Report No. 32675-ET

Ethiopia

Accelerating Human Development Outcomes in a Decentralized Service Delivery Context

Decision Draft June 16, 2005

AFTH3

Human Development Department

Africa Region

Document of the World Bank

CURRENCY EQUIVALENTS

(Exchange Rate Effective August 6, 2004)

Currency Unit / = / Birr
Birr 8.45 / = / US$1
US$1 / = / 1.46275 SDR
FISCAL YEAR
July 8 / - / July 7

ABBREVIATIONS AND ACRONYMS

AAA
BESO / Basic Education Strategic Objective
BOFED / Bureau of Finance and Economic Development
CSAP / CommunitySchool Activity Program
CGPP / Community Government Partnership Program
EMC / Expenditure Management and Control
EPRDF / Ethiopia Peoples’ Revolutionary Democratic Front
GACs / Girls’Advisory Committees
GOE / Government of Ethiopia
HEP / Health Extension Program
HIV/AIDS / Human Immunodeficiency Virus/Acquired Immunodeficiency Syndrome
IMCI / Integrated Management of Childhood Illnesses
KETBs / Kebele Education Training Boards
LIF / Local Infrastructure Fund
MDGs / Millennium Development Goals
PANE / Poverty Action Network of Ethiopia
PEPs / Public Expenditure Programs
NGO / Non-governmental organization
PSCAP / Public Sector Capacity Project
PTA / Parent Teacher Association
REB / Regional Bureau of Education
SDAs / School Development Agents
SDPRP / Sustainable Development and Poverty Reduction Programme
SIA / School Incentive Awards
USAID / United States Agency International Development
WEOs / Woreda Education Offices
ZCs / Zonal Coordinators
Vice President: / Gobind Nankani
Country Director: / Ishac Diwan
Sector Manager: / Laura Frigenti
Task Team Leader: / Andrew Dabalen

-1-

Table of Contents

Acknowledgments

Executive Summary-The promise of decentralization

1.Introduction

2.A framework to understand service delivery outcomes

Ethiopia’s context

3.the promise of decentralization

4.Strengthening local government to improve services and outcomes

Inadequate financing of compacts

Weak administrative capacity

Weak monitoring and enforcement

policy recommendations

5.empower citizens to imrpove services and outcomes.

Improving services through citizen participation

Improving services through provision of information

policy recommendations

6.Reduce the opportunity costs of using services b y poor households

policy recommendations

7.References

8.Annex: Tables and Figures

List of Boxes

Box 31:

Box 41

Box 42: Financial Management Reporting at Woreda Level

Box 43: Expenditure need based approach to block grant allocatin

Box 51: Community participation in service delivery at work-spotlight on BESO

Box 52: Boosting Girls’ Participation in Primary education Through Girls’ Advisory Committees (GACs).

Box 53: Community participation in Hittosa water supply, sanitation and health education scheme, Arsi Zone, Oromiya

Box 54: Ethiopia Pilot Citizen Report Card

Box 61: Measuring the social rates of return to policy interventions (20th Century United States)and Argentina

List of Tables

Table 41: Expenditure and Revenue Assignments in Education, Health and Water and Sanitation, Ethiopia (c. 2005)

Table 51: Major contributors to reductions in under-5 mortality.

Table 81: Summary Statistics, Ethiopia, Welfare Monitoring Survey 1995-2000

Table 82: Summary Statistics, By Predicted Wealth Quartile, Ethiopia, WMS 1995-2000

Table A 1: Basic child health regressions, Ethiopia.

Table A 2: Child health regressions with health facility attendance controls, Ethiopia

Table A 3: Health wealth quartiles, Ethiopia

Table A 4: Child Mortality /Morbidity (Children Born Dead), Ethiopia

Table A 5: Wage premium to doctors and nurses, Ethiopia

Table A 6: Bivariate probit model estimation results

Table A 7: Multinomial logit model estimation results

List of Figures

Figure 21: Key Relationships of Power

Figure 22: Percentile Rank on Governance indicators, Ethiopia, 19996-2002

Figure 23: Governance Structure in Ethiopia

Figure 31: New school construction, total schools and enrollment (grade 1-8), Ethiopia, 1967-2001

Figure 32: Public Education Spending and Enrollment Outcome, Ethiopia, 1982-2004

Figure 41: Total and Per Capita Block Grant Transfers to Regions, Ethiopia, 2004-05

Figure 42: Ethiopia, the share of regional transfers passed on to woredas

Figure 43: Total and per capita transfer to woredas, Ethiopia, 2004/05

Figure 44: Perceptions about the working obligation and lottery system

Figure 51: Community Contributions to School Finance Relative to Incentive Awards

Figure 52: Girls Enrollment in BESO Program Schoools

Figure 61: Attendance by age and level of primary schooling available in a woreda, Ethiopia, 2000

Figure 62: Participation in School and Work by Income Group, Ethiopia, 2001

Figure 63: Ethiopia, Kernel density estimates of distance to health center

Figure 64: School Participation and Work by Age, Urban Ethiopia, 2001

Figure A 1: Average score on Grade 8 National Learning Assessment Exam, Ethiopia 1999/2000 and 2003/04

Figure A 2: Average Score on Student Learning Assessment, Grade 8 Student,, By Region and Subject, Ethiopia 2003/04

Figure A 3: Average Score on Grade 4 Nattional Learning Assessment Exam, Ethiopia 1999/2000 and 2003/04

Figure A 4: Average Score on Student Learning Assessment, Grade 4 Students, By Region and Subject, Ethiopia 2003/04

Figure A 5: : Ethiopia: Trends in Health Expenditure, Public Recurrent and Capital, 1990/91—2001/02 (current and constant prices).

Figure A 6: School Attendance Status of Primary School Age Children (Ags 7-17) Ethiopia, 2001

Figure A 7: School and Work Participation Rates by Age, Ethiopia, 2001

Figure A 8: Reasons Why Children are not in School Ethiopia, 2001

Figure A 9: Reasons for Working by Sub-Sample of working children, Ethiopia, 2001

Figure A 10: Participation and Work by Age, Male Children Ethiopia, 2001

Figure A 11: Participation in School and Work, Female Children, Ethiopia, 2001

Figure A 12: School and Work Participation by Age, Rural Ethiopia, 2001

Figure A 13: Participation in School and Work, By Region, Ethiopia, 2001

-1-

Acknowledgments

  1. This report was prepared by a team comprised of Mmes/Messrs Andrew Dabalen (AFTH3, Team Leader), Nada Eissa and Mike Ingram (Georgetown University), Tor Halvorsen (University of Bergen) and Robert Langley Smith (LIND, Oslo University) and Ato Mammo Kebede (Addis Ababa), Rahel Kassahun (Consultant), Ephraim Kebede (Consultant), Mirafe Marcos (OED, World Bank), Gizaw Molla (Consultant), Endashaw Tadesse (AFTH3), William Wiseman (AFTH3).
  1. The report was prepared under the general guidance of Ms. Laura Frigenti (Sector Manager, AFTH3), Mr. Ishac Diwan (Country Director, AFC06), Jeni Klugman (Lead Economist, AFTP2) and Trina Haque (Lead Economist, AFTH3) and Harold Alderman (Lead Economist, AFTHD). Invaluable comments were received from the Ethiopia country team. In particular, we thank Anwar Bach-Baouab, Mesfin Girma Bezagawaw, Getahun Gebru, Jemal Omer, Christine Lao Pena, Agnes Soucat and Gary Theisen.
  1. The report benefited from comments by: Shantayanan Devarajan (Chief Economist, SASVP) and Elizabeth King (Sector Manager, DECRG).
  1. The team would like to express its special appreciation and gratitude to Woizerit Samia XXX (??), Deputy Commissioner/Manager (??), Central Statistical Authority, Ministry of Finance and Economic Development, and her team for facilitating access to much of the household survey data underpinning many of the results.

-1-

Executive Summary-The promise of decentralization

  1. Decentralization of service delivery holds great promise for Ethiopia as a way to accelerate the achievement of MDGs. However, the promise is threatened by several weaknesses in the service delivery chain. A major constraint is inadequate financing. But that is not the only hurdle. Another problem is that the capacity of the local governments, who are mandated with delivering the services key to achieving the targets, is weak. Equally weak is the role of citizens in holding service providers accountable for delivering quality services.
  1. The strongest evidence in favor of decentralization is the rapid increase in the coverage of basic services following the first phase of the decentralization process. Primary enrollment growth has been impressive and there are indications that the HIV/AIDS infection rates are either leveling or decreasing. But at current pace of progress the targets for primary completion, child and maternal mortality, and clean water and sanitation will not be attained. Only 30% of an entering cohort of children complete grade 6 and even lower (20%) complete grade 8. So far the rising completion rates have been sustained primarily by increasing entry rates, which will be harder to maintain as populations that are easier to reach are covered. Reaching the targets in child and maternal mortality are even more challenging. However, a few key weaknesses in the system pose the greatest challenge to achieving these targets.
  2. Inadequate financing is one major constraint. The revenue bases of nearly all woreda governments are weak. The woreda governments, who have the responsibility to deliver on these targets, rely almost exclusively on unconditional block grants. About 90% of the grant is spent on salaries and operational costs, so little is left for investment that would be essential to reach the goals. An additional problem is that the divisions of responsibilities, between tiers of government are not well-defined, thereby raising the possibility that government responsibilities are not well-funded (between federal and regional, between regional and woreda, and between woredas and kebeles). Although inadequate financing is a major constraint, resolving it will not be enough.
  3. Another is weak local government capacity. In many woredas, there is a shortage of skilled personnel. Most woredas also have poor infrastructure to support transformation: water, electricity, communication networks are either non-existent or old and obsolete.

Citizen participation and power vis-à-vis the Government and service providers is weak and is a major hurdle to strengthening accountability for results. Citizen involvement in holding service providers accountable is hobbled by poor information of local government finances and their relative performance on outcomes, as well as by limited choices in service providers and mechanisms for participating in institutions of monitoring service delivery such as PTAs, local health management boards, and so on.

  1. In addition, many households face huge opportunity costs of using the services. Even if all the four hurdles above are overcome, it may not be possible to reach the targets, if the opportunity costs of many households remain the same. It must be kept in mind that Ethiopia is, after all, a poor country and many poor households will find it difficult to use education and health facilities because the opportunity costs of doing so are very high. Many children are out of school because they are required to work in household enterprises (mostly farms). Distances to school and health facilities have been shown to reduce the use of education and health facilities. But the constraints go beyond the distance to these facilities. Even if distances to school and health facilities are not a binding constraint, distance to water and firewood reduce use of facilities. Effectively, the opportunity cost of adult time is transferred as a negative externality to their children. To accelerate the achievement of MDGs, it will be necessary to,

(a)improve local government capacity and financing, by

  • scaling up the implementation of Public Sector Capacity Project (PSCAP) and Local Infrastructure Fund (LIF);
  • reforming civil service remuneration packages within the on-going reform of the performance based civil service reform program, in order to improve retention and attract more qualified workers into the system;
  • introducing user fees, especially in the water sector, but also in education. This will relief the local government’s budget constraint and improve quality and accountability in the delivery of services.

(b)to strengthen systems of accountability in service delivery by,

1)Reforming the compact between tiers of government through,

  • establishing performance contracts between federal/Regional and local level governments,
  • extending the gim-gama systems to evaluate, not just individual’s improprieties as it currently does, but a group’s (say woreda government’s) performance in service delivery relative to well-defined and agreed goals.

2)
Empowering citizens to improve service delivery through,

  • institutionalizing citizen participation in institutions of local management of services, such as parent teacher associations, local health boards, etc.
  • institutionalizing citizen report cards, which can be an input into a general system of evaluation of local governments against agreed performance contracts.

(c)To reduce opportunity costs of using servicesby poor households by,

  • introducing conditional transfers in education and health.
  • undertaking public information campaigns.
  • adopting multi-sectoral approaches to achieveing education and health outcomes. To achieve education and health MDGs, it is not enough to tackle the constraints within these sectors. It may be necessary to improve the services in other areas, such as easy availability of safe water or cheap energy for households.
  1. Finally, there is a need for more systematic evaluation of key reforms. Ethiopia is in the process of implementing a major effort at transforming its political economy. As a consequence, much is going on in Ethiopia, and this report has focused on a few of the promising interventions being undertaken in making decentralization work in improving accountability and services in the social sectors. Yet there is no systematic attempt to evaluate many of these interventions to understand what works. This need not be. More relevant to the content of this report, there has been far reaching reforms in curriculum, structure of the education system, school fees (in education) and structure of the health systems, user fees, and mode of delivery of primary health services through health extension workers (in health). In order to ensure that the scaling up effort that is being proposed is informed by lessons learned from these reforms, it is necessary to undertake rigorous evaluation of these reforms. It will be 3 years since the start of decentralization to woredas in some regions. While stocktaking in the first couple of years would have been too soon, now can be an opportune time to undertake such an assessment simply because the transition issues would probably have been worked out and regional and woreda preferences would have began to emerge.

-1-

1.Introduction

1.1At the beginning of 2005, exactly a decade before the date the international community has set itself the goal of attaining specific millennium development goals (MDGs), Ethiopia can claim to have made much progress. Nearly 70% of Ethiopian children were enrolled in primary schools, which is a result of an impressive 9% annual growth in enrollment since 1993. Similarly impressive expansion has also been achieved at secondary and post-secondary levels. Immunization coverage stood at 60% from a low of 20% in 1994. Child malnutrition has fallen to 56% from 63% in 1995, while child (under 5) mortality declined to 14% in 2003 from 21% in 1990. More citizens also report visiting health facilities to seek treatment, from 0.3 visits per person in 2003 to nearly 0.4 visits per person in 2005. Population with access to clean water rose from 26% in 1995 to 47% in 2000.

1.2But, though impressive, these results tell a parallel story. First, it says that still far too many children remain out of school, are in poor health, and die from easily preventable diseases. In education alone, arguably the most successful effort of all, at least 4 million children are still not served by any school, while many who are within reach of a school cannot attend or are not learning for a variety of reasons (see Figures A1 and A2 for latest grade 8 results of national examination assessment results and Figure A3-A4 for grade 4). Second, these achievements still lag behind outcomes observed in other poor countries. Finally, for all these success stories there are many areas –in service quality, in equity of access, in complementary services-- where progress has not been as rapid or has deteriorated. In both education and health, quality of inputs (staff, books, drugs, etc.) is inadequate, while in water, sanitation and electricity, huge inequities of access between the rich and the poor remain.

1.3There are multiple reasons for low human development outcomes in Ethiopia. Understanding some of these reasons and what could be done to accelerate improvements in these outcomes is the subject of this report. A key message of the report is that decentralized governance structure which has been introduced in phases since 1991 holds great promise as a mechanism for accelerating the achievement of MDG outcomes. However, there are weaknesses that can derail these possibilities. It argues that in the current Ethiopia context, while policymakers, providers and citizens must work together to strengthen the accountability mechanisms, there is a particular need to strengthen local government and enhance the role of service beneficiaries, that is, the citizens.

1.4The rest of the report is organized as follows. After this introduction, we discuss briefly the framework for understanding service delivery outcomes, by looking at all the actors in various nodes of service provision. It also discusses how these relationships work in the context of Ethiopia. Chapter 3 provides a glimpse of the promise of decentralization via a short review of the accomplishments of the first phase of decentralization. Chapter 4 looks at the first key hurdle to maximizing the promise of decentralization, namely the financial and capacity problems of local governments. Chapter 5 examines the next big problem, the minimal role of citizen participation. It discusses in detail the potential gains in improved service delivery outcomes from involving citizen in the financing, monitoring and management of services. Chapter 6, also the last, cautions that even if the first two key constraints are overcome, MDG targets may become elusive simply because the opportunity costs faced by too many poor households are large enough to preclude them from using and benefiting from services.

1.5Before proceeding with the rest of the report, it is necessary to note the link between this report and other recent and related reports. Both the Education and Health Country Status reports (World Bank 2004a, 2004b) for Ethiopia have just been completed and they are rich both in content and policy recommendations. They examine in greater detail at determinants of education and health outcomes, respectively, and propose many insightful policy options, using both administrative and household survey data. The Poverty Assessment (World Bank 2004e), draws from and enriches the findings in the country status reports by linking these outcomes to poverty status. It also extends the analysis to cover more general question of economic opportunities and welfare. Finally, there is an Institutional Governance Review (2005) under preparation. The focus of this report is more on the actors and their role in accelerating the progress towards achieving MDG outcomes in Ethiopia. It therefore complements, rather than duplicates these recent AAA activities on Ethiopia.

2.A framework to understand service delivery outcomes

2.1Explaining the progress (or lack thereof), on key human development outcomes have enjoyed a resurgence from recent efforts of the international community to accelerate the attainment of these outcomes. In the past, much of the analysis focused on the interactions of behaviors and actions (preferences, effort, prices, income, etc.) of individuals and households and the availability, quality and financing arrangements within and, sometimes, outside (but related) sectors. Naturally, such a focus have promoted solutions that call for technical fixes and more budgets, both of which, while important, have been tried but found insufficient. To broaden the focus of the analysis, the World Development Report, Making Services Work for Poor People, 2004--- (henceforth WDR 2004) --- calls for a more encompassing approach. The key message of the WDR 2004 is that improving outcomes will require making services work for all, but especially for poor people. And making services work calls for understanding, establishing, and strengthening the “service delivery framework (or the chain)” (WDR 2004, p.48).