Africa Region Human Development Working Paper Series EDUCATION IN ETHIOPIA Strengthening the Foundations for Sustainable Progress A Summary of the Key Challenges Africa Region The World Bank II Contents Foreword ..................................................................................................................................vii Acknowledgements ....................................................................................................................ix 1. Introduction .....................................................................................................................1 2. Achievements and Concerns .............................................................................................3 3. Designing the Overall Policy Framework ..........................................................................7 4. Getting Resources to Schools, Communities and Households .........................................15 5. Transforming Resources in Schools Into Educational Outcomes ....................................21 Conclusion .................................................................................................................................25 Annex Tables and Figures ..........................................................................................................27 Table A1. Enrollments by Level in Government and Nongovernment Institutions, Ethiopia, 1967­68 to 2001­02................................................................................27 Table A2. Number of Students by Type and Level of Education, Ethiopia, 2001­02................28 Table A3. Primary Gross Enrollment Ratios (GER), Ethiopia, 1993­94 to 2001­02 ...............30 Table A4. Three Indicators of Entry to Grade 1, Ethiopia, 1993­94 to 2000­01.....................31 Table A5. Composite Cohort Survival Rates and Intercycle Transition Rates, Ethiopia, 1993­94 and circa 2000­01 .....................................................................32 Table A6. Efficiency of Student Flow in Primary Education, Ethiopia, 1993­94 and 2001­02............................................................................................................33 Table A7. Trends in Public Spending on Education, Ethiopia, 1980­81 to 2001­02................34 Table A8. Recurrent Public Education Spending Trends and Composition by Subsector, Ethiopia, 1993­94 to 2001­02...........................................................35 Table A9. Regional Recurrent Spending on Primary, Secondary, and Technical and Vocational Education and Training, Ethiopia, 2001­02 ....................36 Table A10. Teacher Wages and Educational Supplies as a Percentage of Recurrent Spending in Primary and Secondary Education across Regions. Ethiopia. 2001­02 ..........................................................................37 Table A11. Public Recurrent Spending Per Student by Level and Type of Education in Government Institutions, Ethiopia, 2001­02 ..................................38 III IV EDUCATION ETHIOPIA IN Table A12. Public Recurrent Spending Per Primary Pupil, Teacher Wages, Pupil­Teacher Ratios, and Spending on Inputs Other Than Teachers, Ethiopia and Selected World Regions, circa 2000 ....................................................39 Table A13. Gross Enrollment Ratios by Region, Ethiopia, 1993­94 and 2001­02 ....................40 Table A14. Gross Enrollment Ratios by Locality, Gender, and Wealth, Ethiopia, 1999­2000 ..............................................................................................................41 Table A15. Female Share of Primary and Secondary Enrollments across Regions, Ethiopia, 2001­02...................................................................................................42 Table A16. Distribution of Primary and Secondary Students and Schools by Sector, Ethiopia 2001­02 ....................................................................................42 Table A17. Prevalence of Double Shifting and Section and School Sizes of Government and Nongovernment Primary and Secondary Schools, Ethiopia, 2001­02...................................................................................................43 Table A18. Distribution of Teachers by Gender and Qualification in Government and NonGovernment Primary and Secondary Schools, Ethiopia, 2001­02...................................................................................................44 Table A19. Average Weekly Teaching Loads by Teacher Certification and Grade in Which Teacher is Teaching, Oromiya, Ethiopia, 2002­03....................................45 Table A20. Incidence of Poverty by Gender and Education of Household Head, Ethiopia, 1999.........................................................................................................46 Table A21. Unemployment Rate by Educational Attainment and Age, Ethiopia, 2001 ..............46 Figure A1. Enrollments in Regular Programs in Higher Education, Ethiopia, 1985­02.............47 Figure A2. Primary Gross Enrollment Ratio, Ethiopia and Other Countries, circa 2000...........48 Figure A3. Secondary Gross Enrollment Ratio, Ethiopia, Selected African Countries and Country Group Averages, circa 2000................................................................49 Figure A4. Relation between GDP Per Capita and Coverage in Higher Education in Low-Income Countries, circa 2000......................................................................50 Figure A5. Trends in Recurrent Spending on Selected Components of Education, Ethiopia, 1993­94 to 2001­02................................................................................51 Figure A6. Spending Per Pupil in Primary Education, Ethiopia and other African Countries, circa 2000.........................................................................52 Figure A7. Cohort Entry Rate to Grade 1 and Completion Rates in Grades 4 and 8 Across Regions, Ethiopia, circa 2000.............................................................53 Figure A8. Primary School Pupil-Teacher Ratios in Ethiopia (2001­02) and other Countries (circa 2000)....................................................................................54 Tables Table 1: Low Rates of Entry to Grade 1 and Survival to Grade 4, Especially in Rural Areas Despite Significant Improvement Since 1993­94 ...................8 Table 2: High Teacher Cost Leads to Adverse Tradeoff Against other School Inputs in Ethiopia ............................................................................................13 Table 3: Urban-Rural Disparities in Primary School Participation are Especially Wide in Ethiopia .........................................................................................15 Contents v Table 4: Primary Schools are Still too Inaccessible for Many Children in Rural Ethiopia ........................................................................................................16 Table 5: The Direct Cost of Primary Schooling can be Significant for the Poorest Families ....................................................................................................18 Figures Figure 1: Historically Unprecedented Growth in School Enrollments since 1993­94 ...................4 Figure 2: Persistent Deterioration of Pedagogical Conditions in Schools Since 1994 ....................4 Figure 3: Persistent Deterioration of Pedagogical Conditions in Schools Since 1994 ....................5 Figure 4: Recurrent Spending on Education Has Risen But Share of Primary Education Still Needs to Grow ......................................................................10 VI Foreword T his paper summarizes recent progress resources and to provide it in ways which in the education sector in Ethiopia, maximize governments' own powers of deci- highlightssomeoftheremainingchal- sion-making. In Ethiopia, the Sustainable De- lenges, and presents some policy op- velopment and Poverty Reduction Programme tions. It is based on the analysis undertaken for (SDPRP) provides the overarching framework preparingthecomprehensiveEducationCountry for poverty reduction and the Education Sector Status Report (CSR) by a team from the Minis- Development Plans feed into this. The results of try of Education and the World Bank--though the analyses undertaken for the Ethiopian Edu- neither organization explicitly endorses all of cation CSR report are being used in preparation the views expressed. That report provides a of the Government's most recent version of the detailed snapshot of the education sector up to Plan for 2005/06 to 2010/11. 2001­02 (and, in some cases, up to 2002­03) The SDPRP requires the sector to progress using both administrative data and information toward universal primary education, improve from household surveys.1 The focus is mainly the quality of instruction and learning achieve- on costs, finance, and service delivery, and their ments throughout the system, and produce a impact on learning achievement, in primary and trained workforce that is responsive, in quantity secondary schooling, in an effort to discover and skills mix, to the demands of the country's potentially important areas for further policy modernizing economy. At the same time, de- development. In addition to summarizing the centralization is opening the way for regional mainreportthispublicationalsoincludesasetof and woreda governments (and through them, supplementarytablesandfigureswhichdescribe local communities) to take greater responsibil- many aspects of the educational system.2 Welldevelopedsectorstrategies,underpinned bythekindofthoroughanalysisaccomplishedin theCSR,havebecomeincreasinglyimportantas 1 The main report has the same title as this sum- governments renew their commitment to reduce mary and can be ordered from the World Bank poverty and develop detailed strategies, and as bookstore at www.worldbank.org 2 The sources for all the information in the tables the international donor community pledges to and figures in this publication can be found in the provide a larger part of the needed financial main report. VII VIII EDUCATION ETHIOPIA IN ity, financial and otherwise, for managing their supported including the Government's Educa- own affairs, including the delivery of social tion Sector Development Programmes I and services such as education. In responding to the II. The most recent support has been for the challenges, many questions face the education Post Secondary Education Project which aims sector, including: to help improve the quality of the university system and parts of technical and vocational · Isthecurrenteducationpolicyframeworkad- education and training. This financial sup- equate for fulfilling the high expectations? port for projects and programs has provided · What resources are available to achieve the continuity in the Bank's engagement in policy stated goals and are they sufficient to meet development and implementation in the sector the requirements; and if not, how will the over many years. gap between resource availability and need In addition to project financing, World Bank be closed? assistance to client countries also takes the form · What changes in the financing of education, of analytical activities and advice. In the case in resource allocation across sub-sectors and of Ethiopia, this publication, the larger report, schooling inputs, and in the arrangements for and a recent study of higher education, are all servicedeliverywillhelptheeducationsystem examples of such assistance. Like project and to develop over the long run in a financially program financing, the ultimate objective of and pedagogically sound manner? these activities is to help countries to expand educationalopportunitieswhileimprovingqual- The full CSR, and this summary, suggest ity throughout the system. Unlike these projects that pressures on resources in the sector do and programs, however, this type of assistance necessitate some re-ordering of the immediate focuses on improving understanding of the edu- objectives and the adoption of some alterna- cation sector and diagnosing major constraints tive modes of service delivery, as well as more on its further development so as to help support financial resources. the search for appropriate policy responses. It The work that is summarized in this publica- creates a larger knowledge base for deepening tion has resulted from collaboration between the current understanding of the education sys- members of the Ministry of Education and the tem, for engaging a diverse audience in dialogue World Bank and has been funded in part by the on education sector policies, and for developing Norwegian Education Trust Fund. I am pleased a shared vision for the future. that this collaboration was initiated early in the ThissummaryoftheEthiopiaEducationCSR process and that it was maintained through all is presented in Amharic as well as in English in the subsequent stages of collecting and analyz- an effort to widen further the dissemination and ing data, report writing and dissemination, and discussion and it is my hope that its publication policy dialogue. This interaction is critical not in an easily available format will make a small only for undertaking the analyses but also for buteffectivecontributiontothecountry'soverall ensuring that the important issues raised receive efforts to define educational policies which will consideration at the highest appropriate levels have a positive impact on reducing poverty in of Government. Ethiopia. More broadly, it is noteworthy that World Bank engagement in educational development Ishac Diwan in Ethiopia dates back to the 1960s. Since Country Director for Ethiopia, then, many projects and programs have been World Bank Acknowledgements T his summary was prepared by Keith many others who also contributed in one way Hinchliffe and Jee-Peng Tan, based or another, special acknowledgements are owed on a comprehensive study published to Aleazar Tilahun; Yezid Reshid (and his col- under the same title as a World Bank leagues working on the Education Management Country Study. The full study was a product Information System at the Oromiya Bureau); of close collaboration between the Ethiopian Jutta Franz and Horst Sommer of the German Ministry of Education and the World Bank. The TechnicalCooperation(GTZ);BerndtSandhaas Ethiopian team was led by Setotaw Yiman, then of Institute for International Cooperation of the HeadoftheFederalMinistryofEducation'sPro- German Adult Education Association; Mulat gramming and Planning Department under the Demeke of the Addis Ababa University; Ale- guidance of Minister of Education Woiz. Genet; bachew Tiruneh; and Tesfaye Kelemework and and comprised Awash Gebru, Belete Demissie, Thomas Tilson of the United States Agency for Kassaw Ali, Mebratu Birman, and Workiye International Development; and World Bank Tegegn (members of the Steering Committee) colleagues Laura Frgenti, Getahun Gebru, Gary and Damte Demeke, Adefris Belachew, Almaz Theisen,TrinaHaque,MesfinGirma,MamyRa- Beyene, Amare Bizuneh, Getachew Tadesse, kotomalala, Lianqin Wang, Dina Abu-Ghaida, Getahun Workneh, Masresha Geleta, Mulugeta EleniAlbejo,SouthsavyV.Nakhavanit,Cornelia Mekonnen and Teshome Lemma (members of Jesse, and Elsie Maka. Publication of this sum- theTechnicalWorkingGroup).TheWorldBank maryhasbeenmadepossiblebyagrantfromthe team was led Jee-Peng Tan and included Gerard Norwegian government through a grant chan- Lassibille,JulieSchaffner,AndrewDabalen,Ten- neled via the Norwegian Education Trust Fund zin Chhoeda and Kiong Hock Lee. Among the to support educational development in Africa. IX X CHAPTER 1 Introduction T his paper summarizes recent progress and their children; and to the public at large for in the education sector in Ethiopia, whom a strong education system is important highlights some of the remaining for the country's overall economic and social challenges, and presents some policy well-being.Inadditiontosummarizingthemain options. It is based on the analysis undertaken report, in English and Amharic, this publica- for preparing the Education Country Status tion also includes a set of supplementary tables Report (CSR) by a team from the Ministry of and figures which describe many aspects of the Education andtheWorldBank--thoughneither educational system.3 organization explicitly endorses all of the views TheEthiopianGovernmenthasmadepoverty expressed. That report provides a detailed snap- reduction the centerpiece of its development shotoftheeducationsectorupto2001­02(and, strategy and has continued to advance the pace insomecases,upto2002­03)usingbothadmin- of decentralization. These policy directions are istrative data and information from household important for the education sector. Ethiopia's surveys. The focus is mainly on costs, finance, SustainableDevelopmentandPovertyReduction andservicedelivery,andtheirimpactonlearning Program (SDPRP) requires the sector to: achievement, in primary and secondary school- ing,inanefforttodiscoverpotentiallyimportant · progress toward universal primary educa- areas for further policy development. The sum- tion, mary is aimed at a wide audience. In the first · improve the quality of instruction and learn- instance, it is directed toward Ethiopia's policy ing achievements throughout the system, makers and managers in the education sector at and all levels of government, and toward education · produce a trained workforce that is respon- practitioners and researchers. It should also be sive,inquantityandskillsmix,tothedemands of interest to non-governmental organizations of the country's modernizing economy. which are active in providing various education services and to donors; to teachers and other school personnel, in both the government and 3 The sources for all the information in the tables non-government sectors, who are at the front and figures in this publication can be found in the line agents in providing education; to parents main report. 1 2 EDUCATION ETHIOPIA IN At the same time, decentralization is opening between resource availability and need be the way for regional and woreda governments closed? (and through them, local communities) to take · what changes in the financing of education, greater responsibility, financial and otherwise, in resource allocation across sub-sectors and for managing their own affairs, including the schooling inputs, and in the arrangements for delivery of social services such as education. servicedeliverywillhelptheeducationsystem In responding to the challenges, many ques- to develop over the long run in a financially tions face the education sector, including and pedagogically sound manner? · isthecurrenteducationpolicyframeworkad- This paper hopes to contribute to the con- equate for fulfilling the high expectations? tinuing dialogue on these, and other, strategic · what resources are available to achieve the questions and to offer some policy options for stated goals and are they sufficient to meet wide discussion. the requirements? If not, how will the gap CHAPTER 2 Achievements and Concerns Remarkable Progress since 1993­94 secondary education rose from 8 to 12 percent in the same period, while that for higher educa- T here has been a dramatic growth in tion climbed from 0.5 to 1.7 percent. enrollments throughout the educa- tion system in recent years. Aggregate enrollments in Grades 1­12 rose at Some Emerging Signs of Distress in a steady pace of around 9 percent the System a year between 1992­93 to 2001­02; and in Grades 1­4, the first cycle of primary school- In spite of these advances at least one third of ing, they grew even faster at an average of 15 all children are never entering school and only percent a year. By 2002­03, the education sys- around 60 percent of those who do enter sur- tem had grown to around 8.6 million students vive to the end of the first four year cycle. In in primary schools, with more than 600,000 addition,whileveryimpressive,thesystem'sex- in secondary schools. Enrollments in technical pansionhasnotbeenproblem-free.Theclearest and vocational training and education grew distress signals are the deteriorating conditions from less than 3,000 students in 1995­96 to in classrooms throughout most of the country. an estimated 54,000 students in 2001­02. In Since 1993­94, the pupil-teacher ratio, the pu- post-secondaryeducation,enrollmentshavealso pil-section ratio (the average number of pupils ballooned, from around 18,000 in 1990­91 to in a sction) and the real spending per student more than 48,000 in 2001­02. These trends are on non-salary inputs have steadily worsened a remarkable achievement given the patterns of (figure 2). In 2001­02, Ethiopia's pupil-teacher stagnation, reversals and uneven growth in the ratios of 65:1 in government primary schools past (figure 1). and 52.1 in government secondary schools The growth in enrollments has in turn in- were among the highest in the world. Inevita- creased the gross enrollment ratio (GER), a bly, section sizes at both levels have also risen common indicator for measuring coverage, at to extremely high levels, averaging about 74 all levels in the system. In primary education, and 79 students per section, respectively. It is it more than tripled between 1993­94 and not unusual to find classes with more than 100 2001­02, from 20 to 62 percent. The ratio for students. 3 4 EDUCATION ETHIOPIA IN Figure 1 Historically Unprecedented Growth in School Enrollments since 1993­94 Haile Selassie's reign Mengistu's rule New Governm ent 1994­ 1947­74 1977­91 7.0 6.0 Severe famine Severe famine New constitution 5.0 (millions) Gr. 1­4 4.0 students 3.0 of Gr. 5­8 2.0 Number 1.0 Gr. 9­12 0 967­681 969­701 971­721 973­741 975­761 977­781 979­801 981­821 983­841 985­861 987­881 989­901 991­921 993­941 995­961 997­981 999­001 001­022 Note: data include only students in regular programs in government and non government schools. Figure 2 Persistent Deterioration of Pedagogical Conditions in Schools Since 1994 90 9 80 8 (PSR) 70 7 Birr) ratio 60 6 (1995 section 50 5 pupil student or per 40 4 (PTR) ratio 30 3 spendingy 20 2 salar teacher Non Pupil 10 1 0 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 PTR, Gr. 1­8 PSR, Gr. 1­8 Non-salary spending PTR, Gr. 9­12 PSR, Gr. 9­12 Achievements and Concerns 5 Figure 3 Persistent Deterioration of Pedagogical Conditions in Schools Since 1994 80 70 o Rati 60 Section Pupil 50 or eacherT 40 Pupil 30 20 1994­95 1995­96 1996­97 1997­98 1998­99 1999­00 2000­01 2001­02 2002­03 Urban PTR Rural PTR Rural PSR Urban PSR Note: because data are available separately for government and non-government schools only after 1997-98, the figure uses data for both sectors in all years. Data after 1997-98 indicate that the difference is small between the government only series and that for both sectors. Conditions are Especially Desperate in translates into much heavier teaching loads for Rural Primary Schools teachers assigned to rural than urban schools. In Oromiya, for example, the weekly load Conditions have worsened particularly rapidly in rural schools averages 29 hours in Grades in rural schools (figure 3). By 2002­03 pupil- 1­4 and 24 hours in Grades 5­8, compared teacher ratios in these schools exceeded those with the corresponding averages of 22 and 18 in urban schools, by an average of 60 percent hours, respectively, in urban schools. These in Grades 1­4 and 50 percent in Grades 5­8. disparities imply that in rural schools, there is Similarly, classrooms in rural Ethiopia are as almost no scope for specialized teaching; they crowded as they are in urban areas whereas the also make rural teaching jobs even more unat- patterninmostcountriesisforruralclassrooms tractive than they already are, and increase the to contain far fewer students. The combination difficulty of rationalizing teacher deployment of high pupil-teacher ratios and section sizes across schools. 6 EDUCATION ETHIOPIA IN CHAPTER 3 Designing the Overall Policy Framework E ducationinEthiopiaisatacrossroads Goals and Targets today.Isthemodelthathassupported thesystem'sexpansionsince1994still Ethiopia's SDPRP states that the education sec- validtoguideitsfuturedevelopment? tor is expected to help reduce poverty by pro- If the answer is negative, as the data on pupils gressingtowardstheuniversalizationofprimary per teacher and per section increasingly appear educationandbyproducingaworkforcecapable to suggest, how then might it be adapted to the of filling jobs requiring skilled labor. What do conditionsinEthiopiatodayandintheyearsto these goals mean in practice? In the process of come? Beyond this, some searching questions moving towards a primary education of eight also need to be addressed on ways to improve years for every child in the country, is it useful to management of the system, so that all avail- setintermediategoals?Ifso,whatmighttheybe? able resources make their fullest contribution Does producing a skilled workforce responsive to the sector's development. There are at least to the demands of the economy simply mean four major areas where new thinking would turning out as many graduates as the education be useful: system can put through the upper levels of the system, or should expansion be linked closely · the educational system's goals and targets; to the demand for such labor? · the financial allocations, especially for pri- These are all difficult questions. As a basis for mary education; debate, some suggestions are made below. · the diversification of options to manage the costs of educational services; and · Provide four years of quality schooling for · the key indicators for monitoring progress; every single child as an immediate and practi- cal priority. These are discussed further below and some suggestions are made. The recent study of poverty in Ethiopia cal- culates that if all adults in the country had at least four years of primary education, the share of households living in poverty would drop by 18percent.Thissizableimpactisconsistentwith 7 8 EDUCATION ETHIOPIA IN thewidespreadagreementamongpolicymakers cohort ever enrolling in Grade 1, and less than and analysts that the core skills that primary twothirdsofthesereachingGrade4(table1).In educationimparts--literacyandnumeracy--can otherwords,lessthan40percentofchildrencur- boostachild'slifechances,evenifheorsheeven- rently complete four years of schooling. Under tually earns a livelihood in agriculture. Because thesecircumstances,implementationconstraints fiveyearsofprimaryschoolingaregenerallycon- alonewouldprobablymakeuniversalizing eight sidered a minimum for a person to become per- years of schooling for the whole population an manently literate and numerate, a large number impractical goal in the short- or even medium- ofcountriesmadethecommitmentintheUnited term. In this context, putting the emphasis on Nation's 2000 Declaration on the Millennium getting every child into Grade 1 and making Development Goals (MDGs) to ensure that all sure that they complete the first four year cycle children would be able to complete at least five (by reducing the very high dropout rate) might years of primary schooling by 2015. be the first priority followed by an expansion Ethiopia's education system has a 4­4­2­2 of Grades 5­8. Shifting the emphasis in imple- structure which at first sight makes it somewhat mentation would not change government's long awkward to state a clear cut goal for coverage. term policy of universalizing the completion of Given the current starting point, universalizing all eight years of the primary cycle. eight years of schooling is probably financially and practically unviable even in the medium · Adaptshortandmediumtermgoalsforcover- term, while universalizing only four years seems age to the different conditions in urban and too modest in light of international experience. rural areas. In practice, at least in the short/medium term, the issue is simple to resolve because coverage Because the current conditions differ so is still modest, with only 60 percent in each age widely across rural and urban areas, setting Table 1 Low Rates of Entry to Grade 1 and Survival to Grade 4, Especially in Rural Areas Despite Significant Improvement Since 1993­94 2001­02 Indicator 1993­94 Overall Urban Rural Entry rate to Grade 1 a/ 0.37 0.61 0.92 0.46 Survival rates from Grade 1 To Grade 4 0.52 0.59 0.77 0.55 To Grade 8 0.39 0.36 0.80 0.20 To Grade 12 0.15 0.08 -- -- Transition rates Grades 4 & 5 b/ 0.94 0.88 1.07 0.79 Grades 8 & 9 0.84 0.91 -- -- Grades 10 & 11 0.86 0.33 -- -- Notes: a/Data for this indicator refer to 1995­96 and 2000­01, respectively. b/The transition in urban areas in 2001­02 is slightly above 1.0, suggesting possible underlying data flaws. Designing the Overall Policy Framework 9 different time-bound targets for these areas is workers are either self-employed or used as a practical necessity. In urban areas the goal of unpaid family labor--features which put a limit universalizing eight years of primary schooling on the economy's capacity to absorb large num- is well within reach, given that most children bers of educated labor into modern sector jobs. already enter Grade 1, and the majority who Data from the most recent labor force survey in survive past Grade 2 eventually reach Grade 8. 1999suggest,forexample,thatforrecentschool Themainchallengehereistominimizedropping leavers the unemployment rate was about 25 out between the first two grades. In rural areas, percent among those attaining Grades 5­8 or by contrast, less than half the children in each vocational/technical education and 44 percent age cohort enter Grade 1, and of those who do, among those attaining general secondary edu- only 55 percent survive to Grade 4 and only cation. Only among university graduates was 20 percent, to Grade 8. These patterns imply the unemployment rate at a modest 4 percent; that over the next few years, universalizing the and with the large increase in graduates over completion of the first four years of primary the past five years even this favorable situation education in rural areas will itself be a demand- may change unless jobs outside of the public ing goal. administration(whichhaveseenalargeincrease in recent years because of decentralization) · Allow labor market conditions to influence become the major source of employment in the the pace of expansion, particularly beyond future. These patterns underline the importance Grade 8. of aligning the pace of expansion of post-basic education to labor market conditions. As the economy matures, the demand for educated labor typically expands, making it · Improve the education system's responsive- increasingly appropriate and feasible to univer- ness to labor market signals by increasing salize eight, or even more, years of schooling. information, encouraging private sector pro- At issue is how fast to expand the education vision of post-primary education and manag- system. If the system grows too fast and pro- ing student flows. ducesmoregraduatesqualifiedbeyondthecore skills of basic literacy and numeracy than can To foster a closer link between the education be absorbed in the prevailing job market, the system and the labor market will require not investment would be wasted to the extent that only regular monitoring of the labor market to school leavers fail to find work appropriate to generatetherelevantdata,butalsofairlyflexible their level of training and so lose the expected and responsive arrangements for the provision return on their investment. With frustrated of schooling. An option here is to enlarge the job seekers on the streets, the situation could role of the private sector at post-primary levels. easily escalate into serious social unrest. In Because private providers charge fees to cover part to minimize this risk and to institute a most,ifnotall,ofthecostofservices,parentsare merit-based system to allocate scarce places in likely to insist on getting a reasonable return on post-primary education, most governments in their investment, and providers tend to respond low-income countries use standardized exami- morenimblytolabormarketsignalsindesigning nations to regulate the flow of students through their course offerings. the system. Beyond encouraging private providers, the In Ethiopia the demand for educated labor government can also tighten the management has been growing but only slowly. Nearly 80 of student flow through its own institutions. In percent of the country's workforce is still in recent years, the pressure on preparatory and agriculture, and more than 90 percent of the higher education has risen sufficiently for the 10 EDUCATION ETHIOPIA IN Figure 4 Recurrent Spending on Education Has Risen But Share of Primary Education Still Needs to Grow Overall government recurrent spending Education recurrent spending 250 Percentage shares in 2001­02: 29 % for Grades 1­4 21 % for Grades 5­8 200 12 % for Grades 9­12 17% for higher education 11% for system administration 150 10% for other spending (1993­94=100) spending 100 Recurrent 50 0 1993­94 1994­95 1995­96 1996­97 1997­98 1998­99 1999­00 2000­01 2001­02 Note:thespendinginbothseriesareexpressedin1994­95constantBirr,indexedtospendingin1993­94,thefirstyearintheseries; sharesofspendingin2001­02includesamountspentonadministrativeoverheadsoffederalministryandregionalbureaus,prorated by sub-sector in proportion to their share of non-administrative spending. government to limit the number of places in Financial Allocations, Especially for Grade 11, and to divert the remaining Grade Primary Education 10 completers who do not directly enter the labor force to programs in teacher training, During most of the 1980s and 1990s, education and in technical and vocational education and was not a priority for public spending. Since training. A selection mechanism also exists in 2000, however, the government has been spend- Grade8wherestudentssitstandardizedregional ing more on education and aggregate public examinations. However, as table 1 shows, in recurrent spending on education has now been practice there is at present little or no selection restored to three percent of GDP, similar to the at this point, as nearly 91 percent of the stu- rate in 1993/94 (figure 4). Much of the increase dents eventually continue to Grade 9. This fact, hasgoneintoadministrativeoverheadsandpost- coupled with the absence of a formal selection secondary levels of education, however, leading mechanism earlier than Grade 8, may not mat- toasignificantshiftintheallocationofspending. ter much at present because nearly two-thirds The share of primary education--50 percent of of each cohort of first graders eventually drop the total in 2001­02--remains much less than outlongbeforetheyreachthislevel.Butasbasic the international benchmark of 67 percent (for education expands and survival rates improve, eight years of schooling), which is based on the as they should, there may be a need for a more pattern of spending in countries that have made active selection mechanism at Grade 8. good progress toward universalizing primary Designing the Overall Policy Framework 11 school completion. Two broad recommenda- since the 1970s, averaging about 5 percent of tions follow. the total. This level of participation is low even by low-income country standards, so the scope · Continue to allocate more to education while forencouragingthegrowthofprivatesecondary ensuring wise use of the resources. education should be quite promising. Lookingtothefuture,theincreasedattention being given to education, both domestically Diversifying Options for Managing the and internationally, implies that the prospects Costs of more resources for education are good. Yet the magnitude of the increase will inevitably Giving more resources to primary education be constrained by the competition from other would not remove the need to make tradeoffs sectors for the government's limited resources. within the sub-sector itself. The distress signals Among low-income countries, very few have already present in the system suggest that a managed to allocate levels of recurrent spend- simplereplicationofthecurrentmodelofservice ing on education beyond four or five percent deliveryisunlikelytobeeffective.Forthesystem of their GDP. Thus, if Ethiopia faces the same to continue to expand rapidly in a financially internal pressures as the governments in these viable manner and without compromising its other countries, the amount of additional re- ability to function well would probably require sources that the sector can count on is likely to significant change in some key dimensions of be limited even under the best of circumstances. service delivery. Broadly the options for manag- If so, tradeoffs within the education sector will ing schooling costs fall into two categories: beunavoidable.Toachievethegoalofuniversal primary school completion, primary education · those that involve the use of alternative deliv- up to Grade 8 will require a bigger share of re- ery arrangements; current spending in education than it currently · and those that involve lower cost inputs or receives. increasedefficiencywithintheregulargovern- ment sector. · Tap into households' willingness to spend on post-primary education. In both cases, a useful approach would be to make the list of sensible options as wide Because of likely constraints on the public as possible, and empower and motivate deci- purse, household contribution for post-pri- sion makers on the ground in the regions and mary education may need to increase as part of woredas to choose among them and to adjust an overall strategy for the education sector in the national criteria and standards in order to Ethiopia. Mobilizing this contribution at post achieve the educational goals within the budget primary levels would help the government to (and human) constraints they face. prioritize primary education in the allocation Some of the options are discussed below. It of public spending. In Ethiopia, private post is worth restating that these are meant to of- secondary education and training institutions fer promise as short to medium term solutions have begun to flourish in recent years, and a enabling Ethiopia to meet its enrolment targets continuation of this trend would make it more while providing sufficient resources to ensure possibleforthegovernmenttoprioritizeprimary that learning meets acceptable standards. educationintheallocationofpublicspending.In secondaryeducation,theshareofenrollmentsin · Support alternative basic education centers privately-financed institutions has been modest (ABECs) as an option for service delivery. 12 EDUCATION ETHIOPIA IN In Ethiopia, the ABECs provide a three year schools(morethan20percentofthegovernment low-cost schooling covering the first four years primary schools enroll more than 900 pupils; of the primary schooling syllabus. By compress- and about 25 percent of government secondary ing the program, the centers are particularly schools enroll more than 2,500 students). And responsive to the needs of the large numbers at the same time many villages in Ethiopia do of overage children currently seeking primary withoutagovernmentprimaryschoolandmany schooling. Although the centers have achieved children must travel long distances to get to a reasonably good outcomes in terms of student school. learning, the majority operate without govern- In Ethiopia, the use of criteria for teacher re- ment funding. As a result, they still only serve cruitment that are tied to civil service pay struc- a limited number of pupils totaling around tures has reduced the recurrent budget available 500,000 in 2001­02, or about 8 percent of for other school inputs that are also essential for total enrollments in Grades 1­4. If alternative effectiveteachingandlearning.Acomparisonof basic education is to play a more central role in the pattern of resource allocation across school universalizing primary schooling, it will need inputsinEthiopiaandothercountriesillustrates to be financed by the government, and the this point (table 2). Expressed relative to the per centers will need to be made an integral part of capitaGDP,theamountthatEthiopiaspendsper the whole system rather than left to continue student in Grades 1­8 is comparable to what as ad hoc experiments. In the process it will the average low-income country spends. But the be important to retain the features that make composition of its spending differs strikingly alternative basic education highly relevant in from that in other countries. In the first cycle, the first place, particularly in rural contexts:--a the recurrent public spending per pupil is com- flexibleschooldayandacademiccalendar,useof parable to that in Asian and Eastern European teachers from the community, simple classroom countries, but Ethiopia achieves the result by facilities, proximity to children's homes, strong combining a much higher cost of teachers with in-service teacher support, and reliance on com- a substantially less favorable pupil-teacher ratio munity participation. and a significantly smaller share of spending on inputs other than teachers. In Grades 5­8, even · Use teacher recruitment criteria which leave when compared with other African countries, room for improving other aspects of condi- where teacher remuneration as a multiple of per tions in classrooms. capita GDP typically exceeds that in other low- income countries, the composition of spending Teacher recruitment criteria that are set with- inEthiopiareflectsasubstantialtradeoffagainst outadequateattentiontotheirfinancialimplica- pupil-teacher ratios and allocations for inputs tions can produce unintended consequences. If other than teachers, in favor of higher teacher they are set beyond what the country can afford costs. With regard to Ethiopia's low level of (or can mobilize through donor assistance), spending on inputs other than teachers, this schools may meet the requirements, but only can be seen in terms of a scarcity of pedagogical by cutting back in other areas that eventually materials, including textbooks, throughout the undermine the ability to provide a good educa- system, as well as limited provision for teacher tion. The result is a tradeoff by default. That a supervision and other support services for ef- poortradeoffhasoccurredinthepastisreflected fective teaching. in the continuous deterioration in pupil-teacher Because spending on teachers makes up the ratios and pupil-section ratios since 1993­94 bulk of education costs, any strategy for cost- (shown earlier in figure 2). Other telling signs management must consider ways to keep this are a high number of large and over-crowded cost item under control. A straightforward op- Designing the Overall Policy Framework 13 Table 2 High Teacher Cost Leads to Adverse Tradeoff Against Other School Inputs in Ethiopia % of recurrent Public spending Average Pupil-teacher spending on inputs Country/Region per pupil teacher wage ratio other than teacher/ Ethiopia. 2001­02 Grades 1­4 0.10 6.8 74.3 7.8 Grades 5­8 0.18 8.0 48.9 10.3 Regional averages, circa 2000 Francophone Africa (20) 0.13 4.4 51.0 26.8 Anglophone Africa (13) 0.13 4.3 41.6 20.2 South & East Asia (10) 0.09 2.4 36.5 23.4 Latin America (4) 0.16 4.3 33.6 18.5 Eastern Europe &Central Asia (5) 0.11 1.2 18.2 31.3 Numbers in parentheses following each region's name refer to the number of countries on which the regional averages are based. tionistoreducethepaystructure,butinpractice example, suggests that in some localities there thiswouldbepoliticallydifficultandalsolargely may be as many 20 qualified applicants for each irrelevant since the high rates of attrition among available position. Recruiting at lower levels of those with TTC diplomas and university gradu- certification is admittedly not ideal because of ates suggest that they are not overpaid. possibleadverseeffectsonthequalityofservices. A more promising option is to accept greater Yet it may be the only way to manage tradeoffs flexibility in teacher recruitment standards, within a limited budget so as to achieve a bal- for example, by allowing new teachers to be anced and pedagogically effective combination recruited at lower levels of formal certification. of school inputs, characterized by reasonable This does not mean filling teaching posts with student-teacher ratios and adequate availability clearly unqualified candidates, nor does it mean of books and other pedagogical materials. that once recruited the teachers remain stuck in their careers. Rather, the idea is to lower the · Reduce subject specialization so as to make floor for entry while encouraging teachers to better use of teacher time. upgrade their qualifications over time as part of a long-term career plan. As indicated above, Comparison with the practice in other coun- recruitment of para-professional teachers is al- tries suggests that there is probably some scope readyhappeningtosomeextentinsomeregions. for improving the management of teacher time Such teachers may be tenth or twelfth grade inEthiopia.WhileteachersofGrades1­4appear completers without certification from a teacher to have a relatively high teaching load, those in training institute or college, and they typically the higher grades do not. Further exploration of receive an income that is perhaps only half as this is important because differences in teach- high as that of a certified teacher. Even so, re- ing loads make a big difference to the cost of cent processing of recruitment in Oromiya, for service delivery. Our analysis suggests that were 14 EDUCATION ETHIOPIA IN teachers in Grades 5­8 used as fully as they are and translation of findings into implementation in Grades 1­4 this would lower spending by recommendations is a good way to track prog- around one third. ress. With regard to the production of a skilled workforce, it is particularly important to track · Consider theuseofconstructiontechnologies the transition from school to work among the which would lower classroom costs. graduates of courses of study that are especially expensive for the government to provide (e.g. A classroom in Ethiopia can cost up to TVETandhighereducation).Whileasystematic US$19,000. The Ministry of Education recently and routine tracking of this transition does not commissionedtwostudiestolookintotheoptions appear to be in place at present, the gap could for more affordable designs for school buildings. be bridged by exploring collaboration with the One study found that construction costs could Central Statistical Authority to generate the be reduced to between $4,400 to $5,900 per required data through its regular surveys. classroom depending on choice of materials and design. Costs are naturally higher in the · Keep the set of indicators for primary educa- more remote areas, but the second study found tion small and relevant. that even after taking this factor into account, existing construction costs could be reduced by Inthepast,mostcountrieshaveusedmeasures about a third. Across Africa, the benchmark rate such as the gross and net enrollment ratios to is around $8,000 per classroom. The scope for monitor progress in the coverage of primary reducing construction costs in Ethiopia thus ap- schooling, in part because these indicators are pearstobesubstantial.Whilea$6,000classroom easy to compute. However, while appropriate may not be as good as a $20,000 classroom, for some aspects of education planning, they are choosing the latter standard effectively means not adequate for tracking the country's progress that for a given budget less than a third as many toward ensuring that all children have access to new classrooms would be built. a full cycle of primary schooling, particularly in the context of Ethiopia where the system is still catering to a large backlog of overage children. Key Indicators in Measuring Progress Increasingly, countries are instead using the pri- mary school completion rate (typically at Grade In designing the overall policy framework, it 5 or 6) as the relevant measure of success. Be- is not enough simply to define goals and make cause this indicator reflects underlying patterns tradeoffs; it is also critical to track progress to- of access and continuation in school, two other wards the intended goals. As mentioned above, measures, the entry rate to Grade 1 and survival three goals in education have been included in rates to subsequent grades, are also included as theSDPRPcontext:universalprimaryeducation, part of the package. In addition, repetition rates improvement in the quality of services, and the are typically reported as well, for two reasons. production of a skilled workforce. Monitoring First, an education system characterized by high the first goal is described below. With regard to rates of repetition costs more to operate; and the second goal, in 2000 the government com- second, high repetition rates point to possible pleted a national baseline assessment of student problems in managing student flow, which, if achievement in Grades 4 and 8 and has just re- left unattended, are likely to make it harder to peated the exercise in 2004. Regular follow-up improve completion rates. CHAPTER 4 Getting Resources to Schools, Communities and Households Pay Special Attention to the Needs of 1, the share is only 45 percent among rural chil- Rural Populations dren. Once in school, rural children survive to Grade 4 at about 72 percent of the rate among T here are large differences in primary urbanchildren,buttheirsurvivalratetoGrade8 schoolparticipationacrossthreecom- is only 25 percent as high. Overall, these results monlydefinedpopulationgroups:girls suggest that to improve primary schooling out- and boys; children in urban and rural comesinEthiopia,andtoreducesocio-economic areas;andchildrenfromrichandpoor disparities in the higher levels, the needs of rural households (table 3). The gap is widest between localities are paramount. children in urban and rural areas: participation Managers in the education system can influ- for rural children is, on average, less than one ence progress toward the country's goals in third that of their urban counterparts. Whereas basic education through their management of 90percentofurbanchildreneverenrollinGrade the supply- and demand-side factors that affect Table 3 Urban-Rural Disparities in Primary School Participation Are Especially Wide in Ethiopia Index of parity in primary school participation Girls Rural Areas Poorest Quintile Average across all (Boys=1.00) (Urban = 1.0) (Richest = 1.0) three sets of comparisons Ethiopía 0.77 0.31 0.69 0.59 Sub-Saharan Africa 0.83 0.63 0.50 0.65 Note: indices of parity between the groups compared are based on the averages of the following cross-sectional indicators: gross enrollment ratio for Grades 1­6; non-repeaters in Grades 1 and 6 relative to the populations ages 7 and 12, respectively. 15 16 EDUCATION ETHIOPIA IN Table 4 Primary Schools Are Still Too Inaccessible For Many Children in Rural Ethiopia Distance from home to nearest primary school Locality 0 1­2 km. 3­4 km. 5 km. + All Percentage of all sample children Urban 47.4 45.3 6.2 1.1 100 Rural 15.4 28.1 27.8 33.3 100 Total 19.6 30.4 24.9 25.1 100 Percent of children registered for school Urban 85.5 83.1 78.9 71.0 84.0 Rural 43.6 38.8 32.6 19.8 32.4 Total 57.0 47.4 34.1 20.0 41.6 schooling behavior. While the supply- and de- tendance is to be improved, other constraints mand-sideissuesarediscussedseparatelybelow, willalsoneedtobeaddressed.Statisticalanalysis in practice they often interact. suggests that the quality of the learning envi- ronment also matters: the combined variables associated with this aspect of supply explain Act on the Supply of Services a large fraction of the remaining differences in enrollment rates across localities that are not · Improve the availability and accessibility of explained by differences in the characteristics schools in rural areas. of the households. Distance to school is a problem. One third · Deploy teachers across schools more consis- of all rural children live at least 5 kilometers tently in relation to size of enrollments. from the nearest primary school. Statistical analysis indicates that when socio-economic An important action required to improve differences across households are controlled the quality of the learning environment is to for, each additional kilometer from the nearest ensure that teachers are available in adequate primaryschoolreducesregistrationratesbytwo numbers relative to the number of students in to three percentage points, up to distances of the school. Because pupil-teacher ratios are so 12 to 15 kilometers. In other words, reducing high on average in Ethiopia, most schools suffer the distance to school from 5 kilometers to zero fromchronicshortagesofteachers.However,the (i.e. to within the neighborhood) could raise situation is much worse in some schools than registrationratesby10­15percentagepoints--a in others. Among government primary schools substantial increase. enrolling about 1,000 pupils, for example, the However,distanceisnottheonlydeterminant number of teachers ranges from less than five to of school registration: among children in rural more than 30. Across all government primary areas with a school in the immediate neighbor- schools an estimated 28 percent of the variation hood, only 44 percent registered. If school at- in the number of teachers on staff is accounted Getting Resources to Schools, Communities and Households 17 for by factors other than the size of enrollments; that a complete instructional program can be across government secondary schools, the cor- offered. responding share is 30 percent. The relation between number of teachers and number of pupils is even weaker in Grades 1­4. Indeed Address Demand-side Constraints at this level in the system, Ethiopia's degree of randomness in teacher deployment exceeds The demand-side constraints on school partici- that of all other 21 African countries for which pation are many. They include household-level there is information. Making the allocation of factors such as the affordability of schooling teachers more consistent and purposeful across relativetohouseholds'incomes,theopportunity schools is clearly an area that better administra- costofchildren'stime,andthevalueofschooling tive management of the system can and should as perceived by parents; individual-level factors seek to bring about. such as child's personal characteristics such as gender and being an orphan; and community- · Encourage schools progressively to offer level variables such as language. While many of complete instructional programs. thesevariablescannotbealteredintheshortrun (oratall),documentingtheirimpacthelpsinun- Another factor that affects the quality of derstanding the dependence of progress in edu- services is the completeness of the instructional cation on advances in solving problems in other program offered by schools. Where schooling sectors,andinidentifyingpossibleinterventions cycles are incomplete some students drop out and target groups for special attention. sooner than desired or repeat the highest grade offered simply because they have nowhere else · Reduce the direct costs of schooling for chil- to go. In Ethiopia, only a fifth of government dren from poor families. primary schools offer instruction up to Grade 8,whilejustover31percentstopafterGrade6, In Ethiopia schooling is tuition-free but and34percent,afterGrade4.Ruralschoolsare families do incur other fees and expenses. The morelikelythanthoseinurbanareastoofferan direct costs of schooling can be significant: in incomplete instructional program. The legacy 2001, households spent an amount equal to of the previous 6­4­4 structure of education nearly 30 percent of the government' recurrent system is evident in the large share of schools spending on education that year, and slightly that stop at Grade 6. Most of these schools more than half of the total was for primary were created around 1980. Schools that stop at schooling. Across households, the direct costs Grade 4werecreatedonaverage,around1987, of primary schooling weigh heavier on the and most also appear not to have developed budgets of the poorest households than on that beyond the scope of their initial instructional of their richer neighbors, not just because the program. Schools that offer less than the first per child outlay is a larger share of household four grades make up less than eight percent of income, but also because they tend to have Ethiopia'sprimaryschoolbuthave,onaverage, largerfamilies(table4).Itisimportanttoavoid higher pupil-teacher ratios than other schools, increasing the already large burden on poor suggesting that staffing constraints may be one households. Indeed, if private funding for post- factorstandinginthewayoftheirdevelopment. primary education is successfully mobilized, Part of the task of strengthening the adminis- the result should be to free up public funds for trative management of the system is to ensure primary schooling, perhaps even to eliminate that constraints at the school level, whether in the cost that families still face to enroll a child staff or classroom facilities, are addressed so in primary school. 18 EDUCATION ETHIOPIA IN Table 5 The Direct Cost of Primary Schooling Can Be Significant For the Poorest Families Annual education outlay per child in primary school a/ (Birr unless otherwise indicated) Type of school Poorest quartile 2 3 Richest quartile Government schools 21.9 27 36.5 57.2 Non-government school 37.4 56.9 64.1 340.9 Weighted average 22.6 28.1 38.2 80.8 As a percentage of consumption expenditure per adult equivalent 3.3 2.6 2.5 2.4 Memorandum: Annual average consumption expenditure per equivalent adult 688 1,095 1,555 3,366 a/Includes expenditure on fees, Books, school supplies and other education-related items. The estimates are based on data for the 8,112 households in the sample which have children enrolled only in primary education in the indicated type of school. The estimates are underestimated to the extent that spending on evening programs are excluded. · Offer additional help where needed to reduce a child's chances of being enrolled is influenced the opportunity costs of schooling. bythepercentageofhouseholdheadsinthecom- munity who are literate, even after controlling When a child's attendance at school conflicts for such factors as household poverty, distance with participation in work activities--whether from school, and whether or not the child's own at home, on the family farm, or in the labor household head is literate. The fact that the market--families incur an opportunity cost in impact of parental literacy diminishes sharply enrolling a child at school. The data suggest with the inclusion of community-level literacy that more than half of rural Ethiopian boys and rates suggests that beliefs about the importance nearly a third of rural girls work in a range of of schooling may be formed at the community- activitiesthatmightconstitutea"job".Children level, and that exposing communities to the who work are generally less likely to attend potential benefits of schooling may be a very school, although it appears that not all work is useful approach in the short run for enhancing incompatible with schooling. school participation. · Helpparentsandcommunitiesbetterappreci- · Removebarrierstoschoolparticipationposed ate the value of schooling. by gender and being orphaned. In a country where schooling rates are low, a Gender gaps are primarily a rural phenome- barrier to the enrollment of children may sim- non, and appear to be region-specific, implying ply be that parents themselves have not been to that efforts to increase school participation by schools and have no idea what schools can do girls will require a region-by-region approach. for their children. Statistical analysis shows that For orphans, statistical analysis shows that af- Getting Resources to Schools, Communities and Households 19 ter controlling for a variety of household, com- orphans, and that the effect is twice as great munity and regional characteristics, children for children who have lost both parents. Again, who have lost one parent are 5­6 percentage specific programs targeted as such children are points less likely to attend school than non- required. 20 EDUCATION ETHIOPIA IN CHAPTER 5 Transforming Resources in Schools Into Educational Outcomes O nce children are in school it is the The results of the 2004 assessment are not yet responsibility of teachers and school available. directors to ensure that they not only A key challenge for pedagogical manage- remain there but that they learn. Suc- ment is to improve upon these baseline scores. cessful delivery of education services Finding the right interventions will not be easy. must be reflected in progressive improvements The national baseline assessment evaluated in the level of student achievement. the importance of such factors as school infra- structure;schoolorganizationandmanagement practices; teacher characteristics and practices; Make it a Key Goal to Raise Student and the availability of instructional materials Achievement and equipment. In both Grades 4 and 8, the study confirmed the importance of textbooks In the 2000 assessment of learning achievement and the use of radio for supplemental instruc- of Grade 4 and Grade 8 students, about 10,500 tion, findings that are consistent with those fourth graders were tested in Reading (in the reportedinmanyothersimilarstudies.InGrade language of instruction), English, Mathematics 8, schools differ in the share of teachers who and Environmental Science; and some 5,500 graduated from a teacher training college or eighth graders were tested in English, Math- higher levels of training but the study reported ematics, Chemistry and Biology. The average that this variable had no impact on student percentageofcorrectanswersforallthesubjects achievement. By contrast, it did find that vari- combinedwas48percentintheGrade4sample; ables that reflect attitudes and behaviors were and 41 percent in the Grade 8 sample. Given important, including how well teachers prepare that the test items were chosen from a range of their lessons; how much the director focuses on key topics in the curriculum for the grade tested schoolmatters;andtheextenttowhichstudents as well as that for the previous grade, these engage in homework. scores indicate that a large number of students were not achieving the curriculum objectives. 21 22 EDUCATION ETHIOPIA IN Taking Advantage of Decentralization · decentralization of education to sub-national to Get Results governments does not in itself empower par- ents or improve school performance. Further Ethiopia is currently undergoing a far-reaching decentralizationtoschools(schoolcouncilsor decentralization of the structure of government. school boards) or local communities is more Budgets are allocated to the regions in block likely to empower parents and can improve grants, and the regions in turn are expected to school performance. make block grants to the woreda governments. · for decentralization to schools to be suc- In the new governance landscape, these sub- cessful, principals must acquire new skills in national governments are directly responsible motivating teachers and the community, and fordeliveringprimaryandsecondaryeducation. in financial management. For them, two of the issues discussed above are · the specific design of financial transfers to immediately relevant: (a) the mis-allocation of sub-national governments or schools can teachers; and (b) the encouragement of those have powerful effects on both efficiency and inputs and attitudes which boost pupils' per- equity. formance. Making tradeoffs among alternative · failure to restructure national and/or regional goals and means will become increasingly rel- ministries of education is a serious obstacle evant as authority, finances and responsibilities to realizing the benefits of decentralization. are transferred to these lower level authorities. · the decentralization of teacher management Two broad recommendations are offered. is critical to creating accountability and real- izing the potential benefits of decentraliza- · Learn from International and Domestic tion. Experience Some of the foregoing ideas are already being Eventhoughtheprocessofdecentralizationis madeconcreteinEthiopia.In2002,theMinistry still at an early stage, Ethiopia might take note of Education issued a set of "Guidelines for Or- ofsomeofthelessonsfortheeducationalsystem ganization of Educational Management, Com- emergingfromtheexperienceincountrieswhich munity Participation and Educational Finance" are further along in the process. Experiences in toclarifyrolesandresponsibilitiesatalllevelsof Latin America and Eastern Europe suggest: management in the education system, as well as to give explicit recognition to the importance of · efficiency and effectiveness are most likely to community participation and parental involve- improve under decentralization when service ment in school-based management. Identifying providers--schools, local governments, or the most critical elements in these guidelines, regional governments--are held accountable puttingthemintopracticeandevaluatingtheim- for results. pacts of alternative arrangements are important · accountability requires clear statements of tasksthatwarrantattentionintheyearstocome. authority and responsibility and transparent If carried out systematically and routinely, they and understandable information on results will provide lessons from Ethiopia's own imple- (both educational and financial). mentation of decentralization in the sector. · decentralization of real decision making power to schools or school councils can sig- · Mobilize and Disseminate Information to nificantly increase parental participation in Enhance Accountability for Results. the school, and high levels of parental and community participation are associated with As education is decentralized in Ethiopia, the improved school performance. challenges of ensuring accountability for results Transforming Resources in Schools Into Educational Outcomes 23 will be increasingly important. The task is dif- relevant sub-national levels. Report cards for ficult but can be reduced by creating the right school districts and other higher levels of ag- kinds of information flow. In the first instance, gregation can also provide similar comparative theavailabilityofappropriateinformationatthe information. The availability of such informa- levelofschoolscanhelpstimulateacommunity's tion may be expected to improve conditions in interest in ensuring that its schools get a fair schools in two ways: by creating incentives for share of the public resources for education. In under-endowed schools to seek redress; and by Uganda, for example, a large-scale public dis- pinpointing specific schools in the system that semination of information on grants for school- warrant the attention of those in charge. ing gave schools and parents the information Finally, a greater transparency at the local they needed to monitor the grants. Over the government level in the formulation and imple- space of just six years, the leakage of funds fell mentation of budgets, could have a positive from 80 to 20 percent, and schools were able to impact. This would make visible aspects such usetheresourcesnowreachingthemtopurchase as the balance between salary and non-salary textbooks and other pedagogical inputs. expenditures and the utilization rate of the Reportcardsareanotherinterestingexample. capital allocation. In addition there is a great They provide schools with information that need to ensure timely payments and reports compares indicators for the school (e.g. pupil- and to put in place a viable audit system; and teacher ratios, examination results, and parent to encourage a more bottom-up establishment feedback) with averages at the national and of needs. 24 EDUCATION ETHIOPIA IN Conclusion A decade after the launch in 1994 of the of pedagogical conditions in schools across the New Education and Training Policy, country since 1993­94 which has accompanied Ethiopia's government can look back thehighratesofexpansionsoundsanalarmthat with justifiable pride on the progress bears heeding. It is a signal that the existing ar- achieved.Theveryrapidgrowthofprimaryedu- rangementsforservicedelivery--characterizedin cation reflects a genuine commitment to trans- particular by the criteria for teacher recruitment form the country's historically elitist system into and classroom construction--may not be suited onethatservesallofEthiopia'schildren.Andthe for expansion on the scale required to attain the growth of secondary and higher education will goal of universal primary school completion. provide the potential manpower for all sectors The work of finding new approaches will oftheeconomy.Thegovernment'srecentpoverty need to be informed by a hard-nosed, system- reduction strategy paper (SDPRP) reaffirms the wide assessment of potential tradeoffs among goal of universalizing primary education while the options for moving forward. Better ad- also envisioning improvements in quality and ministrative management is also critical for expansion of other levels to enhance the skills continued progress, so that the available human profile of the workforce. These ambitions pres- and financial resources for education are chan- ent sobering challenges on many fronts. neled effectively to schools, communities and Accomplishing them will likely require more households. Transforming these resources into public spending on education than the sector higher levels of school participation and even currently receives, as well as increased spending moreimportantly,intolearningoutcomes, is the on education by households in post-primary ultimatemeasureofsuccess.Thistaskisdifficult education. Money alone is unlikely to produce in the best of circumstances, but it can be made the desired results, however. In primary and less so by cultivating a culture of accountability secondary education the system is truly at a for results and good stewardship of resources cross-roads today. The continuous deterioration throughout the system. 25 26 EDUCATION ETHIOPIA IN Annex Tables and Figures Table A1. Enrollments by Level in Government and Nongovernment Institutions, Ethiopia, 1967­68 to 2001­02 Technical and Higher Vocational Teacher education School Preprimary Secondary Education Training (Regular Year (Grades 1­8) Primary (Grades 9­12) and Training Institutes programs only)a/ 1967­68 -- 496,334 26,690 -- -- -- 1970­71 -- 728,548 53,220 -- -- -- 1975­76 -- 1,226,124 90,091 -- -- -- 1980­81 -- 2,341,437 216,876 -- -- -- 1985­86 -- 2,811,910 292,385 -- -- 18,457 1990­91 -- 2,871,325 453,985 -- -- 17,895 1995­96 -- 3,787,919 402,753 2,738 5,900 17,378 2000­01 109,358 7,274,121 649,221 8,639 6,224 46,812 2001­02 118,986 7,982,760 684,630 38,176 6,080 48,143 2002­03 123,057 8,572,315 626,714 54,026b/ 7,002 77,946 2003­04c/ -- 9,343,428 725,059 -- -- 80,698 -- Data not available at time of report writing. Note: All data in this table refer to enrollments in regular programs only. a. Includes students in diploma-level teacher training colleges, but excludes part-time students and those in evening or summer (Kiremt) courses. b. Data probably incomplete because of missing information from several regions. c. Preliminary data available as of September 2004. 27 28 EDUCATION ETHIOPIA IN 0 0 0 0 0 0 n.a. n.a. ~500 n.a. Adult 100~ } education ~500,000 a/ 26 15 -- 0 Higher 48,143 12,619 44,567 6,472 9,119 102 47­57 50­56 education } cycle 2 3 0 0 9­12) 766 684,630 15,667 80,011 2,695 1,080 89­93 9­13 } Secondary (grades 5 37 0 0 92 6 12 otalT 8,644 7,982,760 369,315 161,577 59,951 ~500,000 cation. 5­8 6 35 0 0 0 97 9 9 certifi 2001­02 cycle 140,050 73,714 26,145 2,331 Grades 2,256,806 Primary Ethiopia, diploma-levela 1­4 4 38 0 0 91 5 13 seeking 6,314 Education, 5,725,954 229,265 87,863 33,806 of Grades 500,000~ teachers Level trainee and c/ for regular training. ypeT students financed applicable. (%) (%) b/ in not not mainly by establishments establishments teacher share share education education n..a. courses students students fee-paying students courses e/ Students distance Kiremt nonformal (thousands) : of of of d/ available; of student daytime students in in in classes diploma-level summer A2. to of nongovernment nongovernment total government data ableT No Number ypeT Total In Nongovernment Total In Nongovernment Percentage daytime Percentage Percentage by Includes Refers Regular Evening Students Students Students Grand Memoranda -- a. b. Annex Tables and Figures 29 to the in to and, to years figure ever students. transfers first Kiremt four fail the allows correspond the daytime of even to For that evening,in or regular instead longer as education programs. threein take curriculuma of adult enrollments studies fact in types studies follows under their may various converting their group grouped the after complete are in to programs complete younger added long group are as The programs older enrollments they times well-organized older. and thein up 1.5 2001­02. figure, less 15 adding take the and in of aged second well-organized Participants the programs thein those 1997­98 for somein 1­4. and assumptions such in pupils pupils 7­14 gradesin between programs; while and data of aged students cycle on alternative to types that because country, those based all the primary groups, the correspond across assumption attempted across two estimate into under the notis uniform here roughais addition on education divided simple notis total higher by calculation program. are equivalents classified The and similara formal computed arrangement the daytime education thereforeis courses. are the with the secondary and of for regular education, students. to schools, nonformal in basic system content shown enrollments equivalency courses primary fee-paying and formal ranges range, grade nonformal formal Participants the intent The each distance For in reach Includes c. d. e. 30 EDUCATION ETHIOPIA IN Table A3 Primary Gross Enrollment Ratios (GER), Ethiopia, 1993­94 to 2001­02 Based on household surveys Based on school census and GER Survey and data School year government population estimates a/ Estimate collection period b/ 1993­94 20.5 -- 1994­95 26.2 -- 1995­96 30.1 34.0 WMS, Jan./Feb. 1996 1996­97 34.7 -- 1997­98 41.6 45.8 WMS, March/April 1998 1998­99 45.8 41.3 LFS, May 1999 1999­00 51.0 53.9 WMS, Jan./Feb. 2000 51.6 DHS, Feb./May 2000 2000­01 57.4 -- 2001­02 61.6 -- -- Estimate not available in the absence of a survey Note: The gross enrollment ratio (GER) in this table refers to grades 1­8. a. For 2000­01 and 2001­02, the GER includes evening students in the numerator (which was augmented as a result by 1.2 per- cent); the increase would have been much larger if children in nonformal basic child education programs were also included. In this case, the GER would have risen to 61.2 percent in 2000­01 and 65.6 percent in 2001­02, assuming that the share of students in regular day programs is 92 percent. b. The household surveys are the following: WMS, Welfare Monitoring Survey; LFS, Labor Force Survey; DHS, Demographic and Health Survey. The Child Labor Force Survey conducted in March 2001 could have extended the series based on household surveys, but the data files were unavailable at the time of data analysis for this study. Annex Tables and Figures 31 Table A4. Three Indicators of Entry to Grade 1, Ethiopia, 1993­94 to 2000­01 Percentage of Apparent intake Cohort entry rate Ratio of AIR children ever Year rate (AIR, %)a/ (CER, %) b/ to CER enrolled by age 7 1993­94 45.2 -- -- -- 1994­95 55.8 -- -- -- 1995­96 71.3 37.2 1.9 12.6 1996­97 80.6 -- -- -- 1997­98 90.5 44.6 2.0 17.3 1998­99 92.9 54.5 1.7 21.9 1999­00 95.7 52.4 1.8 20.6 2000­01 99.5 60.5 1.6 -- -- Data not available. a. Definedasthenumberofnewentrantstograde1,regardlessofage,relativetothepopulationof7-year-olds;dataexcludeevening students. b. Refers to the estimated share in a cohort who ever entered school. The rates are estimated from data for sampled youths ages 7­18 in the Welfare Monitoring Surveys of 1996, 1998, and 2000; youths ages 7­14 in the 1999 Labor Force Survey; and youths ages 7­17 in the 2001 Child Labor Force Survey. Because the percentage ever enrolled tends to peak by ages 13­14, it provides a reasonably good estimate of the share of children in a cohort who would ever enroll. For 1995­96 and 1997­98, the peak occurs slightly later, so the data shown in the table correspond to the percentage ever enrolled by ages 15­16 and 14­15, respectively. 32 EDUCATION ETHIOPIA IN Table A5. Composite Cohort Survival Rates and Intercycle Transition Rates, Ethiopia, 1993­94 and circa 2000­01 Indicator 1993­94 1999­00 2000­01 2001­02 Cohort survival rates from grade 1 To grade 4 0.52 0.60 0.61 0.59 To grade 5 0.49 0.53 0.54 0.51 To grade 6 0.48 0.47 0.49 0.45 To grade 8 0.39 0.36 0.40 0.36 To grade 10 0.23 0.20 0.31 0.27 To grade 12 0.15 0.19 0.13 0.08 Intercycle cohort transition rates grades 4 & 5 0.94 0.88 0.90 0.88 grades 8 & 9 0.84 0.89 0.98 0.91 grades 10 & 11 0.86 -- a/ 0.34 0.33 Notes: a. Survival rates and transition rates are computed from data for two adjacent school years, the first of which is indicated in each column in the table. Shaded cells draw attention to the impact of policy changes implemented in 2000­01 to regulate student flow in grades 11 and 12. b. Estimate exceeds unity, which is by definition inadmissible; it signals probable inaccuracies in the counts of repeaters and re- admitted students as a new policy to regulate student flow in preparatory secondary education was implemented from 2001­02 on. Annex Tables and Figures 33 Table A6. Efficiency of Student Flow in Primary Education, Ethiopia, 1993­94 and 2001­02 1993­94 2001­02 Number left Number left from an initial Repetition Pupil-years from an initial Repetition Pupil-years Grades cohort of 1,000 ratea/ invested cohort of 1,000 ratea/ invested 1 1,000 0.139 1,162 1,000 0.191 1,235 2 580 0.075 627 746 0.141 867 3 539 0.055 570 652 0.132 751 4 518 0.086 567 587 0.161 699 5 489 0.064 522 515 0.159 612 6 483 0.137 559 450 0.115 508 7 413 0.147 484 437 0.226 564 8 385 0.243 508 363 0.221 467 Cumulative pupil years 4,406 4,999 4,749 5,704 Index of efficiencyb/ Dropout-related 0.70 0.61 Repetition-related 0.88 0.83 Overallc/ 0.62 0.51 (0.71) (0.61) a. Defined as the number of repeaters in grade X in year Y+1 relative to the number of students in grade X in year Y. b. The index is defined as the ratio between the cumulative pupil-years invested in a system with neither dropping out nor grade repetition, and the actual cumulative pupil-years invested given the pattern of dropping out and repetition. Thus, the closer is the index to 1.0, the more efficient is the student flow. c. Figures in parentheses refer to the index associated with student flow patterns in grades 1­5. 34 EDUCATION ETHIOPIA IN Table A7. Trends in Public Spending on Education, Ethiopia, 1980­81 to 2001­02 Total recurrent and capital spending Recurrent spendinga/ As percentage As As As of total percentage percentage percentage government As of total of government As of total current spending Gregorian percentage government revenues net percentage spending on net of debt E.C. Calendar of GDP spending of grants GDP education interest payment 1973 1980/81 2.1 10.0 13.0 -- -- -- 1974 1981/82 2.4 9.7 14.4 -- -- -- 1975 1982/83 2.2 7.3 12.7 -- -- -- 1976 1983/84 2.5 8.8 12.9 -- -- -- 1977 1984/85 2.4 8.4 14.1 -- -- -- 1978 1985/86 2.4 8.5 12.3 -- -- -- 1979 1986/86 2.4 9.3 12.8 -- -- -- 1980 1987/88 2.5 8.2 11.6 -- -- -- 1981 1988/89 2.9 8.5 12.4 -- -- -- 1982 1989/90 2.8 9.4 15.8 -- -- -- 1983 1990/91 2.5 10.1 18.1 -- -- -- 1984 1991/92 2.5 12.6 23.9 -- -- -- 1985 1992/93 2.8 14.3 23.4 1986 1993/94 4.0 15.8 28.5 3.0 75.6 24.6 1987 1994/95 3.5 14.3 20.3 2.7 76.1 20.9 1988 1995/96 3.8 15.7 20.6 2.6 68.2 21.0 1989 1996/97 3.6 15.3 20.1 2.6 71.7 22.5 1990 1997/98 3.5 14.1 19.5 2.6 74.2 18.8 1991 1998/99 3.5 11.5 19.9 2.6 73.6 13.8 1992 1999/00 3.1 9.5 17.3 2.6 82.0 10.6 1993 2000/01 4.1 14.2 21.5 3.0 71.5 16.9 1994 2001/02 4.0 13.6 21.0 3.0 75.9 -- Data not available. a.Excludesspendingonitemssuchasvehicles,equipment,building,andconstructionmaterialswhichwereclassifiedinthebudget data as recurrent spending. Annex Tables and Figures 35 E.C. 373 139 140 128 139 461 330 143 178 10.6 63.1 54.4 8.6 5.1 1.2 1.1 17.3 1.7 100 1994 2001­02 1221 1,799.3 1,696.8 TTC E.C. 309 121 122 113 312 102 380 244 105 143 10.8 68.0 58.5 9.4 1.6 1.1 1.2 15.8 1.6 100 Institute; 1993 2000­01 1,563.5 1,370.5 Training E.C. 268 98 103 75 189 111 310 176 73 115 11.7 69.0 61.2 7.8 1.2 1.5 1.2 14.2 1.3 100 1992 1999­00 1,343.2 1,102.2 Teacher TTI E.C. 266 99 103 76 204 111 329 139 64 112 11.9 71.2 63.1 8.1 1.3 1.5 1.3 11.5 1.2 100 Training; 1991 1998­99 1,265.8 1,070.6 and 2001­02 to E.C. 215 97 100 75 187 86 239 137 60 107 10.2 73.2 64.8 8.4 1.3 1.2 1.0 11.9 1.2 100 Education 1993­94 1990 1997­98 1,174.8 1,018.3 E.C. Ethiopia,, 203 101 105 81 105 103 175 128 55 108 9.4 75.7 66.7 9.0 0.7 1.4 0.7 11.0 1.1 100 1989 1996­97 1,084.1 1,035.5 Technical/Vocational TVET Subsector E.C. 140 100 101 98 99 93 92 99 71 102 6.9 by 80.0 68.5 11.5 0.7 1.4 0.4 9.1 1.5 100 1988 1995­96 979.4 970.2 years. E.C. Calendar 100 95 96 95 117 88 96 97 83 96 5.3 80.7 68.9 11.9 0.9 1.4 0.4 9.4 1.8 100 Composition 1987 1994­95 913.1 913.1 Birr. and Gregorian E.C. row, constant rendsT 100 100 100 100 100 100 100 100 100 100 5.080.9 68.9 12.0 0.7 1.5 0.4 9.3 2.1 100 1986 1993­94 847.6 955.4 second 1994­95 in Birr years, Spending a/ : amounts 100)= Birr 1994­95 Calendar on Education spending spending based (1993­94 of total current constant Ethiopian College. are Public on of of trends A8. secondary& 1­8 9­12 share education secondary& 1­8 9­12 shows Training trends education millions millions row : ableT Grades Grades Grades Grades In In First Teacher Spending Recurrent Item Spending Administration Primary TVET TTI TTC Higher Other Total Percentage Administration Primary TVET TTI TTC Higher Other Total Memoranda Note a. b. 36 EDUCATION ETHIOPIA IN Table A9. Regional Recurrent Spending on Primary, Secondary, and Technical and Vocational Education and Training, Ethiopia, 2001­02 Spending in 2001­02 as a ratio Percentage share of regional of spending in 1993­94a/ recurrent spending on educationb/ Region Overall Grades 1­8 Grades 9­12 Grades 1­8 Grades 9­12 TVETc/ Tigray 2.07 2.07 2.07 75.7 9.9 2.1 Afar 1.35 0.75 3.07 44.0 34.1 0.0 Amhara 1.43 1.29 0.92 75.1 7.0 1.7 Oromiya 1.74 1.45 1.70 68.5 11.9 7.9 Somali 1.14 1.00 0.80 71.8 7.9 2.9 Benshangul-Gumuz 1.71 1.48 1.58 71.6 7.9 0.0 SNNPR 1.69 1.52 1.01 74.5 6.3 2.3 Gambilla 2.53 1.71 5.26 50.3 7.9 1.2 Harare 1.80 1.64 2.80 63.8 20.7 0.0 Addis Ababa 1.62 1.09 0.73 38.5 15.7 22.8 Dire Dawa 2.11 1.42 2.41 46.0 22.1 11.6 Average 1.74 1.40 2.03 61.8 13.8 4.8 SNNPR, Southern Nations and Nationalities Peoples' Republic. Notes: a. Amounts in both the denominator and numerator denominated in 1994­95 constant Birr. b. Columnsdonotaddto100percentbecauseoftheomissionofspendingonadministration(shownintheprevioustable)andother items (e.g., teacher training institutes and colleges in some regions and special education). c. Refers to Technical and Vocational Education and Training; increase since 1993­94 not shown because spending was 0 in that year for most regions. Annex Tables and Figures 37 Table A10. Teacher Wages and Educational Supplies as a Percentage of Recurrent Spending in Primary and Secondary Education across Regions. Ethiopia. 2001­02 Teacher wages as a percentage of Educational supplies as a percentage subsector recurrent spending of subsector recurrent spendinga/ Region Grades 1­4 Grades 5­8 Grades 9­12 Grades 1­8 Grades 9­12 Tigray 89.8 82.2 66.3 1.2 6.9 Afar 71.9 65.5 13.4 8.5 5.2 Amhara 96.4 94.8 82.7 0.8 1.6 Oromiya 90.4 89.3 69.2 4.1 15.2 Somalib/ 84.5 87.4 73.5 2.9 0.0 SNNPR 97.7 93.8 73.1 0.6 1.7 Harari 73.2 82.1 64.0 1.4 5.5 Addis Ababa 75.2 76.7 78.8 4.7 5.1 Dire Dawa 80.5 91.4 73.3 8.8 7.4 Averagesc/ Unweighted 84.4 84.5 65.1 3.8 6.1 Weighted 92.5 89.8 71.0 2.5 8.9 SNNPR Southern Nations and Nationalities Peoples Republic. Notes: a. The percentage share of spending on teachers and educational supplies does not add to 100 percent, as the residual category of spending on administrative staff, foodstuff, uniforms, bedding, and so on have been omitted from the table. b. For Addis Ababa, the source data are not broken down by the subcycles shown; for our purpose, we estimate the desired breakdown by prorating spending according to shares of enrollments in the relevant subcycles. c. Estimate based on assumption that teacher remunerations are the same as in Afar, another region where teacher pay is aug- mented by a 30 percent increase as hardship allowance. d. Excludes data for Somali, Benshangul-Gumuz and Gambella , and spending by the federal government. 38 EDUCATION ETHIOPIA IN be 1­8 may 1.0= 1.0 0.8 1.5 2.7 23.4 32.8 58.8 89.1 spending pupil Grades spending of unit of primary multiple per 1­4 As 1.0= 1.2 1.0 1.9 3.3 29.0 40.6 72.8 110.4 million. estimate student Grades 67.3 of the per that and b/ population data, Spending per GDP multiple of 0.12 0.10 0.18 0.32 2.85 3.98 7.13 10.82 As projected budget capita the by MOFED studies. Institutions, in in Birr 2001­02 107 86 160 283 in spending. reported 2,499 3,495 6,267 9,504 Birr unit postgraduate Amount of in Government current amount in billion those the 59.11 to of of in decomposition excludes refers Education GDP of 7368038 655,664 36,462 6,080 3,618 the 27,924 for students Number students ypeT government institutions 5,311,949 2,056,089 of levels. estimated later subsector the this and on number a/ regional presented Level of and dividing enrollments. by Birr) by data spending of recurrent spending 785.8 457.1 328.7 185.6 91.1 21.2 22.7 education; 265.4 with that counts otalT (millions federal higher Student public current at calculated reader of Per d/ was consistency the incomplete overheads Birr alert institutions c/ training& 878.3 ensure to possible 19 Spending c/ 9­12) of to italics of for 1­8) institutes colleges e/ administrative GDP in Somali 2001­02 (grades education capita shown because estimates A11. Recurrent (grades 1­4 5­8 training training education exclude to : per exclude are ableT Education Data The Data Data unreliable Refers Public Ethiopia, Level/type of Primary Grades Grades Secondary Tech./voc. Teacher Teacher Higher Notes a. b. c. d. e. Annex Tables and Figures 39 Table A12. Public Recurrent Spending Per Primary Pupil, Teacher Wages, Pupil­Teacher Ratios, and Spending on Inputs Other Than Teachers, Ethiopia and Selected World Regions, circa 2000 Public Average Pupil­ Percentage of Number of spending teacher teacher recurrent spending on Country/region countries per pupila/ wagea/ ratiob/ inputs other than teachersc/ Etiopía, 2001­02 Grades 1­4 1 0.10 6.8 75.2 7.8 Grades 5­8 1 0.18 8.0 48.8 10.3 Regional averages, circa 2000 Francophone Africa 20 0.13 4.4 51.0 26.8 Anglophone Africa 13 0.13 4.3 41.6 20.2 South & East Asia 10 0.09 2.4 36.5 23.4 Latin America 4 0.16 4.3 33.6 18.5 Eastern Europe & Central Asia 5 0.11 1.2 18.2 31.3 Notes: a. As a multiple of the per capita GDP. b. Weighted average for government schools, excluding data for Somali, Benshangul-Gumuz, and Gambella. c. Includes spending on administrative personnel at the school level, pedagogical supplies, and unspecified nonsalary items for Ethiopia. Data exclude administrative spending at the federal and regional level. 40 EDUCATION ETHIOPIA IN Table A13. Gross Enrollment Ratios by Region, Ethiopia, 1993­94 and 2001­02 Primary education (grades 1­8) Secondary education (grades 9­12) Index Index Country/region 1993­94 2001­02 (1993­94=1.0) 1993­94 2001­02 (1993­94=1.0) Ethiopia 24.1 61.6 2.56 6.9 13.1 1.90 By region Tigray 43.7 77.6 1.78 4.4 24.8 5.64 Afar 5.0 12.6 2.52 1.2 3.2 2.67 Amhara 17.9 58.1 3.25 5.2 9.1 1.75 Oromiya 21.2 62.4 2.94 5.6 11.6 2.07 Somali 6.5 13.1 2.02 0.4 1.4 3.50 Benshangul 35.4 89.1 2.52 2.4 12.2 5.08 SNNPR 28.8 67.5 2.34 5.1 10.7 2.10 Gambilla 53.9 102.7 1.91 5.3 12.4 2.34 Harare 53.4 107.5 2.01 31.6 50.3 1.59 Addis Ababa 84.9 128.4 1.51 40.8 57.3 1.40 Dire Dawa 41.0 80.2 1.96 19.4 28.4 1.46 SNNPR Southern Nations and Nationalities Peoples Republic. Annex Tables and Figures 41 Table A14. Gross Enrollment Ratios by Locality, Gender, and Wealth, Ethiopia, 1999­2000 Secondary Primary education education Higher Population group Grades 1­4 Grades 5­8 Grades 1­8 (grades 9­12)a/ educationb/ By gender Boys 89.1 53.6 72.6 15.6 2.6 Girls 72.6 43.8 59.3 10.5 0.6 Index (boys = 1.0)c/ 0.81 0.82 0.82 0.67 0.21 By locality Urban 122.9 101.1 111.6 76.3 3.7 Rural 65.3 22.0 46.2 0.4 0.02 Index (urban = 1.0)c/ 0.53 0.22 0.41 0.00 0.00 By consumption quintiled/ Richest 112.3 81.2 95.8 53.7 4.8 2 91.0 68.8 80.6 31.1 1.0 3 89.5 50.9 71.3 26.6 1.0 4 80.0 45.0 64.3 21.6 0.4 Poorest 72.9 42.2 59.4 18.9 0.3 Index (richest 20% = 1.0)c/ 0.65 0.52 0.62 0.35 0.05 Notes: a. Ratios in the first two blocks are shown in italics to alert the reader that the data are based on data from the school census and population projections rather than on data from the survey. b. The denominator for this calculation is the population ages 19­23. c. Theindexiscomputedbydividingtheratiofortheless-favoredgroupineachblock(i.e.,girls,ruralchildren,andchildrenfromthe poorest quintile), by the corresponding ratio for the more-favored group (i.e., boys, urban residents, and children from the richest quintile). d. Households are ranked according to the per adult equivalent expenditure on food and other nonfood items. 42 EDUCATION ETHIOPIA IN Table A15. Female Share of Primary and Secondary Enrollments across Regions, Ethiopia, 2001­02 Percent female Index (Ethiopia=1.0) Country/region Grades 1­8 Grades 9­12 Grades 1­8 Grades 9­12 Ethiopía 40.9 39.3 1.00 1.00 By Region Tigray 48.2 36.8 1.18 0.94 Afar 36.9 39.3 0.90 1.00 Amhara 45.8 42.1 1.12 1.07 Oromiya 36.9 35.6 0.90 0.91 Somali 32.8 23.6 0.80 0.60 Benshangul-Gumuz 36.3 30.8 0.89 0.78 SNNPR 37.4 31.7 0.91 0.81 Gambella 36.3 23.7 0.89 0.60 Harari 41.0 42.7 1.00 1.09 Addis Ababa 52.7 50.2 1.29 1.28 Dire Dawa 43.5 43.1 1.06 1.10 SNNPR Southern Nations and Nationalities Peoples Republic. Note: Includes enrollments in evening programs. Table A16. Distribution of Primary and Secondary Students and Schools by Sector, Ethiopia 2001­02 Percentage share by sector Level of education/item Number Government Nongovernment Primary education Schools 11,754 94.9 5.1 Students 7,876,188 95.3 4.7 Secondary education Schools 436 87.6 12.4 Students 679,377 97.7 2.3 Note: The total number of schools differs from the 12,438 reported in the Ministry of Education's Statistical Abstract for 2001­02 (i.e., 1994 E.C.) for several reasons: (i) schools are counted more than once in this table if they offer instruction at both the primary and secondary levels; (ii) schools from Somali are excluded for lack of data in the school census; (iii) schools identified as offering technical and vocational education and training in the census are excluded; and (iv) schools reporting no teachers or no students are excluded. Annex Tables and Figures 43 Table A17. Prevalence of Double Shifting and Section and School Sizes of Government and Nongovernment Primary and Secondary Schools, Ethiopia, 2001­02 Primary schools Secondary schools Indicator Government Nongovernment Government Nongovernment Official instructional time (hours per week)a/ Single shift 1,100 -- 1,100 -- Double shift 840 -- 840 -- Percentage of schools operating two or more shiftsb/ 44.3 19.3 78.6 16.7 Number of students per sectionc/ 72.1 60.1 80.4 51.1 Grades 1­4 75.9 61.9 n.a. n.a. Grades 5­8 60.7 54.8 n.a. n.a. Number of sections per schoolc/ 9.0 9.9 21.0 6.0 Number of students per schoolc/ 673 611 1,741 290 Memorandum: Number of schools in sample 11,047 596 378 54 -- Data not available. n.a. Not applicable. Notes: a. Computed as follows: average of 7 daily periods of 45 minutes each in single shift schools; and 6 daily periods of 40 minutes in double shift schools, in a school year of 210 days. b. Each shift of students has its own set of teachers, but administrative staff, including the school head, is shared by the various shifts. c. Refers to averages per school or section. 44 EDUCATION ETHIOPIA IN Table A18. Distribution of Teachers by Gender and Qualification in Government and NonGovernment Primary and Secondary Schools, Ethiopia, 2001­02 Percentage distribution of teachers according to their academic qualification Teaching assignment of Number Percentage teacher and of of female Grades 8 Grades 9, TTC & BA & type of school teachers teachers and below 10,11, & 12 TTI diplomaa/ MA Otherb/ Total Grades 1­4 Government 74,408 40.4 0.4 2.7 91.6 4.3 0.0 1.0 100.0 Nongovernment 4,756 39.6 1.0 13.7 60.8 17.9 0.8 5.8 100.0 Grades 5­8 Government 43,710 15.7 0.2 0.6 70.3 26.4 0.1 2.3 100.0 Nongovernment 4,008 15.5 0.1 3.4 31.7 54.3 2.7 7.8 100.0 Grades 1­8 Government 118,118 31.3 0.3 1.9 83.7 12.5 0.1 1.5 100.0 Nongovernment 8,764 28.6 0.6 9.0 47.5 34.6 1.7 6.7 100.0 Grades 9­12 Government 13,306 7.6 0.0 0.1 6.9 59.5 32.8 0.7 100.0 Nongovernment 785 11.0 0.0 0.0 1.8 46.9 49.6 1.8 100.0 BA Bachelor's degree; MA master's degree; TTI teacher training institutes; TTC teacher training colleges. Notes: a. Referstodiploma-levelcoursesfromotherinstitutionsofhighereducation;includesteacherswithgrade12andsummertraining or one to three years of unspecified training beyond grade 12. b. Includes teachers qualified to teach physical education, technical and vocational courses, crafts, agriculture, music, art, or unspecified subjects. Annex Tables and Figures 45 Table A19. Average Weekly Teaching Loads by Teacher Certification and Grade in Which Teacher is Teaching, Oromiya, Ethiopia, 2002­03 (in hours) Teacher Certification Locality & grade to which teacher TTC and is assigned 10+1; 10+2 12; 12+1 TTI equivalent All teachers Oromiya 17.7 21.0 24.2 18.1 23.8 Grades 1­4 18.0 27.4 26.0 22.2 26.0 Grades 5­8 17.5 16.2 21.2 18.0 20.6 Urban schools 17.5 16.6 19.9 17.0 19.5 Grades 1­4 18.1 21.6 21.6 18.1 21.5 Grades 5­8 17.1 15.8 17.9 17.0 17.6 Rural schools 23.0 27.7 27.3 22.3 27.2 Grades 1­4 15.0 28.9 28.6 28.9 28.6 Grades 5­8 27.0 20.3 24.5 21.9 24.3 Note: Refers to government teachers only. 46 EDUCATION ETHIOPIA IN Table A20. Incidence of Poverty by Gender and Education of Household Head, Ethiopia, 1999 Characteristic of household head Percentage living in poverty a/ Male 0.38 Female 0.37 Completed some primary education 0.32 Completed secondary 0.21 Completed postsecondary 0.09 Note: Refers to share of households whose consumption expenditure (food and nonfood) per adult equivalent falls below a lower bound threshold estimated in World Bank 2005, Poverty Assessmert. Table A21. Unemployment Rate by Educational Attainment and Age, Ethiopia, 2001 Educational attainment 10­14 15­19 20­24 25­29 30­34 > 34 All ages No schooling 1.6 5.8 5.3 4.7 4.7 3.6 3.7 (14,250) (7,116) (6,587) (6,078) (4,426) (20,030) (58,487) Primary Grades 1­4 4.4 5.1 5.4 2.8 3.5 2.9 4.3 Grades 5­8 6.7 9.6 11.0 7.0 4.5 5.8 7.8 Sub-total 4.7 6.9 8.2 5.0 4.0 4.3 5.7 (4,355) (5,133) (3,965) (3,438) (2,122) (4,366) (23,379) Secondary Grades 9­10 -- 26.3 24.4 14.0 10.7 6.6 16.4 Grades 11­12 -- 45.3 34.4 21.4 12.6 9.4 22.1 Sub-total -- 33.7 30.6 18.8 11.9 8.3 19.9 -- (875) (2,272) (2,397) (1,307) (1,897) (8,755) Higher education -- -- 16.6 9.1 3.7 3.3 7.1 (485) (776) (598) (1,374) (3,258) --Too few observations to compute reliable rates of unemployment Note: Shaded cells refer to the age groups where first-time entrants from each education category are concentrated. Number of observations in the sample in parentheses. Annex Tables and Figures 47 Figure A1: Enrollments in Regular Programs in Higher Education, Ethiopia, 1985-02 Undergraduate Diploma Post-graduate 25.0 20.0 (thousands) 15.0 students of 10.0 Number 5.0 0 1985­86 1987­88 1989­90 1991­92 1993­94 1995­96 1997­98 1999­00 2001­02 Note: data on enrollments in private institutions and evening classes are excluded for lack of a sufficiently long time series. 48 EDUCATION ETHIOPIA IN Figure A2: Primary Gross Enrollment Ratio, Ethiopia and Other Countries, circa 2000 Indonesia Honduras Uganda Rwanda Nicaragua Gambia Nigeria Congo Rep. Angola Chad Guinea Bissau Senegal Ethiopia Sierra Leone Guinea Sudan Congo D.R. CAR 0 20 40 60 80 100 Percent Note: the GER has been standardized for six years of primary schooling for all countries in this figure; CAR refers to Central African Republic; Congo D.R. refers to the Democratic Republic of Congo. Annex Tables and Figures 49 Figure A3: Secondary Gross Enrollment Ratio, Ethiopia, Selected African Countries and Country Group Averages, circa 2000 Sub-Saharan Africa Developing countries Zimbabwe Ghana Togo Kenya Zambia CÙte d'Ivoire Senegal Congo, Demo Rep. Ethiopia Mali Madagascar Guinea Rwanda Tanzania 0 10 20 30 40 50 60 Percent Note: country group averages are population weighted; to improve cross country comparability, the ratio for Ethiopia refers to Grades 7­12. 50 EDUCATION ETHIOPIA IN Figure A4: Relation between GDP Per Capita and Coverage in Higher Education in Low-Income Countries, circa 2000 A: African country O: Other low-income country 800 O A 600 inhabitants A 100,000 per A O A 400 A A A A A Enrollments A AO A A A 200 A O Education A A O AA A AEthiopia O Higher A A A A 50 A A 100 500 1,000 2,000 3,000 Note:GDP,grossdomesticproduct;PPP,purchasingpowerparity;dataforEthiopiareferto2001­02andincludestudentsinevening and Kiremt courses, adjusted to full-time equivalents assuming that such students take 1.5 times as long to complete their studies as those in regular programs . Annex Tables and Figures 51 Figure A5: Trends in Recurrent Spending on Selected Components of Education, Ethiopia, 1993-94 to 2001-02 500 400 Birr constant 1993­94) 300 in level 1994­95 to in 200 (indexed Spending 100 0 1993­94 1994­95 1995­96 1996­97 1997­98 1998­99 1999­00 2000­01 2001­02 TTC Higher education TTI Administration Grades 1­12 Note: TTI refers to teacher training institutes; TTC teacher training colleges. 52 EDUCATION ETHIOPIA IN Figure A6: Spending Per Pupil in Primary Education, Ethiopia and Other African Countries, circa 2000 Ethiopia, Gr. 1­4 Ethiopia, Gr. 5­8 Niger Burkina Faso Eritrea Lesotho Zimbabwe Kenya Sierra Leone CÙte d'Ivoire Mali Senegal Nigeria Togo Gambia, The Mauritania Ghana Burundi Benin Madagascar Sudan Chad Tanzania Uganda Cameroon Rwanda Malawi C.A.R. Guinea Angola Congo, Rep. of Zambia Guinea-Bissau Congo, Dem. 0 10 20 30 40 Note: for countries other than Ethiopia, the data referSpendingper pupil (% of per capita GDP) to spending per pupil in primary education, typically Grades 1­6. Annex Tables and Figures 53 Figure A7: Cohort Entry Rate to Grade 1 and Completion Rates in Grades 4 and 8 Across Regions, Ethiopia, circa 2000 Gr. 1 entry Gr. 4 completion Gr. 8 completion Ethiopia Tigray Afar Amhara Oromiya Somali Benshangul SNNPR Gambella Harari Addis Ababa Dire Dawa 0 20 40 60 80 100 Entry or completion rates (%) Note: data on completion rates unavailable for Afar and Somali. 54 EDUCATION ETHIOPIA IN Figure A8: Primary School Pupil-Teacher Ratios in Ethiopia (2001-02) and Other Countries (circa 2000) Ethiopia Non-African average C.A.R. Chad Cameroon Mali Congo, Rep. Rwanda Burundi Senegal Mozambique Benin Madagascar Malawi Zambia Eritrea Guinea Burkina Faso Mauritania CÙte d'Ivoire Togo Lesotho Congo, Uganda Tanzania Sierra Leone Zimbabwe Nigeria Guinea Gambia, The Niger Ghana Kenya Sudan Angola 0 10 20 30 40 50 60 70 80 Note: Ethiopia's ratio refers to the weighted average for the country's eight year cycle, while those for the other countries typically refer to primary cycles that last five or six years. C.A.R. stands for Central African Republic. The non-African average refers to the data for 21 countries which in 2002 have yet to achieve universal primacy school completion, and whose data are included in the source cited below.