pe rc en t o f po pu la tio n) 1 0. 6 1 2. 9 1 5. 1 1 7. 0 1. 9 1. 6 1. 5 L if e ex pe ct an cy a t b ir th ( ye ar s) Fe m al es 4 5. 2 4 8. 9 5 3. 4 5 6. 7 0. 8 0. 9 0. 7 M al es 4 2. 3 4 5. 9 5 0. 3 5 3. 8 0. 8 0. 9 0. 8 T ot al 4 3. 7 4 7. 4 5 1. 8 5 5. 2 0. 8 0. 9 0. 8 Po ve rt y he ad co un t ( pe rc en t) a t U S$ 1. 25 a d ay ( PP P) a 6 6. 2 6 0. 5 5 5. 6 3 9. 0 –0 .4 –1 .0 –3 .3 Po ve rt y he ad co un t ( pe rc en t) a t U S$ 2 a da y (P PP ) 8 9. 9 8 4. 6 8 6. 4 7 7. 6 –0 .4 0. 4 –1 .8 so u rc e: W or ld B an k (2 01 0) . n o te s:  Li fe  e xp ec ta nc y  at  b irt h  fo r 2 00 9  is  b as ed  o n  20 08  d at a;  p av ed  ro ad s fi gu re  fo r 2 00 9  is  b as ed  o n  20 07  d at a;  n .a . =  n ot  a va ila bl e;  P PP  =  p ur ch as in g  po w er   pa ri ty . a P ov er ty h ea dc ou nt d at a ar e fo r 19 82 , 1 99 5, 2 00 0, a nd 2 00 5. F ig ur es f or c ha ng es in p ov er ty a re p er ce nt ag e po in ts p er y ea r. 12 Paul Dorosh and Shahidur Rashid 10.6 percent in 1981 to 17.0 percent in 2009. Life expectancy at birth rose from 42.3 to 53.8 years for males and from 45.2 to 56.7 years for females over the same period. Poverty estimates are not available for all years, but the available data also show a steady improvement in that area. World Bank (2010) estimates show the poverty headcount falling from 66.2 percent in 1982 to 60.5 in 1995 to 55.6 in 2000, followed by an even steeper drop to 39.0 percent in 2005.13 Unfortunately, no nationally representative survey has been conducted since 2005, but the high reported levels of agricultural and overall GDP growth sug- gest that poverty rates may have continued to decline. Regional Comparisons Ethiopia is vastly different from its neighbors, particularly in terms of its large land area and population and its widely varying ecologies. Nonetheless, com- parisons with neighboring countries provide a useful perspective on the huge development challenge facing the country. Ethiopia has the largest population in the region (82.8 million people in 2009), approximately double that of Sudan and Kenya (42.3 and 39.8 million, respectively) and nine times larger than that of Somalia (9.1 million) (Table 1.4). In spite of the rapid economic growth of the past two decades, however, Ethiopia remains one of the poorest countries in East Africa. Ethiopia’s growth in GDP per capita outpaced that of neighboring countries in the 2000s (5.7 percent per year compared to 4.3 percent in Uganda and 2.6 percent in Sub- Saharan Africa as a whole). Yet it remains one of the poorest countries in the region, with a GDP per capita of only $201 (2000) per person in 2009, 45 per- cent less than that of Uganda ($366 [2000] per person), less than half that of Kenya or Sudan, and less than one-third the average for Sub-Saharan Africa ($620 [2000] per person). Eritrea, Ethiopia’s small neighbor to the north (popu- lation 5.1 million), is much poorer than Ethiopia, though, with a GDP per capita of only $130 (2000) per person.14 Agriculture accounts for a much higher share of GDP in Ethiopia (47.3 percent) than in neighboring Kenya (22.6 percent), Sudan (29.7 percent) or Uganda (24.7 percent) (see Table 1.4). This, in large part, reflects the low rate  of industrialization and urbanization of Ethiopia relative to its neighbors. Few data are available on poverty for any of the countries in the region. Using a poverty line of $1.25 a day purchasing power parity, the World Bank estimates Ethiopia’s poverty rate at 39 percent for 2004/05. Kenya’s poverty rate was al- most exactly half of Ethiopia’s (19.7 percent); Uganda’s poverty rate was higher than Ethiopia’s (51.5 percent). Note, though, that poverty figures vary  13. Measured at the US$1.25-per-day poverty line. 14. For most countries of eastern Africa, and especially for Eritrea and Somalia, economic data are very scarce, adding considerable uncertainty to cross-country comparisons. T A B L E 1 .4 E as t A fr ic an c ou nt ri es : S el ec te d in di ca to rs , 2 00 9 Su b- Sa ha ra n In di ca to r E th io pi a E ri tr ea K en ya So m al ia Su da n U ga nd a A fr ic a Po pu la tio n (m ill io ns ) 82 .8 5. 1 39 .8 9. 1 42 .3 32 .7 84 0. 3 Po pu la tio n gr ow th r at e (p er ce nt , 2 00 0– 09 ) 2. 6 3. 8 2. 7 2. 4 2. 1 3. 3 2. 5 G D P pe r ca pi ta ( co ns ta nt 2 00 0 U S$ ) 20 1 13 0 45 2 n. d. 53 6 36 6 62 0 G D P pe r ca pi ta g ro w th r at e (p er ce nt , 2 00 0– 09 ) 5. 7 –3 .3 1. 7 n. d. 5. 0 4. 3 2. 6 A gr ic ul tu re , v al ue -a dd ed ( pe rc en t o f G D P) 47 .3 14 .4 22 .6 n. d. 29 .7 24 .7 12 .3 Po ve rt y he ad co un t r at io a t U S$ 1. 25 a d ay ( PP P) 39 n. d. 19 .7 n. d. n. d. 51 .5 n. d. H un ge r in de x 29 .8 35 .7 19 .8 n. d. 20 .9 15 21 .7 U nd er no ur is he d po pu la tio n 44 66 30 n. d. 20 15 27 .5   U nd er w ei gh t c hi ld re n  un de r fi ve  y ea rs  o f a ge   34 .6   35 .3   16 .5   32 .8   31 .7   16 .4   23 .6   U nd er  fi ve  m or ta lit y  ra te   10 .9   5. 8  12 .8   20   10 .9   13 .5   14 .0 so u rc es : v on G re bm er e t a l. (2 01 0) ; W or ld B an k (2 01 0) . n o te s: D at a on th e po ve rt y he ad co un t r at io f or E th io pi a an d K en ya a re f or 2 00 5; d at a on G D P (g ro ss d om es tic p ro du ct ) pe r ca pi ta f or E ri tr ea a re f or 2 00 8, a nd th e gr ow th r at e is f or 2 00 0– 08 ; n .d . = n o da ta f or th e co un tr y. 14 Paul Dorosh and Shahidur Rashid widely depending on the poverty line used (the Ethiopian government does not use the $1.25-per-day poverty line in its official poverty estimates). The estimated percentage of the population consuming inadequate calo- ries in Ethiopia (44 percent) is significantly higher than that of Kenya (30 per- cent) and almost three times higher than that of Uganda (15 percent) but only two-thirds the rate in Eritrea (66 percent). Similarly, the percentage of children under five years of age who are underweight is more than double that of Kenya  or Uganda (34.6 percent, as compared to 16.5 and 16.4 percent in the latter two countries). The under-five mortality rate, which is affected heavily by the inci- dence of malaria and other diseases as well as food intake, is lower in Ethiopia (10.9 percent) than in Kenya or Uganda, however, and below the rate for Sub- Saharan Africa (14.0). The Plan of the Book Part I of the book presents an overview and analysis of Ethiopia’s food econ- omy. Key to an understanding of the complex food production systems of a country as large as Ethiopia are the various development domains, as deter- mined by agroecology, settlement patterns, and road networks (Chapter 2). Given the wide variations in elevation, rainfall, and market access in the coun- try, patterns of crop production often vary sharply across short distances. Yet overall cereal production increased rapidly in the 2000s in much of the country due to both area and yield expansion (Chapter 3). Though increasing, the use of modern inputs such as fertilizer and improved seeds remains considerably low. In 2007/08, only about 40 percent of the total land allocated to cereal cultivation benefited from chemical fertilizers, and most of these chemical fer- tilizers went to maize and wheat. Besides, the amount of land in cereal cultiva- tion declined from 2001/02 to 2007/08. The use of improved seeds in 2007/08 was negligible; they were applied over about 5 percent of the total cereal acre- age.15 In addition, irrigation has seldom been employed, with only 1 percent of cereal acreage irrigated in 2007/08. Expanded use of fertilizer is one major factor behind increases in land productivity, though the level of use of improved seeds remains low and indeed has suffered setbacks in recent years (Chapter 4). Increased cereal production has greatly increased market volumes, and there is substantial evidence of greater market integration across major whole- sale markets and reduced marketing costs and margins, in part due to invest- ments in roads and improvements in telecommunications (Chapter 5). In addi- tion, livestock (mainly cattle, sheep, and goats) and dairy also play major roles in Ethiopia’s food economy and fulfill an important function in coping with  15.  This low figure for improved seed use is based on official CSA farmer survey estimates that  likely do not include seed of improved, open-pollinated varieties saved from their own previous year’s harvest. Overall, some estimates suggest that the use of improved seed for wheat and maize cultivation may be as much as 50 to 70 percent (see Chapter 4, “Improved Seed Adoption,” pp. 92–94). Introduction 15 shocks, accumulating wealth, and serving as a store of value in the absence of credit and savings markets. Yet, productivity in the livestock subsector in Ethi- opia lags behind those in both its neighbors in East Africa as well as the least developed countries, and import bans by the Arab States of the Persian Gulf in some years resulted in major suffering for livestock keepers in both pastoral regions and central highlands. Thus, understanding the causes of low productiv- ity and the problems with marketing is essential for appropriate policy formula- tion (Chapter 6). Overall, gains in agricultural production and improvements in markets have contributed to increases in household consumption and reductions in pov- erty (Chapter 7). The wide diversity in consumption patterns across regions of Ethiopia and the tendency for nearly all household groups to consume more than just one cereal (as well as enset, especially in the Southern Nations, Nationalities, and People’s Region) reduce reliance on a single staple and thereby reduce the risks associated with failure of a single crop. For pastoralists and agropastoralists in dry lowland parts of Ethiopia, livestock products (includ- ing dairy) are major sources of calories and incomes, but both poor and nonpoor households purchase a large share of their food. Part II of the book discusses major agricultural and food policy interven- tions and includes a concluding chapter on key policy issues facing Ethiopia today. Like most other Sub-Saharan African countries, Ethiopia has committed itself to investing in sustainable agricultural growth as part of the Comprehen- sive Africa Agriculture Development Programme. Economywide analysis of continued gains in agricultural productivity (Chapter 8) shows not only that effective agricultural investments lead to gains in production levels that benefit  net buyers but also that there is sufficient demand for food products to keep  agricultural prices from falling so rapidly as to impoverish surplus-producing farmers. Assuming that agricultural growth reaches 6 percent per year, model estimates suggest that overall GDP growth will increase by 1 percentage point per year. This higher growth rate would lift an additional 3.7 million people above the poverty line by 2015. The model results also show, however, that rapid non-agricultural growth can produce similar reductions in overall pov- erty. Nonetheless, in the absence of effective disaster response policy and safety nets, increases in production and moderate reductions in food prices are not sufficient  to substantially  increase  food security  for many asset-poor house- holds. Private-sector imports have been discouraged through restrictions on foreign exchange and uncertainties regarding government policy actions and food aid; as a result, public cereal stocks and food aid continue to be the major mechanisms for disaster response and cereal price risk management in Ethiopia (Chapter 9). Yet the success in addressing disasters cannot be attributed merely to cereal stocks and food aid. A key aspect of the country’s success in address- ing emergencies has been institutional development that has included well- managed grain reserves, early warning systems, and coordination between grain stocks and safety net programs, as well as innovative safety nets and overall 16 Paul Dorosh and Shahidur Rashid improvements in emergency management. In an effort to improve food security among the most vulnerable, the Productive Safety Net Programme (PSNP) aims to target the poor (in terms of assets and income) for a public works program and food-insecure households for a direct support program. Findings suggest that the PSNP has been able to target resources to the poorest households in rural areas using a combination of geographic and community-based targeting and that, compared to other safety net programs, the PSNP is better at support- ing its desired population (Chapter 10). Finally, in spite of the enormous progress of the past two decades, Ethiopia continues to face key challenges to raising incomes, reducing poverty, and achieving food security for all. Chapter 11 summarizes the key findings of the  book and places them in the context of major policy choices regarding raising agricultural productivity, reducing poverty and chronic food insecurity, and instilling mechanisms to prevent severe increases in transitory food insecurity caused by droughts and other production and market shocks. References Africa Watch. 1991. Evil Days: 30 Years of War and Famine in Ethiopia. Washington, DC: Human Rights Watch. Ahmed, H., and P. Dorosh. 2009. Foreign Exchange Rationing, Wheat Markets, and Food Security in Ethiopia. Ethiopia Strategy Support Program 2 (ESSPII) Discus- sion Paper 004. Washington, DC: International Food Policy Research Institute. CRED (Centre for Research on Epidemiology and Disaster). 2011. Emergency Event Database (EM-DAT): International Disaster Database. Brussels, Belgium: CRED, Université Catholique de Louvain. Accessed December 2011. www.emdat.be Dercon, S., and R. V. Hill. 2009. “Growth from Agriculture in Ethiopia: Identifying Key Constraints.” University of Oxford, Oxford, UK. Accessed December 2011. www .economics.ox.ac.uk/members/Stefan.Dercon/Ethiopia%20paper%203_v5.pdf. Dercon, S., R. V. Hill, and A. Zeitlin. 2009. “In Search of a Strategy: Rethinking Agriculture-Led Growth in Ethiopia.” Synthesis paper prepared as part of a study on agriculture and growth in Ethiopia. University of Oxford, Oxford, UK. Accessed December 2011. www.economics.ox.ac.uk/members/Stefan.Dercon/In%20Search %20of%20a%20Strategy_v3.pdf. Devereux, S., and S. Maxwell, eds. 2001. Food Security in Sub-Saharan Africa. London: ITDG Publishing. De Waal, A. 1997. Famine Crimes: Politics and the Disaster Relief Industry in Africa. Bloomington, IN, US: Indiana University Press. Ethiopia, CSA (Central Statistical Agency). 2005–10. Area and Production of Major Crops, Addis Ababa. Accessed April 2012. All issues are available at www.csa .gov.et/index.php?option=com_content&view=article&id=91&Itemid=224. Ethiopia, MoFED (Ministry of Finance and Economic Development). 2002. Ethiopia: Sustainable Development and Poverty Reduction Program. Addis Ababa. ———. 2006. Ethiopia: Building on Progress—A Plan for Accelerated and Sustained Development to End Poverty (PASDEP) (2005/06–2009/10), vol. 1. Addis Ababa. ———. 2008. Dynamics of Growth and Poverty in Ethiopia. Addis Ababa. Introduction 17 Gill, P. 2010. Famine and Foreigners: Ethiopia since Live Aid. Oxford, UK: Oxford University Press. Hammond, L., and D. Maxwell. 2002. “The Ethiopian Crisis of 1999–2000: Lessons Learned, Questions Unanswered.” Disasters 26 (3): 262–279. Jayne, T. S., J. Govereh, M. Wanzala, and M. Demeke. 2003. “Fertilizer Market Devel- opment: A Comparative Analysis of Ethiopia, Kenya, and Zambia.” Food Policy 28 (4): 293–316. Marcus, H. G. 2002. A History of Ethiopia. Berkeley, CA, US: University of California Press. McCann, J. 1995. People of the Plow: An Agricultural History of Ethiopia, 1800–1990. Madison, WI, US: University of Wisconsin Press. Schmidt, E., and P. Dorosh. 2009. A Sub-National Hunger Index for Ethiopia: Assessing Progress in Region-Level Outcomes. IFPRI–ESSP2 Discussion Paper 5. Washing- ton, DC: International Food Policy Research Institute. Schmidt, E., and M. Kedir. 2009. Urbanization and Spatial Connectivity in Ethiopia: Urban Growth Analysis Using GIS. IFPRI–ESSP2 Discussion Paper 3. Washing- ton, DC: International Food Policy Research Institute. Sen, A. 1981. Poverty and Famines: An Essay on Entitlement and Deprivation. Oxford, UK: Oxford University Press. von Grebmer, K., M. T. Ruel, P. Menon, B. Nestorova, T. Olofinbiyi, H. Fritschel, and  Y. Yohannes. 2010. Global Hunger Index 2010. Washington, DC: International Food Policy Research Institute. Webb, P., and J. von Braun. 1994. Famine and Food Security in Ethiopia: Lessons for Africa. Washington, DC: International Food Policy Research Institute. World Bank. 2006. “Ethiopia: Policies for Pro-Poor Agricultural Growth.” World Bank, Washington, DC. Mimeo. ———. 2010. World Development Indicators 2009 Database. Development Data Group. Washington, DC: World Bank. Zewde, B. 2002. A History of Modern Ethiopia: 1855–1991, 2nd ed. Addis Ababa, Ethiopia: Addis Ababa University Press. PART I Overview and Analysis of Ethiopia’s Food Economy 21 The opportunities and constraints facing Ethiopian agriculture are strongly influenced by geographical location. Ethiopia’s diverse landscape defines cer- tain agricultural production potentials, access to input and output markets, and local population densities, which determine both labor availability and local demand for food. Understanding the geographical expression of Ethiopia’s agri- cultural and rural development options provides greater information for more locally targeted policy options. These conditions not only vary over space but change over time as well. New and improved roads, greater telecommunications, improved access to electricity, and ongoing urban growth continue to lower transaction costs and improve market access. Evolving production opportunities and technologies continue  to provide greater flexibility of  livelihood decisions within defined  biophysical endowments. As Ethiopia continues to invest in infrastructure and technology, its agricultural landscape continues to be reshaped and redefined  into broader areas of opportunity and growth. The objectives of this chapter are twofold. First we provide an overview of the geographical features that constitute a basic reference for understanding production systems and the geography of agricultural production in rural Ethio- pia. This characterization is organized around the economic logic of compara- tive advantage for a variety of generalized production decisions of relevance in Ethiopia. Second we extend this framework to organize evidence for and dis- cussion of important areas of dynamism in Ethiopia’s rural economic land- scape, including a discussion of land tenure policy and its effects on the level of investments in the agricultural sector. Thus, this chapter is mainly a descrip- tive assessment of important production contexts and the manner in which these contexts are evolving. Characterizing Production Contexts In large and heterogeneous countries such as Ethiopia, agricultural potential is unevenly distributed over space, and the distribution of production patterns 2 Ethiopian Agriculture: A Dynamic Geographic Perspective JORDAN CHAMBERLIN AND EMILY SCHMIDT 22 Jordan Chamberlin and Emily Schmidt reflects this landscape. Mapped zones of smallholder production systems have  long been recognized as important in Ethiopia precisely because of such land- scape heterogeneity (for example, Westphal 1975; De Pauw and Bruggeman 1988; Hurni 1998). To provide a useful characterization of agricultural land- scapes, geographical perspectives seek a balance between too little information and too much. On the one hand, spatial characterization is motivated by the fact that average national statistics and one-size-fits-all development strategies are  insufficient for effective rural planning. On the other hand, a surfeit of detail  may render mapped classifications useless for planning on the basis of shared  conditions: hundreds of subcategories, whether on a map or in a table, even when they accurately portray local variation, usually contain too much information to be tractable as a planning aid. Thus, characterizations generally seek to reduce complexity by prioritizing characteristics that are most meaningful to the pro- duction systems of interest. Agroecological zones are perhaps the predominant methodology used to understand actual and potential agricultural production across geographic space. Agroecological zonation uses biophysical attributes of soil, terrain, and climate to organize land-use types or production systems into relatively homogenous units (FAO 1978; Hurni 1998). Hurni (1998) implemented a set of agroecologi- cal zone definitions for Ethiopia based on traditional zone designations widely  used by rural residents. He linked these designations with specific elevation and  rainfall parameters, which allowed mappable boundaries to be imposed on agroecological zones (Figure 2.1). In Ethiopia most agricultural production takes place in the weyna dega and dega zones (highland areas from 1,500 to 2,300 and from 2,300 to 3,200 meters above sea level), where land productivity has traditionally coincided with the densest rural populations. Figure 2.1 shows these zones in midtone gray, scat- tered throughout the highlands. Specific crops and livelihood choices within  and outside this band are conditioned by moisture and temperature regimes, among other factors. The crops most suited to grow in the weyna dega and dega zones are also the most commonly produced crops in Ethiopia. Most producers in these zones are smallholders occupying less than a hectare of land per house- hold on average.1 Smallholder production is dominated by five major cereal  crops accounting for almost three quarters of the total cultivated area and about 68 percent of total production. Each of the major cereals—teff, maize, wheat, sorghum, and barley—has its own distribution, defined primarily by bioclimatic  ranges (Table 2.1). 1. Information in this section, unless otherwise cited, is from the Atlas of the Ethiopian Rural Economy (Ethiopia, CSA; IFPRI; and EDRI 2006), whose agricultural data are drawn primarily from the 2001/02 Ethiopian Agricultural Sample Enumeration. Ethiopian Agriculture 23 Biophysical Envelopes of Ethiopian Staple Crops Rural livelihoods are strongly influenced by environment because biophysical  conditions and cropping patterns vary widely across Ethiopia (Figure 2.2). Non- biophysical factors, such as access to markets, labor availability, local demand (and markets) for food, and export linkages, may impose additional limitations on locally viable production options. Here we briefly describe some of the major  crops—cereals, tubers, and pulses—that make up the Ethiopian agricultural econ- omy, emphasizing the conditions under which they are produced. Teff (Eragrostis tef ) is the preferred staple food in much of the highlands. It is grown by nearly half of all farming households (to a greater extent if only highland households are considered) and accounts for 28 percent of all culti- vated land, more than any other single crop. Although traditionally grown in the highlands, teff can be grown under a wide variety of agroclimatic condi- tions, including elevations from zero to 2,800 meters above sea level (masl), KenyaUganda Gambella Amhara Sudan Eritrea Djibouti Somalia Somali SNNPR Oromiya Benishangul-Gumuz Tigray Afar Addis Ababa 10,021–50,000 50,001–100,000 100,001–150,000 150,001–500,000 �500,000 Cities � 10,000 Zones Wet wurch Moist wurch Wet dega Moist dega Wet weyna dega Moist weyna dega Dry weyna dega Moist kolla Dry kolla Bereha Null Trunk Main access International corridors Road type FIGURE 2.1 Characteristics of rainfall and altitude source: The agroecological zones depicted here are according to Hurni’s (1998) characterization. 24 Jordan Chamberlin and Emily Schmidt under a similarly wide variety of moisture, temperature, and soil conditions. Its optimal growing conditions coincide with its traditional production areas: an average elevation of 1,800–2,100 masl, average annual rainfall of 750 –1,000 millimeters (mm), and average annual temperature of 10–27°C. Maize is the second most widely cultivated cereal in Ethiopia in terms of area but is produced by more farms than any other crop. It accounts for the largest share of production by volume at 18.8 percent and appears to be increas- ing throughout Ethiopia. In addition to being a foodgrain, parts of the maize plants are also used as fodder, fencing materials, and cooking fuel. Although maize is becoming more widely grown, it is less tolerant of cold than teff, barley, and wheat. Teff can grow at elevations up to 2,800 masl; there is only limited maize production above 2,400 meters. In Ethiopia, the highest maize yields require an annual rainfall of 800–1,500 mm. More than 60 percent of maize production comes from Oromiya region, followed by Amhara, with about 20 percent of total production. Sorghum accounts for about 17 percent of all area planted with cereals and for about 18 percent of production. Relative to other cereals, sorghum is gener- TABLE 2.1 Traditional ecological zones and altitudinal ranges of selected crops Meters above sea level <900 900–1,400 >1,400 More than 3,700 High wurch (no crops) 3,700–3,200 Moist wurch Wet wurch (barley) (barley) 3,200–2,300 Moist dega Wet dega (barley, wheat, (barley, wheat, pulses) pulses, oilseeds) 2,300–1,500 Dry weyna dega Moist weyna dega Wet weyna dega (wheat, teff, maize) (maize, sorghum, (teff, maize, enset, teff, wheat, oilseeds, barley) oilseeds, barley, enset) 1,500–500 Dry kolla Moist kolla (sorghum, teff) (sorghum, teff, pulses, oilseeds) Less than 500 Bereha (only irrigated crops) source: Hurni (1998). notes: Crops in parentheses have restricted distribution within the zone and/or grow under less than ideal conditions in these areas. Blank cells indicate not applicable. Annual rainfall (mm) Primary Secondary Barley Maize Millet Sorghum Teff Wheat FIGURE 2.2 Dominant cereals in Ethiopia source: Ethiopia, CACC (2003). 26 Jordan Chamberlin and Emily Schmidt ally drought tolerant and is also accepting of excess water conditions. These characteristics give sorghum a large range of feasible climate regimes, although it grows best in semiarid conditions, especially in comparison to other cereals. For example, sorghum can produce grain in areas too dry for maize (those receiving less than 250 mm average annual rainfall). However, it is sensitive to cold temperatures and is rarely found at higher altitudes (2,500 masl may be considered a ceiling). Wheat accounts for similar shares of national cereal production as sor- ghum, with 17 percent of planted area and 19 percent of production. Varieties include Durum, Emmer, and so-called bread wheat. Durum wheat is often grown for domestic pasta fabrication. Emmer (aja in Amharic) is grown mainly in eastern Oromiya (Arsi and Bale) and Amhara (Shewa and Wollo). Bread wheat is widely grown throughout the highlands and transitional areas. Ethio- pian wheat production typically takes place at altitudes of 1,600–3,200 masl, in areas with average annual rainfall of 400–1,200 mm and average annual temperatures of 15–25°C. Unlike the other major cereals cultivated in Ethiopia, for which the highest cereal yields are achieved at lower elevations (between 500 and 2,300 masl), barley grows well at high altitudes, and its share of area cultivated generally increases with altitude. At elevations above 2,500 masl it is frequently the only cereal grown. Nationally, barley is grown by about a third of all cereal produc- ers and occupies about 9 percent of cultivated land. Many local varieties of barley are produced in Ethiopia, with a range of production and consumption characteristics such as rate of maturation and size of grain. Ethiopian barley varieties are reported to be relatively quick growing and more drought tolerant than other varieties (NRC 1996). In addition  to  the five major cereals grown  in Ethiopia, enset (Ensete ventricosum, sometimes referred to as false banana) is an important staple in large parts of the southern highlands, where it has been estimated that more than 10 million people depend on enset for food, fiber, and other uses (Brandt et al.  1997). Enset, which is cultivated solely in Ethiopia, is planted at elevations ranging from 1,100 to 3,100 masl, although its optimal range is 2,000–2,750 masl. Enset does not tolerate frost, and it is not drought resistant. Thus, opti- mal growing environments are humid and temperate. Annual rainfall in enset- growing areas ranges from 1,100 to 1,500 mm, with average annual tempera- tures of 10–21°C. Agroecologies and Farming Systems Farming systems, encompassing bundled sets of agricultural livelihood choices, including crop type and production technologies, are related to agroecological space because production choices must be viable given the available technol- ogy and the physical parameters required for plant growth. To the extent that Ethiopian Agriculture 27 the latter are reflected in agroecological zones, the spatial expression of tradi- tional farming systems in Ethiopia is at least partially coincident with agro- ecological zone maps. Westphal (1975) identified four major farming systems for Ethiopia: seed  farming, enset planting, shifting cultivation, and pastoral complexes. The seed- farming complex focuses on the production of grain, particularly cereals, but also pulses and oilseeds. Grain-based, seed-farming production systems are found throughout the central, northern, and eastern highlands and involve the majority of Ethiopian small farmers. Crop choice within the grain-based sys- tems varies widely, with these systems found from kolla (lowlands, between 500 and 1,500 masl) to wurch (highlands, between 3,200 and 3,700 masl), and in moisture conditions ranging from dry to wet. Similarly, the enset-planting complex (in moist and wet dega and weyna dega) allows for flexibility of pro- duction whereby enset could be the principal staple, a co-staple with cereals and/or tubers, or a minor component of cereal- or tuber-based systems. Finally, shifting cultivation and pastoral complexes are most common in the western and eastern lowlands, respectively. In the humid western lowlands (primarily moist kolla), disease is a limiting factor for livestock. The arid and semiarid lowlands in the eastern part of the country (mostly bereha, less than 500 masl) lack available moisture, which limits rainfed crop production. In both areas, population  densities  are  low,  reflecting  the  low  carrying  capacities  of  land  resources under current technologies. Characterization for Policy Guidance The classification systems described earlier are roughly contoured. They may  easily be (and frequently are) further refined on the basis of more detailed envi- ronmental data such as soil type, seasonality, terrain, technology specificity, or  local production idiosyncrasies. For example, the Ethiopian Institute of Agricul- tural Research (EIAR) organizes crop management research by 18 major and 49 minor agroecological zones, which are refined from the traditional agroclimatic  zones outlined earlier.2 For policy guidance, however, more detail may not be what is most re- quired for identifying overarching challenges to the agricultural sector and cor- responding investment priorities. Emphasizing the importance of moisture availability for the country’s rainfed production systems, the Government of Ethiopia has long framed basic agricultural policy discussions within the 2. EIAR was called the Ethiopian Agricultural Research Organization (EARO) until 2006. The Food and Agriculture Organization (FAO) was involved in identifying 18 major and 42 minor agroecological zones in the late 1980s, and EIAR has now subdivided some of these zones, with a total of 49 currently. These agroecological zones were developed under an FAO-led project within the Ministry of Agriculture, described by De Pauw (1987). 28 Jordan Chamberlin and Emily Schmidt “Three Ethiopias”: moisture-reliable highlands, drought-prone highlands, and pastoral lowland areas.3 For many rural experts, these basic regimes of mois- ture availability and the production systems therein are the critical distinctions when evaluating first-order strategic options for rural development across the  country. However, production choices are a function of not just biophysical endow- ments but also socioeconomic conditions and the man-made environment. These include the local availability of labor, demand for food, cost of transpor- tation between farms, and presence of input and output markets. Development domains are used in an attempt to build on basic information on agricultural potential by adding economic information within a framework of potential pro- duction choices. The Development Domains of Ethiopia Development domains were developed out of work by Pender, Wood, and col- leagues based on household- and community-level research in Ethiopia, Hon- duras, and Uganda (Pender, Place, and Ehui 1999; Wood et al. 1999; Pender et al. 2001a, 2001b; Pender, Scherr, and Durón 2001; Nkonya et al. 2004; Pender 2004a, 2004b; Pender et al. 2004; Pender, Ehui, and Place 2006; Pender, Place, and Ehui 2006). Drawing on the theory of comparative advantage and location theory (von Thünen 1826; Chomitz and Gray 1996) and the literature on the evolution of farming systems in tropical agriculture (Boserup 1965; Ruthenberg 1980; Binswanger and McIntire 1987; Pingali, Bigot, and Bins- wanger 1987; McIntire, Bourzat, and Pingali 1992), we can understand key components of an area’s agricultural development endowment in terms of a relatively reduced set of factors: agricultural potential, access to markets, and population density. Agricultural potential is determined by crop characteristics, inputs (includ- ing seed and fertilizer), and the biophysical environment. The income potential of alternative cropping patterns and livelihoods, however, depends on access to markets and population density, as well as agricultural potential. For example, an area with high and reliable rainfall and good soils may have an absolute advantage in producing high-value perishable vegetables but may have little comparative advantage in this livelihood if it is remote from markets. Popula- tion density, by affecting local land–labor ratios,  influences the comparative  advantage of labor-intensive livelihoods. High land–labor ratios in areas with poor access to markets and low agricultural potential endowments may encour- age labor-intensive but low–external input production strategies. All three fac- 3. This characterization has recently been expanded to “Five Ethiopias” (drought-prone, humid lowland moisture-reliable, moisture-reliable–cereals, moisture-reliable–enset, and pastoral- ist) for the Ethiopian Social Accounting Matrix developed by the Ethiopian Development Research Institute (EDRI). Ethiopian Agriculture 29 tors  together  influence  the profitability of different commodities, production  technologies, and land management practices. Implementing development domains means applying this conceptual frame- work to identify mappable conditions that broadly enable or constrain develop- ment options of national importance. This implies that two decisions must be made. First, the scale of strategic planning must be determined. For national strategy formulation, sectorwide perspectives require some level of generaliza- tion greater than that typically used in commodity- or technology-specific rec- ommendation domains. A second decision is how best to represent these factors given local conditions and the relevant scope of perspective. In practice, this means identifying the best indicators to represent relative levels of agricultural endowments for rainfed (and irrigated) agriculture, access to market opportuni- ties, and the availability of labor relative to land. In  Ethiopia,  Chamberlin,  Pender,  and Yu  (2006)  defined  development  domains by starting with the long-standing moisture regime distinctions fea- tured in policy discussion of the “Three Ethiopias” and further distinguishing between highland and lowland areas of rainfed agriculture. This yielded a total of  five  basic  zones  of  agricultural  potential  (moisture-sufficient  highlands,  moisture-sufficient lowlands, drought-prone highlands, drought-prone lowlands,  and pastoralist areas). Socioeconomic characteristics were also included in the characterization: two classes of market access (low and high) and three classes of population density (high, medium, and low). The resulting 25 domains were significant  predictors  of  a  range  of  rural  livelihood  variables  at  the woreda (district) level. These domains highlight important aspects of the Ethiopian rural context. Of particular note is the predominance of low market access conditions. In 1994, 92 percent of land resources and 69 percent of rural populations were located in areas with low market access, as defined here (Figures 2.3 and 2.4;  Tables 2.2 and 2.3). These conditions have improved considerably, with reduc- tions to 79 percent of land and 40 percent of population in 2007, and represent one of the dimensions of greatest change in Ethiopia over the past two decades.4 Nonetheless, the portion of the country’s land and population resources that may be fairly characterized as remote is still very high. Although high-density areas are becoming even denser (and are accounting for larger shares of total populations), almost half of Ethiopia’s land and people are still in low- and medium-density areas. Many of these areas are also characterized by poor mar- ket access. Although crop choice is heavily influenced by biophysical parameters (at  least partially captured in the agricultural potential dimension of the domains), 4.  The domain classification was made on the basis of woreda boundaries; we acknowledge that some shifts in categories may simply be artifacts of changing boundaries. Due to this modifi- able areal unit problem (Openshaw 1984), we restrict our discussion to broad changes in charac- teristics rather than shifts between individual domains. 30 Jordan Chamberlin and Emily Schmidt Moisture-reliable areas Pastoralist areas Highlands, high-density, high-access Highlands, high-density, low-access Highlands, low-density, high-access Highlands, low-