Table of Contents
Foreword ix
Acknowledgments xi
Acronyms and Abbreviations xiii
Executive Summary xv
Improvements in Health and Education Services 1
Changes in Outcomes over the Past 15 Years 2
Was Decentralization Responsible? 2
Decentralization and the Delivery of Basic Services 7
Phasing in Decentralization 7
A Framework for Understanding Service Delivery Outcomes 9
The Scope of Decentralization and Intergovernmental Fiscal Transfers 13
Subnational Structures 13
Decentralization of Functions and Responsibilities 15
Financing Social Services through Intergovernmental Transfers 18
How Did Decentralization to Woreda Level Affect the Delivery of Social Services? 37
Increases in Woreda-Level Spending Following Decentralization 38
Decentralization Appears to Have Improved the Distribution of Expenditures Across Woredas with Worse-off Woredas Benefiting the Most 40
After Woreda-Level Decentralization Began, Some Evidence of Lagging Woredas Partly Catching Up in Social Service Delivery 44
Teacher Redeployment Across Woredas and Within Woredas 46
Regression Analysis: Linking Expenditures with Outcomes 51
Decentralization's High Potential to Improve Service Delivery 53
Making Decentralization Work: Overcoming Constraints in Decentralized Service Delivery 57
Critical Constraints 57
Regional Innovations for Recruiting and Retaining Workers 74
Empowering Citizens and Communities to Improve Services and Outcomes 75
Improving Accountability Mechanisms 85
Appendixes
Calculating Block Grants Allocations From Federal Government to Regions and Regions to Woredas Using the Three-Parameter Formula 91
The "Fiscal Equalization" and "Unit Cost" Approaches for Block Grant Allocations 95
Regression Results on Education Outcomes Before and After Decentralization 101
Estimating the Effects of Decentralization on the Delivery of Human Development Services in Ethiopia 103
Methodology and Technical Details for Data Analysis in Chapter 4 107
References 109
List of Tables
Index of Real Government Expenditures and Spending as Percentage of GDP, 1999-2005 4
User Satisfaction with Government Health Services, 2005 6
User Satisfaction with School Services, 2005 6
Trends in Service Delivery, 1995/96-2004/05 8
Assignment of Expenditure and Revenue Responsibilities for Education, Health, and Water and Sanitation, by Tier of Government, circa 2005 16
A Wide Variation in Per Capita Block Grant Transfers to Regions 20
Two Different Approaches for Allocating Federal Resources to Regions 22
Regional Budgets and Share of Budgets Transferred to Woredas in Four Regions, 2005/06 28
Block Grant Allocations in Oromiya, 2002/03-2004/05 30
Block Grant Allocations in SNNPR, 2002/03-2004/05 30
Trends in Regional Expenditure, 1993/4-2005/06 31
Aggregate Fiscal Performance 35
Woreda-Level Spending on Education and Health in SNNPR, by Category of Spending, 2001 and 2004 42
Woreda-Level Per Capita Spending on Education and Health in Oromiya, by Category of Spending, 2001 and 2004 43
Woreda-Level Spending on Education and Health in SNNPR, by Type of Woreda, 2001 and 2004 43
Recurrent Expenditure Per Primary Student in SNNPR, 2001 and 2004 44
Recurrent Expenditure Per Secondary Student in SNNPR, 2001 and 2004 44
Education Outcomes in SNNPR, 2001 and 2004 45
Primary Education Outcomes in Oromiya, 2001 and 2004 45
Block Grant Transfers as a Proportion of Federal Revenues, 1996/97-2005/06 60
Number of Regional and Woreda-Level Personnel in Oromiya Before and After Decentralization 69
Staffing in Sector Offices in Delanta Dawnt Woreda, Amhara Region, 2003/04 70
Staff Vacancies in Selected Woredas 71
Community Contributions to Education in Boset Woreda, Oromiya Region, 2004/05 78
Community Contributions to Education in Boritcha Woreda, SNNPR, 2004/05 79
Per Capita Block Grant Transfers and Population 94
School-Level Regressions Explaining Educational Outcomes Before and After Decentralization 101
Woreda-Level Regressions Explaining Educational Outcomes Before and After Decentralization 102
Effects of Changes in Expenditures on Educational Outcomes in the Primary Education Subsector: Results of Woreda-Level Regressions for SNNPR 105
List of Figures
Primary Enrollment Increased Rapidly Beginning in the Mid-1990s 3
Increases in Primary School Enrollment Since 1995 Occurred Despite Only Modest Increases in Spending on Education as a Percentage of GDP 5
Decentralization has Devolved Responsibility to Subnational Levels of Government 9
Accountability for Providing Services Can Follow a Long Route or a Short Route 10
Federal Transfers to Regions Using Block Grants have been Rising but Not as Fast as Federal Discretionary Spending 19
A Very Close Inverse Relationship Between a Region's Population and its Per Capita Transfer from the Federal Government in 2005/06 21
Following Decentralization, Transfers from Regional Governments to Woredas and Zones Increased, Except in SNNPR, Where They Were Already High 23
Regional Budgets With and Without Special Purpose Grants 32
Regional Revenue as a Share of General Government Revenue, 2002/03-2004/05 32
Real Per Capita Government Spending on Education Rose After 2000, but Much of the Increase went to Tertiary Education 33
Overall Real Per Capita Spending on Health Increased Between 2000/01 and 2004/05, but Real Per Capita Subnational Government Spending Declined 34
Woredas' Share of Regional Recurrent Budgets for Amhara, Oromiya, SNNPR, and Tigray (the Four Main Decentralizing Regions) Together 36
Expenditures of Woredas Increased After Decentralization 39
Recurrent Woreda-Level Spending Rose in All Sectors in SNNPR Following Decentralization 40
The Share of Total Recurrent Spending for Primary and Secondary Education in SNNPR Rose Following Decentralization 41
Gross Enrollment Rates have Improved in All Categories of Woredas in SNNPR Since Decentralization 46
Grade 8 Pass Rates in SNNPR Have Increased Since Decentralization, with Improvement in Pastoral Areas Greater than in Urban Areas 47
Repetition Rates have Fallen Sharply in Urban Areas and Pastoral Areas in SNNPR Since Decentralization 47
Teachers have been Redeployed from Urban to Non-Urban Woredas Since Decentralization 48
Teacher-Section Ratios Tended to Equalize Since Decentralization Began in SNNPR 49
The Gap in Pupil-Teacher Ratios Across Woredas has Narrowed Since Decentralization 49
No Reallocation of Teachers or Narrowing of Gaps in Pupil-Teacher Ratios Appears to Have Taken Place in Oromiya Since Decentralization 51
Differences in Pupil-Teacher Ratios Across Schools Within Woredas in SNNPR have Narrowed Since Decentralization 52
Per Capita Block Grants to the Regions Averaged just Birr 82 Per Capita in 2004/05 58
Inadequate Financing Represents an Even Greater Constraint at the Woreda Level 59
Capital Spending by Subnational Governments as a Share of Total Spending has been Declining, Especially in the Social Sectors 64
Since Decentralization to the Woredas, There has been an Acceleration in the Production of Physical Capital Stock Related to Education and Health Service Delivery 65
Nursing and Medical Students in Ethiopia Perceive that the Assignment of Posts is Subject to Manipulation 74
Community Contributions Represent a Much Larger Source of School Financing Than Incentive Awards 81
Enrollment by Girls Rose in BESO Schools 81
School Attendance is Higher in Woredas in which Primary Education Continues to Grade 8 87
List of Boxes
The "Unit Cost" Approach to Block Grant Allocation 25
Performance Agreements in SNNPR 26
Devolution of Power in Theory and in Practice 27
Improving the Legal Framework for Decentralization in Tigray 62
Improving Financial Management Reporting at the Woreda Level 69
Community Participation in Service Delivery at Work: Ethiopia's Basic Education Strategic Objective (BESO) 80
Boosting Girls' Participation in Primary Education Through Girls' Advisory Committees 82
Community Participation in Water Supply, Sanitation, and Health Education Schemes in Oromiya 83
Using Citizen Report Cards in Ethiopia 86