STUDY ON UNDERSTANDING THE ROLE OF NON-STATE EDUCATION PROVIDERS IN SOMALIA FINAL REPORT Prepared by Altai Consulting for the World Bank | Somalia - August 2018 © World Bank May 2018 Unless specified otherwise, all pictures in this report are credited to Altai Consulting CONTENT ACRONYMS ........................................................................................................................................... 8 EXECUTIVE SUMMARY ...................................................................................................................... 10 1. INTRODUCTION ........................................................................................................................ 14 1.1. Background of the Study ........................................................................................................ 14 1.2. Objectives ............................................................................................................................... 14 1.3. Structure of the Report ........................................................................................................... 14 2. EDUCATION SECTOR BACKGROUND .................................................................................. 15 2.1. Brief History of the Education System .................................................................................... 15 2.2. Current Education System ...................................................................................................... 16 Overview of the Education System ................................................................................................ 16 Regional Differences in Education ................................................................................................. 19 Education Policy Framework ......................................................................................................... 19 2.3. Main Education Stakeholders ................................................................................................. 20 Government ................................................................................................................................... 20 Umbrella Associations ................................................................................................................... 21 Civil Society ................................................................................................................................... 23 International Actors ........................................................................................................................ 25 2.4. Study Methodology ................................................................................................................. 25 2.5. School Typology ..................................................................................................................... 27 Lack of Accepted Typology ............................................................................................................ 27 Proposed Typology ........................................................................................................................ 28 3. FINDINGS .................................................................................................................................. 31 3.1. Overview of Survey Sample ................................................................................................... 31 School Types ................................................................................................................................. 31 Student Demographics .................................................................................................................. 32 3.2. School Management ............................................................................................................... 32 3.3. School Ownership ................................................................................................................... 33 Definition of School Ownership in the Current Somali Context ..................................................... 34 Overview of Ownership .................................................................................................................. 34 Ownership by School Type ............................................................................................................ 35 Ownership by Location .................................................................................................................. 36 3.4. School Funding ....................................................................................................................... 36 Major Funding Streams.................................................................................................................. 37 Expenditure and Funding Gaps ..................................................................................................... 39 3.5. Education Quality.................................................................................................................... 41 Evaluating Quality Performance .................................................................................................... 41 Key Staff Qualifications, Availability and Performance .................................................................. 42 Educational Content and Outcomes .............................................................................................. 49 Availability and Condition of Physical Infrastructure and Learning Materials ................................ 51 Evaluating Quality Beyond the Proposed Typology ...................................................................... 56 3.6. Accessibility & Inclusivity ........................................................................................................ 57 Financial Drivers of Exclusion ........................................................................................................ 58 Distance to School ......................................................................................................................... 60 Gender Barriers ............................................................................................................................. 61 Other Drivers of Exclusion ............................................................................................................. 62 Implications for School Choice ...................................................................................................... 63 3.7. Accountability.......................................................................................................................... 64 Long Route of Accountability ......................................................................................................... 64 NON-STATE EDUCATION PROVIDERS Altai Consulting 3 June 2018 Short Route of Accountability ........................................................................................................ 66 4. RESEARCH IMPLICATIONS .................................................................................................... 68 4.1. Policy Implications .................................................................................................................. 68 4.2. Suggestions for Further Research .......................................................................................... 71 NON-STATE EDUCATION PROVIDERS Altai Consulting 4 June 2018 LIST OF FIGURES Figure 1. Structure of the MoECHE ...................................................................................................... 20 Figure 2. Surveyed locations................................................................................................................. 26 Figure 3. Reasons headmasters consider their school to be public (n = 43, multiple response) ......... 28 Figure 4. Regional distribution of survey sample according to new typology (n = 170) ....................... 31 Figure 5. Management by school type (n = 170, multiple response) .................................................... 33 Figure 6. Features of school ownership according to headmasters (n = 170, multiple response) ....... 34 Figure 7. School ownership by school type (n = 170, multiple response) ............................................ 35 Figure 8. School ownership by location (n = 170, multiple response) .................................................. 36 Figure 9. Sources of funding for teacher salaries by school type (n = 170, average per school type) . 39 Figure 10. Main expenses by school type (n = 169) ............................................................................. 40 Figure 11. Average reported annual funding from fees and expenditure by school type (n = 134) ..... 41 Figure 12. Average education quality index score by school type (n = 170) ........................................ 42 Figure 13. Teacher qualifications by school type (n = 130) .................................................................. 44 Figure 14. Average monthly salary of primary teachers by school type (n = 123) ............................... 47 Figure 15. Average monthly salary of secondary teachers by school type (n = 96) ............................. 47 Figure 16. Selected educational outcomes by school type ................................................................... 51 Figure 17. Access to water by source and school type (n = 170) ......................................................... 52 Figure 18. Availability of computers and Internet by school type.......................................................... 54 Figure 19. Textbooks at two independent schools in Bosasso ............................................................. 55 Figure 20. Quality of infrastructure mentioned during FGDs ................................................................ 56 Figure 21. Average time spent travelling to school by school type (n = 170) ....................................... 60 Figure 22. Average time spent travelling to school by location (n = 170) ............................................. 61 Figure 23. Government inspections by school type as reported by headmasters (n = 170) ................ 65 Figure 24. Frequency of government inspections by school type (n = 92) ........................................... 66 NON-STATE EDUCATION PROVIDERS Altai Consulting 5 June 2018 LIST OF FOCUS BOXES Focus Box 1. The typical Somali school and student ........................................................................... 17 Focus Box 2. Origins of umbrella associations ..................................................................................... 22 Focus Box 3. The path to becoming a teacher in Somalia ................................................................... 45 Focus Box 4. Life as a teacher in Somalia ........................................................................................... 48 Focus Box 5. Variation in education quality beyond the proposed typology ........................................ 56 Focus Box 6. Major drivers of cost ........................................................................................................ 59 Focus Box 7. Stakeholder perspectives on inclusivity .......................................................................... 62 Focus Box 8: Factors influencing school choice ................................................................................... 63 Focus Box 9. Perspectives on the short route of accountability ........................................................... 67 Focus Box 10. Country Examples: Afghanistan and South Sudan....................................................... 71 NON-STATE EDUCATION PROVIDERS Altai Consulting 6 June 2018 LIST OF TABLES Table 1. Current state of the education system .................................................................................... 16 Table 2. Methods used during study on non-state education in Somalia ............................................. 25 Table 3. Proposed new school typology ............................................................................................... 29 Table 4. School size and student gender ratio by school type ............................................................. 32 Table 5. Percentage of schools receiving funding from each source, by school type (n = 170) .......... 37 Table 6. Percentage of schools receiving funding from each source, by location (n = 170) ................ 37 Table 7. Availability and gender ratio of teachers ................................................................................. 43 Table 8. Remuneration of teacher compared to other professions....................................................... 48 Table 9. Percentage of schools that have select facilities by school type (n = 170) ............................ 52 Table 10. Student-latrine and student-classroom ratios by school type ............................................... 53 Table 11. Average cost of education to the parent per surveyed school in USD (n = 170).................. 58 Table 12. Average annual cost of education per surveyed school by school type in USD .................. 58 Table 13. Availability of scholarships by school type (n = 168) ............................................................ 60 Table 14. Drivers of exclusion ............................................................................................................... 62 NON-STATE EDUCATION PROVIDERS Altai Consulting 7 June 2018 ACRONYMS ABE Alternative Basic Education ANOVA Analysis of Variance CBO Community Based Organization CEC Community Education Committee COO Chief Operating Officer DCM Data Collection Manager DCO Data Collection Officer DEO District Education Officer DFID United Kingdom Department for International Development DG Director General EA Enumeration Area EMIS Education Management Information System ESP Education Sector Plan ESSP Education Sector Strategic Plan EU European Union FENPS Formal Education Network for Private Schools FGD Focus Group Discussion FGS Federal Government of Somalia FMS Federal Member State FPENS Formal Private Education Network (Note: The name has been changed to UCES, but the umbrella association is still commonly referred to as FPENS.) GER Gross Enrolment Ratio GPE Global Partnership for Education GPS Global Positioning System GTEC Garowe Teacher Education College IDP Internally Displaced Person (I)NGO (International) Non-governmental Organization IQS Integrated Quranic School IT Information Technology JRES Joint Education Sector Analysis KII Key Informant Interview LTMMIS Learning and Teaching Materials Management Information System M&E Monitoring & Evaluation MoE Ministry of Education MoU Memorandum of Understanding MoECHE Federal Ministry of Education, Culture and Higher Education NGO Non-Governmental Organization NON-STATE EDUCATION PROVIDERS Altai Consulting 8 June 2018 OIC Organization of Islamic Cooperation PTA Parent-Teacher Association REO Regional Education Officer SABER Systems Approach for Better Education Results SAFE Schools Association for Formal Education SEDA Somali Education Development Association SMC School Management Committee SNU Somali National University SOFE School Organization for Formal Education SOFEL Somali Formal Education Link SOFEN Somali Formal Education Network TMIS Teacher Management and Information System ToR Terms of Reference TVET Technical and Vocational Training UCES Umbrella Community Education in Somalia Umbrellas Umbrella associations UN United Nations UNESCO United Nations Educational, Scientific and Cultural Organization UNICEF United Nations Children's Fund UNOSOM United Nations Operation in Somalia USAID United States Agency for International Development WFP World Food Program NON-STATE EDUCATION PROVIDERS Altai Consulting 9 June 2018 EXECUTIVE SUMMARY Introduction This document presents the findings of a "Study on Understanding the Role of Non-State Education Providers in Somalia" commissioned by the World Bank. The goal of the study was to develop an evidence base for understanding non-state education providers in the broader Somali education ecosystem, and to identify opportunities for cooperation between state and non-state actors. The report is divided into three main chapters: an overview of Somalia’s education sector and its implications on study methodology, presentation of the study’s findings, and key research implications. Education Sector Background Somalia has developed a fragmented education system stemming from differences introduced by British and Italian colonialists and compounded by the civil war starting in 1991. The war effectively destroyed the public education system and contributed to regional differences as several states claimed greater autonomy. In the absence of a central education authority, non-state actors became the most prominent providers of education. This remains the case, although the federal government has been leading a push towards the reestablishment of public education. The study focused on formal primary and secondary education which comprises eight years of primary schooling and four years of secondary. According to government data there are 925 primary schools and 372 secondary schools with a total enrolment of 987,108 and 195,804 students, respectively.1 Due to funding shortages and limited institutional capacity, outcomes remain poor with low enrolment (20% in primary gross enrolment ratio) and high drop-out rates (64% survival rate to Grade 5).2 Among key stakeholders, the Ministry of Education, Culture and Higher Education (MoECHE) is responsible for the development of education on federal level while Ministries of Education (MoE) operate on the state level. Umbrella associations3 are the most important non-state actor in the Somali education landscape. They fulfill a quasi-governmental role by setting minimum standards for member schools, developing curricula, and facilitating teacher training. Finally, civil society and the international community are also involved in the education sector. In addition to owning and operating schools, community-based organizations (CBOs), non-governmental organizations (NGOs), and private individuals provide funding and technical assistance. Study Methodology A mixed-method approach was used combining quantitative and qualitative methods. These included a quantitative survey of 170 primary and secondary schools across four locations; 27 Focus Group Discussions (FGDs) with students, parents, and teachers; and 15 Key Informant Interviews (KIIs) with government officials, education experts, and private education stakeholders. Given the limitations induced by the sample size, the need to capture a wide variety of school types and a lack of a comprehensive countrywide sampling frame of primary and secondary schools, the decision was taken to sample from high-density areas in three different Federal Member States (FMS). The data collection took place in four urban and peri-urban locations across Somalia: 1 Education Sector Analysis 2018 - 2020 2 Ibid 3 Umbrella associations refer to specific networks of non-state schools in Somalia. Many of these ‘umbrellas’ were originally founded by former teachers to associate schools in the absence of central authority in the 1990s and have grown to incorporate hundreds of schools across the country. Over 1000 schools are either directly managed by an umbrella or more loosely associated with an umbrella network. NON-STATE EDUCATION PROVIDERS Altai Consulting 10 June 2018 Bosasso and Garowe (Puntland), Kismayo (Jubbaland), and Mogadishu (Benadir Regional Administration). School Typology In the absence of a commonly agreed-upon school typology among local stakeholders, one was developed for this report to allow comparisons between different types of schools. Using management to differentiate schools, the survey sample has been divided into four categories:  Government schools managed solely by the government  Mixed-management schools managed by the government together with a non-state actor  Umbrella schools managed solely by an umbrella association  Independent schools managed by a non-state actor other than an umbrella association School management, as used in this report, involves responsibility for key operational decisions such those related to staffing, curriculum, and fees. Findings 1. Overview of Survey Sample Within the sample (N=170), independent schools are most common (39% of the sample), closely followed by umbrella schools (36%), then by government schools (15%), and mixed management schools (10%). There is also regional variation as school types are not uniformly distributed across all locations. Most of the surveyed schools (61%) offer both primary and secondary education, are day schools (99%), and operate in two shifts (79%). The median number of students per school is 391, with a large gender gap in the student body (median of 156 males per 100 females). 2. Ownership In Somalia, the definition of school ownership is contested between state and non-state actors, as the government tends to claim ownership of schools that it built but that are now privately run. Local communities and the government are involved in the ownership of many schools (25% and 22%, respectively), but most umbrella-managed schools are privately owned (85% owned exclusively by private individuals). 3. Funding Most schools rely on fees paid by parents as their principal – and often only – source of funding. The reliance on fees makes schools dependent on parents’ ability to pay, and therefore vulnerable to economic shocks. Government funding is scarce and appears to be mostly limited to schools managed at least in part by the state. Other sources of funding include NGOs, communities, and umbrella associations but, even when combined with revenue from fees, are often insufficient to cover operating expenses. Schools operating at a deficit cannot afford to maintain facilities and equipment or make regular salary payments to teachers. 4. Quality of Education Most of the surveyed schools have the basic infrastructure (access to water, electricity, security wall), facilities (recreation area) and equipment (chalkboard, chalk, basic stationary). The number of students per classroom ranges from 10 to 233 , with a median value of 31. Students have to share textbooks in 65% of surveyed schools and 34% of umbrella schools do not use textbooks at all . Most schools teach a similar variation on the national curriculum. Interviewed students and parents requested more courses in Information Technology (IT), which is taught at 37% of surveyed secondary schools. Surveyed schools tend to mix three languages of instruction: Somali, English and Arabic, although as of August 2018 primary schools are formally required to teach in Somali. Both primary and secondary schools are staffed by teachers with varied levels of education and training, as MoECHE does not yet regulate teacher qualifications. Independent schools have the best qualified teachers, with 75% in possession of a bachelor’s degree, however, overall, 12% of teachers NON-STATE EDUCATION PROVIDERS Altai Consulting 11 June 2018 do not have post-secondary education. The average pay, issued irregularly, tends to range between USD 50 and 350 per month, which is comparable to that of a day laborer or a vegetable vendor. Consistent measure of educational outcomes is not currently available. Although MoECHE now issues a universal graduation exam for secondary schools, the results were not available at the time of research. Interviewed headmasters estimated the exam pass rate (for both primary and secondary) to be 88% at surveyed schools.4 Nevertheless, independent schools were found to produce significantly better outcomes, including the exam pass rate, dropout rate and repeat rate. 5. Accessibility and Inclusivity Accessibility and inclusivity are limited by financial, geographical, and cultural barriers to education. The prevalence of school fees is the main driver of exclusion. On average, annual school fees are USD 99 for primary and USD 149 for secondary, with the total cost rising to USD 132 and 197, respectively, when additional costs are included. Although limited scholarships are available at some schools, they are insufficient to guarantee universal access to education. The large gender gap (see above) in student enrolment indicates that there are additional barriers for girls. Finally, distance is likely a barrier in rural areas, although less so in urban areas where the research took place and where the average student at 92% of schools travels less than 30 minutes to get to school. 6. Accountability Accountability mechanisms are underdeveloped with limited efficacy. CECs and Parent-Teacher Associations (PTAs) interact with school management to voice their concerns and improve quality and outcomes in an ad-hoc manner, while umbrella associations set and monitor standards in response to market pressures with limited and unstructured interaction with government institutions. On the government side, the MoECHE facilitates accountability by collecting data during school inspections and collating it within the Education Management Information System (EMIS), to a limited degree of success. Within the survey sample, only 54% of schools were inspected by the government . Research Implications Policy Implications As demonstrated by the findings, the education sector in Somalia suffers from a number of weaknesses. The education landscape is fragmented, with a number of state and non-state stakeholders, who engage in a limited and unstructured interaction. The government agencies lack capacity for policy planning, standards setting, financing, collaboration with private providers and data collection. The individual schools are under-funded, struggle with a lack of qualified teachers and are therefore unable to provide adequate service to all school-aged children in surveyed areas. Finally, there is a significant gender gap in enrolment, as described above. Stakeholders involved in the sector might consider the following interventions: 1. Induce cooperation between the government institutions (MoECHE and MoEs) and umbrella associations to encourage the spread of common standards and build the FGS capacity. The FGS should take a collaborative approach to engaging with the umbrella associations and establish a formal discussion platform. Umbrella schools suffer from the same set of challenges as other school types and umbrellas are therefore likely to be receptive to assistance with funding, textbooks and equipment, and teacher capacity. 2. Clarify the roles and responsibilities of the MoECHE and the state-level MoEs to reduce stakeholder fragmentation within the education landscape. 4 In the survey, headmasters were asked to provide a pass rate for any final exam that is administered at their school. This includes the government-mandated final secondary exam, as well as final primary exams that tend to take place at individual schools but that are not nationally harmonized. NON-STATE EDUCATION PROVIDERS Altai Consulting 12 June 2018 Establishing a jointly agreed upon and implemented system for school inspections could serve as one of the first building blocks of a constructive relationship between MoECHE and the MoEs. 3. Expand, reinvent and refine the business model for education to improve quality and stimulate expansion of the school network. To deliver cash support directly to individual schools, consider a grant delivery mechanism managed jointly by the donors and the government that issues grants to individual schools, conditional on fulfilment of a set of criteria. In addition to direct financial support, lessons learnt on sustainable business models should be collected and distributed. 4. Extend access to education for women and children from families that cannot afford education through systemic change. To address the substantial gender gap, ensure that gender mainstreaming is fully integrated into the teacher training curriculum and that an adequate number of female teachers, on primary and secondary level, is receiving their qualifications. Financial support for vulnerable students could be extended through grant delivery mechanisms mentioned above. 5. Invest in systems that support teacher training and education. Teacher performance baseline, Teacher Management and Information System (TMIS) and teacher training curriculum should form the basis of improvement to teaching quality. Alternatives to in-service training, including ICT solutions and introduction of mobile tutors should be explored. However, as long as teacher compensation remains minimal, any gains in improving teaching capacity are easily lost. 6. Encourage and stimulate data collection and reporting within umbrella associations and on the level of individual schools. Strengthen EMIS data collection procedures to compile a more comprehensive audit of primary and secondary schools in Somalia. Lessons from other fragile countries, such as South Sudan and Afghanistan, offer lessons on taking advantage of ICT and adopting a decentralized approach to data collection. Suggestions for Further Research To build upon the findings of this study, it is recommended that future research initiatives consider investigating: (1) rural schools, including schools catering to nomadic populations; (2) non-state actors with special focus on the in-depth study of umbrella associations; (3) the decision-making process on school leadership as one of the main determinants of education quality in the current system, where questions of curriculum, teacher hire and overall resource allocation are decided on the school level; (4) teacher performance baseline to establish the quality of teaching in the absence of government regulation; and (5) funding to obtain more granular data on how schools raise, allocate and spend funds. NON-STATE EDUCATION PROVIDERS Altai Consulting 13 June 2018 1. INTRODUCTION 1.1. BACKGROUND OF THE STUDY In February 2018, Altai Consulting was commissioned by the World Bank's Somalia Education Sector Programmatic Advisory Services and Analytics to conduct a " Study on Understanding the Role of Non-State Education Providers in Somalia". The study focuses on formal primary and secondary schools in urban and peri-urban areas across four locations in Somalia: Mogadishu (Benadir Regional Administration), Kismayo (Jubbaland), and Garowe and Bosasso (Puntland). 1.2. OBJECTIVES Overall Objective The overall objective of this study is to help relevant Somali government institutions and the international community in developing an evidence base for understanding non-state education providers in the broader Somali education ecosystem, and in identifying opportunities for streamlining private initiatives into federal and state-level education sector development policy. Specific Objectives More specifically, the study aims at: a) Developing an in-depth understanding of the characteristics and role of non-state education providers in primary and secondary education in Somalia; b) Analyzing their operating environment and identifying its constraints; c) Identifying areas and approaches for public-private partnership; and d) Delivering recommendations for the government that can feed into a possible regulatory policy framework supporting non-state education service providers. 1.3. STRUCTURE OF THE REPORT The report is structured in the following way:  The second chapter gives a brief overview of the Somali education sector and outlines the study’s methodology.  The third chapter presents findings related to objectives a) and b) listed above. Separate sections within this chapter present findings on school management; school ownership; school funding; education quality; accessibility and inclusivity; and accountability.  The final chapter relates to objectives c) and d), providing recommendations for governmental stakeholders and identifying areas of opportunity for public-private partnerships. NON-STATE EDUCATION PROVIDERS Altai Consulting 14 June 2018 2. EDUCATION SECTOR BACKGROUND 2.1. BRIEF HISTORY OF THE EDUCATION SYSTEM Somalia’s education system has been fragmented by the country’s tumultuous past. British and Italian influences, the absence of a central education authority between 1991 and 2006, and the subsequent emergence of non-state education providers are at the heart of this fragmentation. The following sub-section will briefly outline the evolution of education in Somalia. In pre-colonial Somalia, education was dispensed through informal systems of communal interaction, and pastoralist parents taught their children orally in flexible sessions, relevant to their basic needs.5 Additionally, Quranic schools were among the first educational institutions in Somalia, teaching basic literacy skills and the Quran in the Arabic language. The importance of religion grew in the 14th century as Islamic religious groups including Sufi religious orders spread prayers, practices, and Arabic religious writing among the Somali population. The mixture of Islamic scholarship and Western ideas introduced by Somali travelers fueled debates on education and religious reform, even prior to European colonization.6 In the late 19th century, Somalia was divided between the United Kingdom and Italy. British Somaliland was established in 1884 in the northwest while the eastern coast became Italian Somaliland in 1890. Both colonial powers sought to formalize education through an elitist approach focusing on urban schools and higher education institutions that primarily served the colonial population and local elites. Consequently, English and Italian were the main languages of instruction. In addition, limited financial support was extended to existing Islamic schools and Quranic teachers who taught literacy, numeracy, Arabic, and English alongside religion in a moderate attempt to reach the rural nomadic population. However, it was the emphasis on urban and higher education that continued to define the Somali education system into the modern era. The origin of the modern Somali education system can be traced to the unification of the British and Italian colonies and the creation of an independent Somali state in 1960. All schools were consolidated into a single system of around 1400 primary schools and 50 secondary schools , with the assistance of international donors including the United States Agency for International Development (USAID). Additionally, several vocational-technical institutes, a National Teacher Education Center, and a National University were created. The Somali written language was introduced in 1972 under the military regime of Siad Barre and literacy campaigns were launched aiming to bring education to the masses. In the same year, the Somali language was promoted as the official language in public administration and the formal education system, leading to a steep increase in enrolment and a struggle to develop Somali-language textbooks. In 1975, education was made free and compulsory for children up to 14 years old. The education system built in the 1960s and 1970s was effectively destroyed during the civil war that erupted in 1991. The displacement of large segments of population, including many teachers, led to the closure of schools and the premises being used for alternative purposes.7 According to estimates, 5 Abdi, A., Education in Somalia: History, Destruction, and Calls for Reconstruction, in: Comparative Education 34(3), 327-340, 1998; UNICEF, Despite All Odds, n.d. URL: https://www.unicef.org/somalia/SOM_NewDawndoc.pdf 6 For more information on history of education in Somalia, see Cassanelli, L. and Abdikadir, F., 2007. Somalia Education in Transition, in: Bildhaan, 7 (1), 91-125; Abdinur A., 2008. Community Assumes the Role of State in Education in Stateless Somalia, in: International Education, 37(2), 43-61 7 Cummings, K., 2003. Basic Education and Policy Support Activity, Somalia Education Sector Assessment: with Special Attention to the Northwest Zone. NON-STATE EDUCATION PROVIDERS Altai Consulting 15 June 2018 90% of school infrastructure and materials were destroyed.8 Once the United Nations Operation in Somalia (UNOSOM) troops withdrew in 1994, most initiatives to rehabilitate schools failed as a result of the absence of a central authority and lack of coordination among key stakeholders. In the absence of a public education system, private umbrella associations were formed to fill the void left by the government. These education umbrellas – along with international donors, local communities, and the Somali diaspora – assumed key roles in reconstructing schools, education service delivery, and regulation. The following sections will describe the role of the government, umbrella associations and other actors in the education sector in more detail. 2.2. CURRENT EDUCATION SYSTEM OVERVIEW OF THE EDUCATION SYSTEM The current Somali education system emerged in 2012 with the establishment of the Federal Government of Somalia (FGS). The new government established free primary and secondary education as the basic right of all Somali citizens 9 and has taken steps to standardize education across the country. The Ministry of Education, Culture and Higher Education (MoECHE) manages the education system although non-state actors maintain their influential roles. Education begins with early childhood education (including pre-schools and kindergartens), then progresses to primary and secondary schooling, and finally to post-secondary education.10 This report focuses on primary and secondary education , which generally comprises eight years of primary (Grades 1-8) followed by four years of secondary (Forms 1-4).11 In Mogadishu and Garowe schools are officially supposed to be open 10 months a year, while in Kismayo and Bosasso schools run for nine months a year, in part to avoid the heat of summer. As a result of overcrowding, many schools operate in two shifts – one in the morning and one in the afternoon. 12 Although the education sector is undergoing reforms, it remains underdeveloped and educational outcomes remain poor, especially for girls and in rural areas. Key indicators characterizing the current state of the education system are presented in Table 1 below. Table 1. Current state of the education system13 Indicator Urban Rural Male Female Total Number of schools * Primary — — — — 925 Secondary — — — — 372 Alternative basic education — — — — 11 8 Barakat et al., 2014. Beyond Fragility: Conflict and Education Analysis of the Somali Context; Cummings, K., 2003. Basic Education and Policy Support Activity, Somalia Education Sector Assessment: with Special Attention to the Northwest Zone. 9 The Federal Republic of Somalia, 2012. Provisional Constitution. 10 MoECHE, 2017. Education Sector Analysis 2018-2022. 11 A 9-3 structure is also used in some private Arabic institutions but informants indicated that the 8-4 system is most prevalent in the survey locations and is therefore used throughout the report. Additionally, there are plans to begin using 4-4-4 system with an intermediate school level in August 2018 (UNSOM, 2018). 12 MoECHE, 2017. Education Sector Analysis 2018-2022. 13 All data is from the Education Sector Analysis 2018-2022. Integrated Quranic schools are included with primaries in government data. NON-STATE EDUCATION PROVIDERS Altai Consulting 16 June 2018 Quranic schools — — — — 2,967 Number of teachers Primary 12,366 5,275 15,509 2,132 17,641 Secondary 5,015 412 5,265 162 5,427 Number of students Primary 638,538 348,570 563,680 426,385 987,108 Secondary 184,955 10,849 119,593 76,211 195,804 Gross enrolment ratio * Primary 48.2 4.4 22.2 19.8 20.1 Secondary 36.6 0.3 15.5 10.9 13.3 Survival rate to Grade 514 Total 70.8 47.9 65.9 61.7 64.0 Exam pass rate (2016) * Grade 8 — — — — 92% Form 4 — — — — 87% * Excludes Somaliland and Puntland — Indicates data either unavailable or not applicable Based on the information presented in this chapter and the results of the survey presented in the chapter 3 ‘Findings’, it is possible to create a picture of the typical school and student in Somalia (Focus Box 1). Focus Box 1. The typical Somali school and student15 Typical School Typical Student A typical Somali school would be a primary A typical Somali student would be between 8 school located in an urban center, running 232 and 10 years old and enrolled in the first years days (or about 9 months) a year. It would be of primary (Grades 1-3). He would be a boy managed by a non-state actor such as the local living in an urban center and would travel community, or an umbrella association, and between 10 to 30 minutes to get to a typical non- owned by private individuals. There would be state school such as the one described on the around 400 students, 13 teachers, and 12 left. His parents would work at the local market classrooms. Subjects would include Arabic, or as day laborers. Although they might struggle Math, English, Somali, Islamic studies, and to pay his school fees, the boy would be 14 Percentage of a cohort enrolled in Grade 1 expected to reach Grade 5 15 The analysis presented in the Focus Box draws on the 2016 EMIS report that indicates that three quarters of primary and secondary enrolment is located in urban areas and on findings from the survey implemented predominantly in urban areas for this report. In both cases of a ‘typical school’ and a ‘typical student’, the most common characteristics were used to construct the example. Note that the descriptions of both the school and the student refer to the formal education system, which is the focus of the report. NON-STATE EDUCATION PROVIDERS Altai Consulting 17 June 2018 Science and would be taught in a mix of Somali fortunate to be part of the population that can and English. School fees would be charged but afford to go to school. Drought, conflict, and would not fully cover running costs. economic uncertainty would likely prevent him Consequently, facilities and learning materials from completing his studies. However, if he does are limited to the basics such as water, manage to stay in school, he would likely pass electricity, and a few textbooks. his final exams. Religious education is highly valued in Somalia with many children attending informal Quranic schools, sometimes in addition to attending a formal primary or secondary school.16 These institutes teach students to read, write, and recite the Quran. In addition, the United Nations Children's Fund (UNICEF) introduced integrated Quranic schools (IQS) that employ qualified teachers who teach both the secular and religious curricula.17 Unlike traditional Quranic schools, IQS are considered part of the formal education system. More information on the characteristics of the IQS visited for the purpose of this research can be found in Focus Box 5. Alternative Basic Education (ABE) schools and centers operate in parallel to formal schools to provide education to out-of-school children, especially marginalized groups such as nomads, orphans, and displaced people. ABE schools provide a condensed and accelerated form of the formal primary education program for children 7-14 years old,18 with the aim of transferring learners to the formal system or adult literacy programs. Government sources acknowledge that their record of 11 ABE schools is likely outdated;19 in comparison, USAID’s initiative to provide education to pastoralists has recently opened 93 ABE centers.20 ABE is considered part of the formal education system aimed at school-aged children by the FGS, although non-formal basic education programs are also being promoted as a means of reaching uneducated youth and adults. Post-secondary education in Somalia has been developing quickly over the past 20 years. Prior to the collapse of the education system in 1990, the Somali National University (SNU), a government institution founded in Mogadishu in 1954, was the only post-secondary institution in the country. The SNU remains prominent but has been joined by over 50 – mostly private – registered universities, the majority of which are located in Mogadishu.21 The MoECHE oversees universities but its ability to monitor them is limited. Education quality is further limited by the lack of available funding. Technical and Vocational Education and Training (TVET) institutes are also proliferating as young Somalis seek to acquire employable skills. However, TVET institutes also face funding shortages and the system is fragmented and largely unregulated. Post-secondary institutions are crucial to the education sector as the source of new teachers. The government supports teacher training through the SNU while government teacher training institutes22 16 Available data indicate that there are more informal Quranic schools than formal schools (2,967 and 1,297, respectively, excluding Puntland and Somaliland - Education Sector Analysis 2018-2022). 17 The first IQS was initiated in a pilot project together with the government in Somaliland in 2006. The concept was expanded within the Strategic Partnership for the Recovery and Development of Education in Somalia that UNICEF, UNESCO and DfID formed (UNICEF, Evaluation Report. The Integrated Quranic Schools Pilot Project, 2011). 18 Ministry of Human Development and Public Services, Programme Implementation Grant Global Partnership for Education, Programme Document 2014-2016. 19 MoECHE, 2017. Education Sector Analysis 2018-2022. 20 USAID, 2018. https://www.usaid.gov/sites/default/files/documents/1860/Fact_Sheet_-_Somalia_ABE_January_2018.pdf 21 MoECHE, 2017. Education Sector Analysis 2018-2022. 22 In addition to government institutes, private teacher training institutes have also been set up. These include Mogadishu University, Benadir University, SIMAD University, the University of Somalia, and the umbrella-owned Benadir Teacher Training Institute. There is currently no unified teacher training curriculum, therefore the syllabus tends to very between institutions. NON-STATE EDUCATION PROVIDERS Altai Consulting 18 June 2018 also exist in Garowe and Hargeisa. However, the SNU’s teacher training department lacks funding and trainers. It was only in 2017 that the SNU’s College of Education hired an external advisor to start developing an undergraduate degree specific to early childhood and primary school teaching. 23 Although SNU representatives claim the teacher training department does not receive any funds from the government, government stakeholders suggest this is a consequence of ineffective internal resource allocation. REGIONAL DIFFERENCES IN EDUCATION Strong regional differences have emerged as the education sector has been shaped by the civil war and subsequent political turmoil. In the aftermath of the war, the country’s administration became decentralized as several autonomous regions, including Somaliland, Puntland and Galmudug, emerged. The turn of the century marked a shift back towards federalism as the Transitional National Government was established in 2000 and was subsequently succeeded by the Transitional Federal Government in 2004. The current national administration, the Federal Government of Somalia (FGS), was established in 2012. Under the Federal Republic of Somalia, state governments, recognized as Federal Member States (FMS), retain some degree of autonomy over regional affairs. Somaliland, the northwest portion of Somalia and former British colony, declared its independence in 1991. Its independence has not been officially recognized by the majority of the international community or the FGS, which considers Somaliland to be part of the federal state. To the present day, relations with the FGS remain strained and border disputes with the neighboring state of Puntland are ongoing. Consequently, Somaliland’s education sector is managed by the regional administration, which has allocated a greater percentage of its budget to education than other FMSs.24 This, along with the relative peace and stability in the region, have allowed the development of a more centralized and regulated education system. Free primary education was introduced in 2011 but the success of the initiative has been limited by a lack of government capacity and resources.25 The situation is similar in the state of Puntland, which was established as a semi-autonomous state in 1998 but, unlike Somaliland, does not seek full independence. Compared to the southern FMSs, Puntland has been relatively secure, enabling a stable regional administration with a greater capacity for service delivery. The Puntland Ministry of Education (MoE) has been actively working towards the goal of achieving universal primary education. The remainder of Somalia has been most affected by the civil war and ongoing Al-Shabaab insurgency which have fostered political instability. Prior to the establishment of the FGS, public institutions were weak and lacked the capacity to deliver services to the population. Consequently, there are much fewer government-run schools as non-state actors, especially umbrella associations, were relied upon to provide education. Despite the progress towards the reestablishment of a public education system, the sector remains fragmented as schools are mostly unregulated. EDUCATION POLICY FRAMEWORK Since the establishment of the FGS, several policy initiatives have been developed to strengthen the education sector. In 2012, the first Joint Education Sector Analysis (JRES) was completed to evaluate the state of the education sector and inform an Education Sector Strategic Plan (ESSP) for 2013/14-2015/16. The MoECHE has also collaborated with national and international stakeholders to develop Education Sector Plans (ESP), the current of which covers 2018-2020. Somaliland and 23 Despite campaigning efforts for more students to apply, the dropout rates are high: out of 60 students that enrolled in the new bachelor for primary education, in May 2018, only 12 students remain. 24 MoECHE, 2017. Education Sector Analysis 2018-2022. 25 IRIN, 25.01.2011 NON-STATE EDUCATION PROVIDERS Altai Consulting 19 June 2018 Puntland have developed their own policy documents, including JRESs and ESSPs. Puntland has signed a memorandum of understanding (MoU) with the FGS to define responsibilities and align strategic objectives. Policy plans aim to engage non-state actors in order to standardize education and regulate schools. Existing documents outline plans for the development of a unified national curriculum, administration of national exams, and regulation of teacher qualifications. Planned policies will further these objectives by supporting the development of Somali textbooks, establishing a standardized teacher training curriculum, and further regulating non-state schools. For more information on existing and anticipated policy plans, please refer to Annex 3 ‘Literature’. 2.3. MAIN EDUCATION STAKEHOLDERS GOVERNMENT On the federal level, the MoECHE is responsible for the overall education policy framework, as discussed in the previous section. It oversees the unified national graduation exams for the secondary school students and is currently in the process of designing a unified national curriculum. Additionally, the MoECHE also directly manages a relatively small number of schools where teachers are at least partially paid with MoECHE funds. These schools are identified as ‘government schools’ throughout this report. The main divisions of the MoECHE are outlined in Figure 1 below. Figure 1. Structure of the MoECHE Minister of Education Secretary General Director General Federal Departments: Federal Departments: Regional Departments Education Development Curriculum Development Policy and International Relations Planning and Higher Education Education Development Examination and Certification Umbrella and Private Education  Funding: According to a government official, 90% of the MoECHE’s funding comes from the international community.26 The MoECHE contributes around 4% to the education sector’s overall funding,27 which mostly goes towards covering its own running costs, national examinations, and incentives for teachers teaching at government schools.  Coordination: The MoECHE and the MoE of the FMS are in regular communication with each other. The Education Minister and the Director General (DG) of the FGS MoECHE meet with the ministers and the DGs of the FMS ministries every three months. At monthly Education Sector Coordination meetings that the MoECHE chairs and UNICEF co-chairs all governmental and non-governmental stakeholders come together for discussions regarding the education sector. 26 Major donors include GPE, UNICEF, EU, and USAID. 27 Interview with a government official in Mogadishu in April 2018 NON-STATE EDUCATION PROVIDERS Altai Consulting 20 June 2018 The Department for Umbrella and Private Education was created within the MoECHE for the purpose of coordinating with private education stakeholders and umbrella associations. The department director together with the DG of the MoECHE hold frequent ad-hoc meetings with umbrella association representatives. Thus far, the interaction between the MoECHE and umbrella associations has not been formalized or regulated. However, the MoECHE does register umbrellas and their affiliated schools. Representatives of the federal-level MoECHE and state-level MoEs both operate on the sub-national level, often with little coordination. On the side of the MoECHE, the Regional Education Officers (REOs) and the District Education Officers (DEOs) work in their region of deployment during the school year and report back to the MoECHE during school holidays. According to the Director of the MoECHE Department for Umbrella and Private Education, the REOs report to him, although their responsibilities go beyond private and umbrella education. The primary role of the REOs is to support the MoECHE in the running and delivery of education programs in the regions, and to establish an institutional framework for the effective delivery of education services. They serve as the direct link between the MoECHE and the schools in the region and are expected to keep a comprehensive record. They are also in charge of human resources in the region, including deployment and discipline of teachers and head teachers, and of quality assurance including inspections. Together with representatives of the MoECHE and the National Examination Board, they coordinate national exams in the secondary schools. The DEOs support and report directly to the REOs. In practice, the REOs receive a salary but are not allocated a budget to cover, for example, the cost of transportation to the schools they are expected to inspect, which limits the extent to which they can fulfill their responsibilities.28 Similarly, due to the lack of funds, most districts in the southern states do not have a DEO present.29 Even where the REOs and the DEOs are present and able to work, a criticism frequently heard in the course of research was that the REO reports to the MoECHE only rarely trigger a reaction from the central Ministry staff. Each state has its own MoE, with their own budget allocated by the state government.30 The more proactive and well-funded MoEs, in states like Puntland or Somaliland, generate their own education policies, as described in the previous section 31 and have their own teacher training institutions. The Puntland MoE has also issued its own Terms of Reference (ToR) for the REOs, requiring them to report to the DG of the Puntland MoE.32 The ESSP 2018-2020 leaves the respective roles and relationship between the MoECHE, its staff and the state-level MoEs undefined, however, a number of the interviewed officials criticized the role of the REO as outdated and designed for a centrally funded and managed education system, which has now become obsolete in the context of the Somali federation. UMBRELLA ASSOCIATIONS Umbrella associations were formed by former teachers now leading the associations, with an interest in re-establishing education after the civil war had left the education system destroyed, and students and teachers scattered. The founders did not contribute significant funding to the umbrellas but turned to the diaspora and international organizations for initial investments. Now, the umbrella associations and their schools run off school fees and smaller contributions from international 28 Interview with a government official in Mogadishu in April 2018 29 Ibid 30 Puntland, for example, allocated 8% of the state's budget to the education sector, about USD 70 million, according to the Director General of the Puntland Ministry of Education. 31 For example, the Puntland Education Sector Strategic Plan 2012-2016 32 Puntland State, Job description for Regional Education Officer, n.d. NON-STATE EDUCATION PROVIDERS Altai Consulting 21 June 2018 organizations. Despite the government restoring its role in the education sector, the position of umbrella associations remains strong as important providers of education and guarantors of minimal education quality. However, the MoECHE is reclaiming core responsibilities such as regulating the education sector, developing curricula and textbooks, and administering national exams that had originally been assumed by umbrella associations in the absence of a central government. Currently, there are 14 different umbrella associations, which vary greatly in size and reach. While umbrellas have a particularly strong presence in the Benadir administrative region, they operate across all FMSs. Government data indicate that there are over 1,000 umbrella-affiliated schools in Somalia providing education to over 250,000 students. To put the umbrella enrolment rates in perspective, the MoECHE, according to its own records, manages a total of 93 schools, with 31,856 enrolled students, and 1,059 teachers across the country. Focus Box 2 provides details on the two largest umbrella associations while information on other umbrellas can be found in Annex 1 ‘Background Information’. Focus Box 2. Origins of umbrella associations Umbrella associations: "education warlords"33 or useful regulators? The Formal Private Education Network (FPENS) is the largest umbrella association in Somalia with 1080 schools. It was founded in 1993 by a group of teachers in an effort to introduce more regulation into a sector devastated by the civil war and thus improve the quality of the service delivered to the Somali population. An additional motivation was the recognized need for a body that could issue secondary school graduation certificates recognized outside of Somalia, so that Somali students could continue their education even in the absence of formal government certification, which was at the time missing in Somalia.34 The Schools Association for Formal Education (SAFE), the second largest umbrella with 300 schools, was founded in 1998. The association started with five schools and quickly expanded by working with schools to introduce standardized certification and curriculums. Since most textbooks had been destroyed during the war, the founders of SAFE worked to reconstruct the syllabus taught by the government schools before the war and used it to provide member schools with lists of topics and lesson outlines.35  Role: The role of an umbrella is to secure minimum standards at the school level, to provide management guidance to schools and to organize teacher training,36 to connect schools to donors and to a larger network more generally, and to provide a platform for schools to express their concerns to the government.37 Over the years, umbrellas have grown in size and importance, exercising significant influence on the education sector. One of the most significant examples of this influence was observed in 2013, when several umbrella 33 A term used by a government official to describe umbrellas according to an interviewed representative of an umbrella. 34 Interview with an umbrella representative in Mogadishu in April 2018 35 Interview with an umbrella representative in Mogadishu in April 2018 36 The Benadir Teacher Training Institute is owned by an umbrella 37 Umbrellas used to provide secondary school graduation exams that the government certified. Since 2016 the government only issues exam certificates if the students sit government-approved national exam. Now, the umbrella associations are tasked to gather information and fees necessary for the government in preparation for the exam. NON-STATE EDUCATION PROVIDERS Altai Consulting 22 June 2018 associations38 agreed on a unified syllabus, introducing a degree of syllabus standardization for the first time since the end of the civil war.  Membership: For a school to join an umbrella, it has to register with the government, pay a registration fee and commit to comply with the standards set by the association. Once a member, the school pays a small monthly fee per class.39 If a school does not wish to be directly managed by an umbrella, there is an option of a looser association. Associated schools pay a registration and a monthly fee in return for a range of services including the right to use the umbrella association branding. While schools that are directly managed by an umbrella association have been found only in Mogadishu, associated schools were found in all survey locations. Within the survey sample, 66% of schools claimed to be associated with an umbrella while the MoECHE estimates this figure to be 81% overall. The differences between schools associated with umbrellas and those that are not are explored at the end of Section 3.5.  Structure: Each larger umbrella association has a regular general assembly gathering, which brings together the member school headmasters and the executive committee, often comprised of the founding members of the association. The executive committee elects a Board Chairman as well as a Chief Operating Officer (COO), who works with the board on strategic development and policy decisions. The frequency with which a general assembly convenes varies among the umbrellas, in the case of FPENS it is every two years.  Regulation and oversight: In the absence of formalized communication, ad-hoc informal meetings take place between the MoECHE staff and umbrella representatives. Umbrella associations are also regularly asked to provide information on the schools they manage, including data necessary in preparation for the government-organized secondary-level final exams. Additionally, umbrella associations require a certificate from the MoECHE and the Ministry of Interior and Federal Affairs that is to be renewed every two years and schools must be registered with the Government.  Relationship with the government: The relationship between the umbrella associations and the government is characterized by frequent tension even as cooperation between the two does take place. Although interviewed government officials acknowledged the umbrellas' importance, there is a tendency among high-level government officials to see umbrellas as competition. Some umbrella representatives also hold government positions, thus further complicating the relationship. This relationship, however, varies by region. For example, in Puntland the state-level MoE has tasked the umbrellas with implementing the national secondary school exams, while the government continues to issue the certificates. On the other hand, there is a concern among interviewed umbrella officials that the 14 associations do not speak to the government as one voice.40 CIVIL SOCIETY The most commonly encountered community-based organizations (CBOs) active in education are Community Education Committees (CEC), community-based oversight bodies that assists the school 38 The following umbrella associations agreed to implement unified syllabus: FPENS, SAFE, FENPS, SOFE, SOFEN, SOFEL, and SEDA. 39 The monthly fee varies between different umbrella associations but can be estimated at around USD2 per class. The one-off registration fee is between USD100 and 200. Some smaller umbrella associations do not ask for fees. 40 An umbrella association advocacy board for private education was mentioned as option during an interview with an umbrella representative. NON-STATE EDUCATION PROVIDERS Altai Consulting 23 June 2018 administration. Introduced to Somalia by UNICEF,41 CECs were designed to fulfill a role comparable to School Management Committees42 (SMCs). However, in practice, they are not effectively involved in school management and operational decision-making. Instead, they often only facilitate communication between school management and the community and help with fundraising. By providing a forum for parents and community stakeholders to voice their concerns to the school management, CECs help hold school management accountable. CEC members may also visit schools to assess quality. On their end, the teachers and headmasters use CECs to communicate with parents, for example to address the issue of students dropping out of school. Additionally, CECs are sometimes involved in teacher recruitment processes and the management of teacher payroll. The CECs tend to comprise parents, members of the local community, and can also include local elders, teachers, education authorities, or other community members with management experience. Members are often selected from parents of current students based on their social standing in the community and their abilities. Interviews with umbrella associations and the chair of a CEC also suggest that it is often school founders who become CEC members. The CEC members tend to be elected or appointed and their precise role tends to differ from school to school. Interviews with government officials of the MoECHE suggest that CECs are not integrated into institutional structures or policies but are rather seen as way for parents to have a say in education at the school-level, independently of the government. However, government – especially at the local level – provides capacity-building and support services to CECs.43 According to government officials, about half of the schools have a CEC. Across the locations surveyed for this report, CECs were found at the majority of schools in Bosasso, Garowe and Kismayo, but only 26% of surveyed schools in Mogadishu. The difference in Mogadishu appears to be driven by a high number of schools managed by non-state actors, including umbrella associations, that do not have CECs. Parent Teacher Associations (PTAs) are school-based organizations comprised of parents and teachers, intended to facilitate parental participation at the school level. The roles of PTAs tend to overlap with those of CECs and include the monitoring of student performance; conflict resolution between parents, students and teachers; and organizing volunteer activities to the benefit of schools or the wider community. The main distinction is their composition, as CECs tend to include members of the community while PTAs do not. One government official, however, reported that both names are often used interchangeably. Local NGOs play a role in funding education including the teacher payrolls at individual schools, building and maintaining school infrastructure, providing teacher training and general capacity training, as well as emergency education. Finally, private individuals play an important role in the education system as founders and owners of schools. Of the schools surveyed for this study, 65% were owned by private individuals. In many 41 UNICEF, 2011. Evaluation Report Impact Evaluation of the Community Education Committee Mobiliser Programme in Somaliland, Puntland and South/Central Somalia. UNICEF has supported the Ministry-led Mobilisers’ Programmes since its inception in 2004 in Somaliland and Puntland. In Central and Southern Somalia, Regional Education Umbrellas work through local organizations to support teachers and CECs in a similar way. 42 SMCs are active across the world and serve to increase accountability by involving parents and the community at large in school management. 43 UNICEF, 2011. Evaluation Report Impact Evaluation of the Community Education Committee Mobiliser Programme in Somaliland, Puntland and South/Central Somalia. NON-STATE EDUCATION PROVIDERS Altai Consulting 24 June 2018 instances, these individuals were members of the diaspora who then left the school’s management to the umbrella associations or local communities. INTERNATIONAL ACTORS Key international stakeholders include bilateral and multilateral donors, national development agencies and international non-governmental organizations (INGOs). Multilateral agencies such as the European Union (EU), the World Bank, and the United Nations’ (UN) constituent agencies – especially UNICEF – tend to provide financial support, capacity-building, and strategic direction. INGOs, the two most active of which are CARE International and Save the Children, often serve as implementing partners for grants to the education sector. The multi-stakeholder funded Global Partnership for Education (GPE), which has contributed over USD 32 million to the education sector since 2012, is among the key sources of international assistance. Examples of specific programs include the ‘Go-2-School’ initiative to increase primary enrolment, the British-led ‘Girls Education Challenge’ to promote female education, and the Qatari ‘Educate a Child’ program to increase access to and quality of education. More information on international stakeholders and programs can be found in Annex 1 ‘Background Information’. 2.4. STUDY METHODOLOGY This section presents a brief overview of the methodology employed to collect, process, and analyze data presented in this report. For additional details, please refer to Annex 2 ‘Methodology’. The list of literature consulted for this report can be found in Annex 3 ‘Literature’. A mixed-method approach was used combining quantitative and qualitative methods (Table 2). This approach allowed information to be collected from various sources and triangulate findings. Table 2. Methods used during study on non-state education in Somalia Method Description Sample Size Structured interviews with headmasters of schools employing a close-ended tool to gather information at the school level. Major Quantitative themes covered in the questionnaire include school 170 schools survey management and ownership, cooperation with the government, school finances, quality of education, and infrastructure and learning materials. Semi-structured discussions with parents, teachers and Focus group students at schools in each of the surveyed locations. FGDs discussions 27 FGDs served to gather qualitative insights and triangulate findings (FGDs) from the quantitative survey. Semi-structured interviews with government officials, education Key informant experts, and private education stakeholders. KIIs served to 15 KIIs (+ 2 formative interviews (KIIs) gather information on the structure and state of Somalia’s interviews) education system. The design of the sampling approach was driven by the objectives of the study, as listed above, but also by the need to accommodate and compensate for the overall lack of reliable school-level data in Somalia. During the inception period, it became clear that existing school-level data, provided by the government and the dominant non-state actors, was not sufficiently comprehensive on its own to furnish a sampling frame on either the countrywide or local level. To generate an adequate sampling NON-STATE EDUCATION PROVIDERS Altai Consulting 25 June 2018 frame and ensure the representativeness of the survey results, enumerators were therefore required to verify the location and basic characteristics of each school in the surveyed area. In light of these challenges and the relatively small sample of 170 schools, the decision was taken to focus the survey on high-density areas where a diverse representation of schools could be found. As seen in Figure 2, four urban and peri-urban locations were chosen across Somalia: Bosasso and Garowe (Puntland), Kismayo (Jubbaland), and Mogadishu (Benadir Regional Administration). Information on each of the surveyed locations is presented below: Figure 2. Surveyed locations  Mogadishu is the capital and most populous city of Somalia with a diverse population of 1,280,939 inhabitants.44 It is predominantly urbanized with many informal settlements. As major economic hub, it attracts large numbers of IDPs from the regions in southern and central Somalia,45 as well as returning members of the diaspora. The security situation is poor, with violence carried out by Al-Shabaab on a regular basis. Although the extremist group officially withdrew from the city in 2011, they maintain strongholds in at least two neighborhoods. Despite this, the economy is booming, with elites focusing on the service sector, on trade (charcoal especially), telecommunications, and money transfer services. 46  Kismayo is the largest city of Jubbaland with 172,861 inhabitants. The city used to be a hub for Al-Shabaab, until a military offensive by Somali National Army (SNA) and troops of the African Union took the town in 2012.47 The regional government is not yet fully established, and the city suffers from a lack of law enforcement.48 Many more rural parts of Jubbaland are either under Al-Shabaab control or managed through the federal government. The MoECHE 44 All population estimates in this section are drawn for the 2014 UNFPA Population Estimation Survey. 45 Rift Valley Institute, 2017. Land Matters in Mogadishu: Settlement, ownership and displacement in a contested city. 46 World Bank. 2017. Somali Economic Update. Mobilizing Domestic Revenue to Rebuild Somalia. 47 AMISOM, 2013. Sector II Profile: Kismayo. 48 OCVP, 2015. Kismayo District Conflict and Security Assessment Report - 2015. The source of violence is tied to land disputes, revenge killings, power struggles and family disputes (ibid.). NON-STATE EDUCATION PROVIDERS Altai Consulting 26 June 2018 coordinates and implements education delivery in the parts of Jubbaland that it has access to. Kismayo’s economy relies on agricultural and charcoal production, livestock, and fishery.  Both Bosasso and Garowe are cities in Puntland, which claimed semi-autonomous status in 1998, breaking with the unstable transitional government in Somalia, before being integrated as a federal member state of the newly forming Somali federal state in 2014.49 Although, in comparison to southern Somalia, Puntland has experienced relative stability, the lack of law enforcement and governmental capacity remain a major issue. 50 Bosasso is the coastal commercial capital of Puntland, with an estimated population of 471,785 inhabitants. Its economy is based on livestock, small-scale farming, fishing, and imports and exports through the port.51 Garowe is the administrative capital of Puntland, with an estimated population of 138,929 people with a security apparatus that is described as sufficiently operational.52 Its economy concentrates on livestock, telecommunication services, khat, and small-scale enterprises. As the chosen sampling approach focuses on high-density areas, rural schools are not equally represented in the study. Nevertheless, the results of this research bear high relevance to the Somali education sector: government statistics indicate that 75% of students in Somalia are enrolled at urban schools.53 In addition, a total of 14 schools located in rural and peri-urban areas around the surveyed cities were also sampled to provide a comparison group. Findings on rural/urban differences are presented in the Focus Box 5 at the end of Section 3.5. 2.5. SCHOOL TYPOLOGY LACK OF ACCEPTED TYPOLOGY Schools in Somalia are managed and funded in complex ways that often involve more than one stakeholder, and multiple bodies can claim ownership over a school. 54 So far, this highly fragmented environment has failed to produce a school typology unanimously accepted by all key state and non- state stakeholders. The current policy framework does not offer a school typology reflective of the current education landscape and, even among stakeholders within the MoECHE, there is minimal consensus regarding basic distinctions between what classifies as a private or as a public school. 55 The distinction between public and private is key to the issue of school typology in Somalia. Government officials are likely to maximize the number of ‘public’ schools in the country by claiming schools housed in buildings constructed by the government before the civil war, in buildings or on the land owned by the government, or schools that receive any financial contribution from the government irrespective of its size. This tendency on behalf of the government to claim wide ownership of schools currently managed or funded by others is disputed by many, including umbrella associations and other actors, who feel they are the rightful owners of schools they manage and fund. 49 IGAD, 2014. Agreement between the Federal Government of Somalia and Puntland State of Somalia. 50 The source of violence is mainly tied to land disputes, youth-based violence, clan-related conflicts and gender-based violence (OCVP, 2015, Garowe District. Conflict and Security Assessment Report - 2015), as well as piracy that is feared to spike again after no incident recorded in recent years (Hassan, 2017. Somali regional anti-piracy chief says sacked over illegal fishing comments, Reuters; Craze, 2017. Only a strong state can put a stop to piracy in Somalia, The Wire). 51 Ministry of Planning and Statistics (Puntland State of Somalia), 2003. Puntland Facts and Figures. 52 OCVP, Garowe District. Conflict and Security Assessment Report, 2015 53 MoECHE, 2017. Education Sector Analysis 2018-2022. 54 Among many others, see MoECHE, 2017. Education Sector Analysis 2018-2022. 55 Interviews with government officials in Mogadishu, April 2018 NON-STATE EDUCATION PROVIDERS Altai Consulting 27 June 2018 The second interpretation of the word ‘public’ sometimes encountered among government officials is that of any school that is ‘open to public’, i.e. that is free and is accessible to all irrespective of their socio-economic (or other) background. Not only does this interpretation not account for who owns, manages or funds the school; there also appear to be relatively few such schools in the country. In addition to the government and umbrella (non-state) officials interviewed for this study, headmasters of the surveyed schools were asked what makes a school ‘public’. Their responses are shown in Figure 3. Figure 3. Reasons headmasters consider their school to be public (n = 43, multiple response) The building is owned by the government The government considers the school public The land is owned by the government The government supports teacher training The school follows the national curriculum The school is managed by the government The teacher salaries are paid by the government Infrastructure is paid by the government The education is free and accessible to all The teacher salaries are paid partly by the government 0 20 40 60 80 100 % of headmasters The lack of a clear consensus on what differentiates state and non-state schools, and more importantly, what rights and responsibilities building or land ownership, as well as running the day-to-day management implies, has led to substantial tensions between the state and non-state stakeholders. According to some of the interviewed key respondents, the current tendency of the government to seek involvement in the management of schools housed in government buildings but not otherwise supported, is symptomatic of the growing effort to return to the pre-war education system.56 Umbrella associations and other private entities feel that this new direction neglects the important roles and successes they have had in managing and supporting schools across the country. PROPOSED TYPOLOGY Although the initial typology proposed in this study’s inception report was useful in ensuring the diversity of the survey sample (see Annex 4 ‘Additional Information’), it did not adequately match how headmasters and key informants defined schools, especially with respect to the role of the government. Therefore, the initial typology has been amended to more accurately capture the Somali educational ecosystem in its current state. As management tends to be less disputed than ownership and more directly impacts the overall school environment and performance, management was taken as the basis for the school typology used in this report. There appears to be consensus among key stakeholders on what defines school 56 In some cases, this has resulted in the government insisting that schools change their current name back to the name the given by the government before the Somali civil war, despite private education providers having managed and funded the schools for 20 years. This decision was taken by the Ministers of Education and the DGs of the MoEs of all states without consulting non- state education providers. (Communiqué issued in April 2018, based on a MoU between FGS and FMS, signed in September 2016) NON-STATE EDUCATION PROVIDERS Altai Consulting 28 June 2018 management. Consequently, and for the purpose of this report, any actor that takes on the following responsibilities is defined as participating in management:57 1) Teaching and administrative staffing decisions 2) Decisions that dictate the day-to-day operations of the school 3) Decisions related to the school curriculum at the primary school level58 4) Determining the amount of school fees charged Four categories of schools based on school management have been proposed as shown in Table 3. ‘Public’ was dropped in favor of ‘government’ as surveyed headmasters showed varying levels of understanding of what is considered ‘public’ in Somalia .59 In order to differentiate between schools managed solely by the government and those managed by both state and non-state actors, a new category ‘mixed-management school’ was added. Additionally, a new category of 'umbrella schools' was added to represent the significant role of umbrella associations in directly managing schools. Finally, ‘independent schools’ consist of all schools that are managed independently of both the government and umbrella associations. Table 3. Proposed new school typology School type Description Schools managed exclusively by the government, as represented by the Government schools MoECHE or the state-level MoE. Schools managed by the government in cooperation with a non-state actor. Mixed-management schools Non-state actors include local community, religious groups, (I)NGOs, umbrella associations or private individuals. Umbrella schools Schools managed exclusively by an umbrella association. Schools managed by a non-state actor other than an umbrella association.60 Independent schools Non-state actors include local community, religious groups, (I)NGOs or private individuals. The following characteristics were assessed as not suitable for the purpose of developing a typology:  Ownership: as mentioned above, the definition of what constitutes ownership is contested between state and non-state stakeholders. See Section 3.3 for more details.  Funding: nearly all schools rely predominantly on tuition fees, including government schools.  Curriculum: according to the survey findings, most schools at least claim to be following the government-issued curriculum. This includes the IQSs that teach the secular curriculum in one shift, and Quranic studies in the other. Majority of schools include Islamic studies in their curriculum. Although the proposed typology is judged to be the best fit for the current education landscape from the available options, it has the following limitations: 57 According to interviews with government officials, CEC, and teachers 58 At secondary level, the curriculum is regulated by the MoECHE 59 In future policy documents, action plans, or research, it should not be assumed that the label 'public' is synonymous with 'government'. It is recommended that the label 'government school’ be used when referring to schools affiliated with the government, as it more closely reflects the Somali term 'dowladda'. 60 Note that independent schools can still be associated with umbrellas. For details, see Section 2.3. NON-STATE EDUCATION PROVIDERS Altai Consulting 29 June 2018  As specified above, the IQSs are not included as a separate category, as it would make them the only school distinguished by a module added to the curriculum. However, differences between IQS and secular schools are explored at the end of Section 3.5.  The typology does not differentiate between primary and secondary schools. Since most surveyed schools combine both primary and secondary levels this was not judged to present a serious obstacle.  The proposed typology presents, by definition, a simplified version of the current Somali educational landscape. NON-STATE EDUCATION PROVIDERS Altai Consulting 30 June 2018 3. FINDINGS 3.1. OVERVIEW OF SURVEY SAMPLE Key Takeaways ▪ Most schools are combined primaries and secondaries. ▪ Based on the proposed typology, Mogadishu is dominated by umbrella schools (59%), Kismayo by government schools (60%), and Bosasso by independent schools (91%). In Garowe, all types except umbrella schools are present, but mixed management schools are most common (35%). ▪ The median gender ratio is 156 boys per 100 girls, indicating a large gender gap. SCHOOL TYPES Most schools included in the sample offer both primary and secondary education (61%). The remainder are separate primaries (31%) and secondaries (8%). Nearly all are day schools, with only a single boarding school being surveyed, and most operate in two shifts (79%). A geographic breakdown of surveyed schools according to the proposed typology is shown in Figure 4. Figure 4. Regional distribution of survey sample according to new typology (n = 170) 80 70 60 Number of schools 50 Government schools 40 Mixed-management schools Umbrella schools 30 Independent schools 20 10 0 Bosasso Garowe Kismayo Mogadishu All schools  Overall, independent schools (39% of the sample) and umbrella schools (36%) are most common. Government schools (15%) and mixed-management schools (10%) make up the remainder of the sample.  Umbrella-managed schools were only found in Mogadishu where they accounted for 59% of the sample. However, schools that are not directly managed by the umbrellas but only associated with them, are found in every location.  Government schools are most common in Kismayo where they represent 60% of the sample.  Mixed-management schools are most common in Garowe (35%), another state-level administrative center.  Bosasso, the fast-growing commercial center of Puntland that is nevertheless far removed from the state capital, showed the highest proportion of independent schools (91%). NON-STATE EDUCATION PROVIDERS Altai Consulting 31 June 2018 The regional differences shown in Figure 4 appear to be genuine and not driven by the sampling approach adopted for this research. As described in the chapter on methodology, verification was conducted in every location where schools were either physically visited and mapped (Mogadishu) or contacted (remaining locations). All headmasters contacted during verification were asked to list other schools in their vicinity to ensure completeness of the sampling frame, and to avoid the risk of the Education Management Information System (EMIS) or umbrella-issued lists of schools dominating the sample frame and skewing the sample composition. STUDENT DEMOGRAPHICS The median school size across all school types is 391 students. The surveyed schools are all relatively similar in size, with most school types having between 300 and 400 students, as displayed in Table 4 below. Umbrella schools tend to be the largest in size, and government schools the smallest. However, schools are often overcrowded and do not have the facilities to accommodate the large number of students. This is discussed later in in Section 3.5 ‘Education Quality’. Table 4. School size and student gender ratio by school type Overall statistics Median by school type Mixed- Government Umbrella Independent Mean Minimum Maximum Median management schools schools schools schools Number of students (n = 162) 541 56 2905 391 301 358 400 380 Males per 100 females (n = 149) 198 0 1850 156 133 167 184 138 Also seen in Table 4, there is a large gender gap at all school types with a median gender ratio of 156 male students per 100 females. All schools in the sample are co-ed with the exception of two girls’ schools. Geographically, the gender gap is smallest in Garowe (with a median of gender ratio of 117:100), and largest in Mogadishu (184:100). In both Bosasso and Kismayo the ratio is 135:100. The reasons for and implications of this gender gap are discussed later in Section 3.6 ‘Accessibility & Inclusivity’. At 90% of schools, the most common professions among the parents are day laborer and vendor. These professions are common among parents at all school types. The remainder of schools are mostly attended by children of farmers (2%), government staff (1%), NGO staff (1%), or a mixture of professions (6%). 3.2. SCHOOL MANAGEMENT Key Takeaways ▪ Umbrella associations, local communities, and the government are the most common managers of schools. ▪ The proposed typology reflects regional differences in management. Since the proposed typology is based on management, it reflects the management structure of surveyed schools (Figure 5). NON-STATE EDUCATION PROVIDERS Altai Consulting 32 June 2018 Figure 5. Management by school type (n = 170, multiple response) 200 180 160 Number of responses 140 Umbrella 120 Private Individual(s) 100 Religious organisation 80 (I)NGO 60 Local community 40 Government 20 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools By definition government and umbrella schools are managed exclusively by their eponymous stakeholder. Mixed-management schools are managed jointly by the government and the local community (in 47% of the cases), the government and an umbrella (47%) or the government, the local community and an (I)NGO (6%). Independent schools are mostly managed by the local community and (I)NGOs. This may be the result of individual community members either founding a school directly or setting up an NGO to do so. The data show significant differences in school management by location :61 in Mogadishu the majority of schools are managed by an umbrella association (59%); in Bosasso by the local community (77%); in Garowe by the local community (35%) or the government (29%); in Kismayo by the government (60%). Schools managed by the government are less prevalent in Bosasso than in other survey locations – the only one that is not a regional administration center. 88% of (I)NGO-managed schools are in Mogadishu, representing 20% of the city’s schools, possibly because Mogadishu is more easily accessible to international stakeholders. Only two schools are managed by religious groups (both of which are in Bosasso), which instead concentrate on Quranic schools that do not have a secular curriculum and are not part of the formal education system. Overall, the umbrella associations are involved in the management of the greatest number of schools (41%), but schools managed exclusively by umbrella associations are only located in Mogadishu. As discussed in Section 2.3, other schools can however associate themselves with an umbrella while keeping their management arrangements separate. Such associated schools comprise 66% of the overall sample, across all school types and locations. 3.3. SCHOOL OWNERSHIP Key Takeaways 61 X2 (33, N = 170) = 210.11, p < 0.001 NON-STATE EDUCATION PROVIDERS Altai Consulting 33 June 2018 ▪ Ownership of schools is contested between state and non-state stakeholders. The government is likely to claim ownership based on construction or ownership of the school building, even if the school is privately operated. ▪ Within the survey sample, private individuals are the most common school owners (ownership stake in 65% of schools) while 20% of schools have multiple owners. ▪ In 53% of cases, different actors own and manage the school. This is especially common for umbrella schools, which are usually privately owned. DEFINITION OF SCHOOL OWNERSHIP IN THE CURRENT SOMALI CONTEXT The definition of school ownership remains contested in the education sector in Somalia. While most interviewed stakeholders agreed that an active role in school management and a contribution to the school funding give an actor right to claim at least partial ownership, there is a tendency among the government officials to also emphasize the relevance of ownership of the school building or the land upon which it is built. The umbrella representatives, on the other hand, tend to speak less of infrastructure and more of the design and regulation of the school, as well as of the school brand, all of which are areas through which umbrella associations exercise influence among their own schools. As explained in the previous section, this can lead to tensions between the government and non- government education stakeholders. To gain further insight into the issue of ownership, headmasters of surveyed schools were asked to identify who owns their school and on what grounds. Figure 6 shows that headmasters prioritized cost-coverage over building and land ownership when assigning school ownership. 62 Figure 6. Features of school ownership according to headmasters (n = 170, multiple response) School owner must own the land School owner must own the building School owner must pay teachers' salaries School owner must pay for running costs to maintain the infrastructure 0 10 20 30 40 50 60 70 80 90 100 % of responses OVERVIEW OF OWNERSHIP  Most schools are owned by private individuals (65%), according to the interviewed headmasters.  50% of schools are owned exclusively by private individuals, and an additional 15% include private individuals as part owners in combination with other owners.  20% of schools have multiple owners. The three most common combinations of hybrid ownership found through the quantitative survey are: 62 Management was not offered as one of the responses to this close-ended question posed to the headmasters, however interviews with stakeholders indicate that an active role in management also plays a role in assigning ownership. NON-STATE EDUCATION PROVIDERS Altai Consulting 34 June 2018 - Community and private individuals - Private individuals and umbrellas - Government and community  Many schools are not managed and owned by the same actor(s). This is the case for just over half of the schools surveyed (53%). In most of these cases (64%) the school is managed by an umbrella association and owned by private individuals. OWNERSHIP BY SCHOOL TYPE Figure 7 shows the breakdown of school ownership, as identified by the headmasters, by school type based on the typology presented in the previous chapter. As explained in the previous section, the proposed school typology is based on management rather than ownership, in part due to the lack of consensus on the definition of ownership. As headmasters were given the opportunity to identify multiple owners, the figure reflects the overall frequency with which different owner categories were identified and a number of responses that is therefore larger than the number of surveyed schools. Figure 7. School ownership by school type (n = 170, multiple response) 250 200 Umbrella 150 Private individual(s) Religious organisation 100 (I)NGO Local community 50 Government 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools  All schools managed by the government (and therefore defined as ‘government schools’) were also described as owned by the government, most of them exclusively (92%). The remaining 8% were identified as partially owned by the local community.  The ownership structure of mixed-management schools most frequently includes the government (in 47% of surveyed schools) and/or private individuals (also 47%). Only 12% of mixed-management schools have a hybrid ownership structure.  The majority of umbrella schools were described as owned by private individuals exclusively (85%). Although the operations of the umbrella schools tend to be managed directly by the association, including the appointment of key staff, only a few were described as being owned by the umbrella (3%). For more on how umbrellas are structured and operate, see Section 2.3.  Independent schools are mostly owned by private individuals exclusively (37%), the local community exclusively (25%), or a hybrid combination of both (25%). A very limited number of schools were described as independently managed but with partial government ownership, as a result of the government ownership of the school building (1.8%). NON-STATE EDUCATION PROVIDERS Altai Consulting 35 June 2018 OWNERSHIP BY LOCATION Figure 8 shows the breakdown of school ownership by survey location. Similar to the breakdown of ownership by school type above, Figure 8 is based on a multiple response question and therefore reflects the overall frequency with which the headmasters have identified a particular type of ownership rather than the total number of schools surveyed. Figure 8. School ownership by location (n = 170, multiple response) 120 100 Umbrella Private individual(s) 80 Number of schools Religious group 60 (I)NGO 40 Local community Government 20 0 Bosasso Garowe Mogadishu Kismayo  In Mogadishu, the large majority of schools are reportedly owned by private individuals exclusively (85%); the remaining schools have a hybrid ownership.  In Bosasso, just under half of the schools have a hybrid ownership comprised of both the local community and private individuals. An additional 31% are owned exclusively by the community; the remaining schools have hybrid ownership of government and community, community, private individuals and (I)NGOs, and community and (I)NGOs.  In Garowe, most schools are owned by the government exclusively (59%); 24% are owned by the local community exclusively; the remaining schools have hybrid ownership. In Kismayo, most schools are owned exclusively by the government (53%). 20% are owned exclusively by the local community, and 20% by private individuals. The remaining schools have hybrid ownerships by the government and the community (7%). 3.4. SCHOOL FUNDING Key Takeaways ▪ School fees are the most important source of funding for schools in Somalia. Within the survey sample, 89% of schools rely on school fees as a source of funding. ▪ Government funding is limited, and generally restricted to government-managed schools. ▪ Most of the MoECHE’s funding is provided by the international community. NON-STATE EDUCATION PROVIDERS Altai Consulting 36 June 2018 ▪ Revenue from school fees is often insufficient to cover expenses, the largest of which is payroll, meaning that many schools operate at a deficit. MAJOR FUNDING STREAMS According to the MoECHE, school fees account for 70% of the overall education funding to primary and secondary schools in Somalia, while government funding accounts for only 4%. Additionally, (I)NGOs are said to contribute 10 – 20%, communities 5%, and the diaspora another 5%. Five major funding streams supporting primary and secondary schools in Somalia were identified in the course of this research:  Student fees  Government funds  (I)NGOs  Local community  Bilateral/multilateral funding Less important sources of funding include umbrella associations, private individuals, and religious groups. Sources of funding are disaggregated by school type in Table 5 and by location in Table 6. Table 5. Percentage of schools receiving funding from each source, by school type (n = 170) Table 6. Percentage of schools receiving funding from each source, by location (n = 170) School fees are the most common source of funding for schools in Somalia, with 89% of surveyed schools being at least partially funded by fees. All school types may charge school fees, although NON-STATE EDUCATION PROVIDERS Altai Consulting 37 June 2018 government-managed schools are less likely to do so: 52% of government schools charge fees, compared to 94% of other schools. Although it is unclear why certain government schools charge fees and others do not, there are indications from non-state actors that this is related to the demographics of the school’s location. Fees tend to be lower in poorer neighborhoods and may be reduced during times of economic hardship, conflict or drought. The average annual volume of funding from school fees is approximately USD 56,600 per school at the surveyed schools63. This figure ranges from nothing at schools that do not charge fees to USD 755,300 at a large independent school in Garowe. Since fees are charged per student, the amount of funding from fees varies as a function of school size. Schools are highly dependent on school fees for funding, and therefore dependent on the parents’ financial capacity as well. Only a relatively small number of surveyed schools receive financial support from the government (13%), with only 4% receiving funding exclusively from the government. Government support tends to cover or contribute towards the school’s running costs, teacher salaries, and equipment. However, a large portion of the government funds do not reach the individual schools directly, as they are used to cover the MoECHE’s operational costs, school supervision, certificate exams, and pay teacher incentives at government schools. The government’s inability to provide a stable and significant source of funding limits its influence over the schools, although influence may be exercised in other ways such as regulation. Most (90%) of the MoECHE funding comes from the international community64. The European Union (EU) is the largest donor, spending USD 71.5 million for the period 2014-2020. EU funding supports the strengthening of public systems for education service delivery, mainly in form of capacity development of the education sector, the development of curriculum framework, and the implementation of centralized exams.65 The EU funds are not disbursed to the individual schools.66 According to the interviewed government officials, (I)NGOs fund mainly the construction and maintenance of infrastructure, make contributions towards teacher salaries, and fund emergency and special needs. In spite of their status as key stakeholders in the education sector, umbrella associations do not offer substantial funding towards school operating budgets. They may however provide scholarships and discounted fees for certain students. This implies that their status is based on their roles as important providers and regulators of education rather than as funders of schools. Also of note, the average surveyed school relies on 1.3 funding sources. Lack of diversity in funding sources and dependence on parents makes schools more vulnerable to disruptions in funding streams and to economic shocks in general. Government schools are more likely to have diverse funding sources: 76% of government schools receive funds from sources other than fees, while only 47% of mixed-management schools and 33% of independent schools do. This shows that diversity in ownership and management does not necessarily translate into diversity in funding. 63 The overall projected yearly volume of student fees is USD 97,723,692 for primary, and USD 24,475,500 for secondary schools, a total of USD 122,199,192. This is calculated using enrolment figures of primary and secondary students published in the Education Sector Analysis 2018-2022 (tables 55 and 84) and multiplying them by the average school fee charged to primary and respectively secondary students of the 170 schools surveyed for this study. 64 Interviews with government stakeholders in Mogadishu, April 2018 65 European Union External Action, press release, 03.06.2016, URL: https://eeas.europa.eu/sites/eeas/files/20160603_en.pdf [18.5.2017] 66 UNICEF’s education cluster received over USD 5.8 million in 2017 for projects in Somalia (UNICEF, Somalia Humanitarian Situation Report, 16.11.2017, URL: https://www.unicef.org/somalia/SOM_sitrep18_15Nov2017.pdf). NON-STATE EDUCATION PROVIDERS Altai Consulting 38 June 2018 Additionally, data was collected on sources of funding for the teacher payrolls, the main expense for schools. Figure 9 presents the relative importance of each source of funding by school type. Figure 9. Sources of funding for teacher salaries by school type (n = 170, average per school type) 100 90 % of funding for teacher salaries 80 Umbrella 70 Private Individual(s) 60 Religious organisation 50 Local NGO 40 Local community 30 INGO 20 Goverment 10 School fees 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools  School fees cover at least 67% of the teacher payroll for each school type, an average of 87% across all school types.  On average, school fees contribute the largest amount to teacher payroll for umbrella schools (95%), and the smallest amount of the teacher payroll for government schools (67%).  Local communities contribute most at the government-managed schools (6% of the teacher payroll), mixed-management schools (4%) and very little in the case of umbrella schools (1%).  (I)NGO support also appears to be concentrated on government-managed schools (8% of teacher payroll) and mixed-management schools (7%). Independent schools also receive support from NGOs with 4% of teacher payrolls being funded by INGOs and 3% by local NGOs.  Government schools are the only school type benefiting from government funding of the teacher payroll (19%). Additionally, most government schools (60% across all locations)67 report receiving any form of funding from the government. Although it is unclear how government funding is allocated, government schools that charge fees are less likely to receive funding than those that do not. 68 According to respondents, government funding is irregular, leaving schools vulnerable to funding shortages. EXPENDITURE AND FUNDING GAPS Figure 10 presents the expenses, disaggregated by school type, as reported by the headmasters. 67 Government schools in all locations reported government funding, with the exception of the one government school in Bosasso. The small number of government schools within the overall sample should also be noted when considering these findings. 68 X2 (1, N = 25) = 9.64, p < 0.01 NON-STATE EDUCATION PROVIDERS Altai Consulting 39 June 2018 Figure 10. Main expenses by school type69 (n = 169) 100% 90% Other recurring expenses 80% Student meals % of total expenditure 70% 60% Registration/approval fees 50% Tax 40% Equipment and learning material 30% Maintenance 20% Utilities 10% Payroll 0% Mixed- Government managementUmbrella schools Independent schools schools schools All schools The main cost driver is by far the payroll, making up at least 50% of the expenses across all school types, followed by recurring expenses, such as rent, learning materials, maintenance expenses and utilities. At the school level, funding for adequate infrastructure is often lacking. Respondents notably identified major funding gaps in terms of infrastructure, with many schools lacking sufficient classrooms, chairs, and tables as well as more advanced facilities such as a library or a laboratory. Teachers are paid poorly, if at all,70 and funding for adequate and sufficient learning materials, especially textbooks, is often lacking. For details, please refer to Section 3.5. Significant funding gaps exist across all school types since the main revenue source, school fees, is insufficient to cover all expenses. Figure 11 contrasts the annual revenue from school fees against the annual overall expenditure, as reported by the headmasters, and thus presents the gap that funding other than fees must fill in order to avoid a budgetary deficit.71 On average, school fees cover 74% of school expenditures. On average, schools spend USD 203 per child annually. Mixed management schools spend the most at USD 302, followed by government and independent schools at USD 228 and 224, respectively, while umbrella schools spend the least (USD 140). 69 These expenses exclude rent for the school building, if applicable. 70 Interviews with government officials and umbrella associations, as well as discussions with teachers 71 The revenue from school fees was determined calculating the average primary school fee using information given by the headmasters, then multiplying the average by the total number of primary students per school as reported by the headmasters and multiplying the average secondary school fee by the total number of secondary students per school as reported by the headmasters, and then averaging the result by school type. The expenditure is as reported by headmasters, then averaged by school type. NON-STATE EDUCATION PROVIDERS Altai Consulting 40 June 2018 Figure 11. Average reported annual funding from fees and expenditure by school type (n = 134) 100000 90000 80000 70000 Annual expenditure 60000 USD/year Annual revenue from school 50000 fees 40000 Annual deficit 30000 20000 10000 0 Mixed- Government management schools schools schools Umbrella All schools Independent schools In order to overcome budget shortfalls, schools leverage community support, reach out to local and international NGOs working in education, and contact individual businessmen. Some schools also establish their own enterprises to help with funding, for example selling food at canteens, or providing printing and photocopying services. One of the schools under the FENPS umbrella generates additional funding through beekeeping, selling honey to school neighbors, and using the excess money for maintenance of the school's infrastructure. Since this activity proved successful, FENPS plans to extend the business model to more schools in the future. Driven by the chronic lack of funding, REOs and CECs spend much of their time fundraising instead of fulfilling their regular functions and duties, 72 weakening accountability mechanisms. 3.5. EDUCATION QUALITY Key Takeaways ▪ Quality index score-based analysis did not reveal clear differences in quality among school types, although a regression model showed independent schools to be more likely to produce positive educational outcomes. ▪ Teacher training and qualifications are not regulated by the government , with 12% of teachers who have not pursued post-secondary education. ▪ In addition, teacher performance is affected by low and inconsistent payment. ▪ Although steps have been taken to standardize educational content, schools continue to use a mix of curricula and teach in several different languages. EVALUATING QUALITY PERFORMANCE The overall approach taken to evaluate education quality within the scope of this report is centered around three main criteria:  Key staff qualifications, availability and performance 72 Interviews with government officials NON-STATE EDUCATION PROVIDERS Altai Consulting 41 June 2018  Educational content and outcomes  Availability and condition of physical infrastructure and learning materials Indicators from each of the components of quality are examined in detail in the following sub-sections. Additionally, select indicators73 from each category were used to generate a simple index of overall quality. For each indicator, schools were awarded between 1 and 3 points based on relative performance. The final index is scored out of 100 to facilitate comparisons between school types. Although the results of this analysis did not reveal any clear differences between school types (Figure 12), there nonetheless exists a wide range in quality between individual schools as index scores ranged from 44 to 89. Figure 12. Average education quality index score by school type (n = 170) 100 90 80 Quality Index ( /100) 70 60 50 40 30 20 10 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools To assess the factors affecting educational outcomes, outcome indicators (such as the dropout rate) were regressed against a set of indicators of quality and several control variables. School type, location, and gender ratio were among the control factors while the indicators of quality included the student- teacher ratio, presence of a library, and teacher salaries. Using stepwise regression74, models were refined by dropping variables that did not significantly explain differences in outcomes. In several models, location and school type were found to be significant determinants of educational outcomes. Based on this analysis, outcomes appear to be best at independent schools. Regionally, there is limited evidence that schools in Puntland produce better outcomes. A larger sample size would be helpful in confirming the significance of these findings. Further details on the results of this analysis are available in Annex 4 ‘Additional Information’. KEY STAFF QUALIFICATIONS, AVAILABILITY AND PERFORMANCE Information on Headmasters Nearly all headmasters interviewed completed some form of post-secondary education (98%), including 83% with a university degree. Many also received additional training to fulfill their administrative roles (61%). This training commonly focused on management, administration, and leadership skills. Finally, the profession appears to be male-dominated as only 3 of the 170 headmasters interviewed during the survey were women. 73 The list of indicators used can be found in Annex 4 ‘Additional Information’. 74 Additional details on this analysis, including variables used in the model, can be found in Annex 2 ‘Methodology’. NON-STATE EDUCATION PROVIDERS Altai Consulting 42 June 2018 Teacher Availability and Characteristics Across all schools the median student-teacher ratio is 28:1 and the average of 37:1 is in line with the FGS’s national estimate of 33:175 (Table 7). Although the median is fairly low and in line with what is recommended for emergency recovery contexts;76 it should be approached with caution. Parents, students, and teachers often expressed that there are too few teachers and in some cases the student-teacher ratio reaches 250:1. Furthermore, despite the number of teachers being significantly related to the number of students at the school,77 the number of students increases faster than the number of teachers. This means that larger schools are more likely to be understaffed .78 In addition to affecting the quality of education, this creates a more challenging working environment for teachers themselves. Table 7. Availability and gender ratio of teachers Overall statistics Median by school type Mixed- Government Umbrella Independent Mean Min Max Median management schools schools schools schools Number of teachers (n 17 4 101 13 15 15 12 12 = 169) Student- teacher 37 10 252 28 32 24 31 28 ratio (n = 153) Males per 100 females 677 117 5900 471 370 358 400 633 (n = 116) As with the student body, there is also a gender disparity among staff. On average, schools employ 6.8 times more male teachers than females. The imbalance is greater in secondary schools since older students may show less respect to female teachers,79 causing women to gravitate towards primary schools. Conversely, men may find it more difficult to handle young children so are more likely to opt to teach secondary. According to a teacher trainer at the SNU: "Many female students at our university decide to take the path of primary school teaching, even though it is less well remunerated. This is because female teachers find it hard to get secondary students to listen to them and respect them." Female teachers can be important role models for girls and inspire them to complete their studies.80 The scarcity of female teachers in Somalia could therefore impact educational outcomes for girls. 75 MoECHE, 2017, Education Sector Analysis 2018-2022 76 The Inter-Agency Network for Education in Emergencies (2010) suggests a ratio of 30-40 students per teacher to be appropriate for chronic crisis or recovery contexts. https://www.unicef.org/eapro/Minimum_Standards_English_2010.pdf 77 Linear regression (F1, 160 = 38.58, p < 0.001, R2 = 0.19) 78 The coefficient of correlation between the student-teacher ratio and the number of students is 0.70 79 Interview with a member of the SNU teacher training department 80 UNESCO. http://unesdoc.unesco.org/images/0014/001459/145990e.pdf NON-STATE EDUCATION PROVIDERS Altai Consulting 43 June 2018 Teacher Training and Qualifications Teachers in Somalia may receive pre-service and in-service training. Pre-service training can cover lesson planning, teaching methods, and classroom management, as well as helping teachers develop knowledge of subject matter. In-service training allows teachers to receive feedback on their performance and advance their professional development. Figure 13 shows the pre-service qualifications of teachers across school types. Overall, 12% of teachers have not pursued any post-secondary studies. On average, mixed-management schools have the least qualified teachers: 30% have no post-secondary education and only 36% have a bachelor’s degree. Independent schools have the highest qualified teachers with 75% of teachers having the highest qualification, a bachelor’s degree, a further 15% with a college degree, and less than 10% with no post-secondary education. Figure 13. Teacher qualifications by school type (n = 130) 100% 90% 80% 70% % of teachers 60% Not completed secondary 50% Secondary school 40% College degree 30% Bachelor's degree 20% 10% 0% Government Mixed- Umbrella Independent All schools schools management schools schools schools As exemplified by the employment of unqualified teachers , the qualifications necessary to work as a teacher are not regulated by the government.81 In theory, the Curriculum and Quality Assurance Department within the MoECHE can withdraw teaching licenses received from training institutions but, in practice, this is not done due to the shortage of teachers and a lack of funding needed for enforcement. Additionally, no unified teacher training curriculum exists at pre-service training institutions.82 Nevertheless, parents and students perceive quality of education to be linked to the qualifications of teachers, as described by a student in Bosasso: “This school delivers good education because the teachers of the school are qualified and very hardworking - most of them have master’s degrees and even those with only a diploma are capable of teaching the students well.” A parent in Garowe stressed the importance of subject matter expertise: “Knowledge of subject matter is important. This may seem obvious, but it is sometimes overlooked. The teachers of this school have incredible knowledge of and enthusiasm for the 81 Interviews with MoECHE and SNU representatives 82 MoU between Faculty of Education at the University of Helsinki and SNU in April 2017 to re-develop and update the current programme for teacher training, including developing a shared curriculum, but so far not been put forward. NON-STATE EDUCATION PROVIDERS Altai Consulting 44 June 2018 subject matter they are teaching. They are prepared to answer questions and keep the material interesting for the students.” In-service training is offered irregularly at some schools during school holidays, depending on funds available; about half of the schools reported that an in-service training session had taken place in the past year (54% overall). These appear to be more common in Puntland than in the southern FMSs. As described by a teacher in Bosasso: “During the holidays, the Ministry of Education calls upon teachers from various subjects and offers trainings that are beneficial to them and to the students. These include leadership skills, class management, practicals, and written skills.” There is a lack of sufficient and adequately resourced pre-service and in-service teacher training institutions across the federal states. Limited institutional capacity contributes to a shortage of teachers exemplified by the high student-teacher ratios recorded at many schools. According to informants, top-performing teaching applicants tend to be recruited to the secondary level, leaving the remainder to teach primary. The federal government is building new training centers that are to be managed and supervised by the existing ones, rather than increasing the capacity of existing facilities.83 Although this option may be costlier, it has the potential of improving access to training across the country. Teacher Recruitment, Contracting, and Remuneration Most schools have an internally defined teacher recruitment process during which the applicant’s qualifications and knowledge are considered, according to key informants. The degree of formality of the process varies from a simple review of applicants’ CVs to standardized testing. Additionally, some schools require proof that the applicant has received a teaching certification. At some schools, CECs and parents are involved in the recruitment process – usually if there is a practical assessment of teaching capability. The recruitment process is more regulated in government schools . At schools where the government is not a primary manager, owners and managers hire teachers at their own discretion. Only a few schools indicated that they do not have a formal recruitment process, instead accepting any willing applicant. As previously discussed, the selection process at teacher training institutions results in the less qualified candidates teaching primary and the better-qualified candidates becoming secondary school teachers. Focus Box 3 below presents the perspectives of teachers on the recruitment process. Focus Box 3. The path to becoming a teacher in Somalia The path to becoming a teacher in Somalia Although there is no clear path to becoming a teacher in Somalia, it may begin with form of post- secondary training. Teacher training programs are still being developed in the southern and central states but are better established in Puntland: “Like teachers throughout the country, teachers in Puntland are licensed by the institution they graduated from. Most receive training and professional development through Garowe Technical College known as GTECH and East Africa University in Bosasso. ” – Former Puntland government official 83 Interview with a representative of the SNU NON-STATE EDUCATION PROVIDERS Altai Consulting 45 June 2018 The next step is to find a teaching job. Some applicants are fortunate and use their connections to land a job. “I graduated from this school, and one day management called me and told me that they wanted to hire me as teacher.” – Teacher in Mogadishu For most, however, becoming a teacher requires applying to a vacancy and successfully completing a selection process. “When the school needs teachers, they advertise on the radio. Teachers send their CV to the management office and are interviewed. If they have enough experience, they are hired. The school managers mostly like to hire Lafole graduates.84” – Teacher in Mogadishu In addition to reviewing candidates’ experience and interviews, the recruitment process often includes a practical assessment of teaching skills. “During my recruitment, I was taken to teach several classes of the same level. After the lesson students commented on my techniques, and finally I got employed.” – Teacher in Bosasso In Puntland, the government may be involved in the hiring process. “I was interviewed by a panel from the Ministry during my hiring process.” – Teacher in Garowe This is also the case for public schools in Kismayo. “Advertisements, shortlisting, and interviews are done by a commission of civil servants and the Ministry of Education, but only for public schools.” – Teacher in Kismayo “[…] for private schools, every school has its own way of employing teachers - you submit your CV to school managers and they decide whether or not to call you for an interview. ” – Teacher in Kismayo Figures 14 and 15 show the average teacher salary by school type. Teachers who have a teaching certificate or diploma earn between USD 50 and 350 a month. Teacher salaries vary significantly both by teacher qualification and school type. 85 Teachers with diplomas or teaching certificates tend to earn more than teachers without a certificate. The average salary of teachers at mixed- management schools is significantly higher compared to other school types at both the primary and secondary level. In group discussions, the teachers from government schools indicated that they earned less than teachers working at private schools. 84 Lafole University is a well-respected teacher training institute in Afgoye District. 85 Two-way ANOVA: Main effect of school type (F3,430 = 12.23, p < 0.001); Main effect of teacher qualification (F3, 430 = 11.74, p < 0.05); post-hoc Tukey test (p < 0.05 for all pairwise comparisons involving mixed management schools) NON-STATE EDUCATION PROVIDERS Altai Consulting 46 June 2018 Figure 14. Average monthly salary of primary teachers by school type (n = 123) 350 300 250 USD/month 200 No official teaching certificate Teaching certificate 150 Diploma 100 Part-time 50 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools Figure 15. Average monthly salary of secondary teachers by school type (n = 96) 400 350 300 250 USD/month No official teaching certificate 200 Teaching certificate 150 Diploma 100 Part-time 50 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools Teacher salaries are often referred to as 'incentives' since they are well below subsistence level. The interviews also indicated that the salary payments to teachers are irregular and that some teachers go for several months without payment. In an attempt to increase compensation, some schools offer monetary incentives to high-performing and motivated teachers. School management may also assist teachers in times of need as a repayment for their loyalty. 86 Additionally, most teachers at the schools surveyed were on long-term contracts (only 22% of teachers were on short-term contracts). This does not vary significantly between different school types. Nonetheless, low teacher salaries affect the appeal of the profession. As seen in Table 8, remuneration is comparable to unskilled professions such as a day laborer or market vendor. In addition, the unpaid years required to become a qualified teacher present a major opportunity cost to young Somalis and therefore greatly diminish the allure of teaching. 86 Discussions with teachers NON-STATE EDUCATION PROVIDERS Altai Consulting 47 June 2018 Table 8. Remuneration of teacher compared to other professions87 Position Monthly salary Teacher with diploma USD 50-350 Medical doctor USD 1500-2500 Civil servant (higher level) USD 500-2000 Civil servant (lower level) USD 150-450 Nurse USD 250-450 Day laborer USD 210-600 Vegetable vendor USD 100-450 Livestock vendor USD 300-400 Teacher Performance Low pay is a challenge for working teachers and affects morale, motivation, and performance. Consequently, teacher absenteeism is a widespread issue in Somalia.88 One responsibility of CECs, if they are in place, is to monitor teacher absenteeism. However, since teachers are paid poorly, if at all, the school management has little leverage over the teachers' presence other than appealing to their morale. According to the headmasters, most schools (91%) provide some sort of supervision for teachers. This can entail classroom observations where teachers are assessed and given feedback on their performance. Teachers can also be provided guidance on class management, lesson planning, and teaching methods. However, in group discussions many teachers claimed they do not receive supervision or guidance. In Puntland, graduates from the Garowe Teacher Education College (GTEC) are visited to provide guidance when they start their work. 89 Focus Box 4 presents teachers' perspectives on the challenges of being a teacher in Somalia. Focus Box 4. Life as a teacher in Somalia Life as a teacher in Somalia Across Somalia, teaching is a challenging profession. The education system is still recovering from the effects of the civil war and remains highly under resourced, creating difficult working conditions for teachers: “The lack of teaching materials, textbooks, and mathematical instruments, as well as the school environment and working conditions, diminish teachers’ effectiveness, their growth and development, and efforts to improve student achievement.” – Teacher in Bosasso In addition to having to work with insufficient resources, attitudes towards the value of education can complicate the jobs of teachers. Students may be unmotivated, either due to general disinterest or because of problems at home, and parents are often indifferent to their child’s education. “If parents do not hold high aspirations for their child’s educational attainment, their child will not see the purpose of staying in or doing well in school. If parents are engaged early in the 87 The remuneration of professions was assessed qualitatively by asking several trusted individuals in different regions for their estimation. 88 GPE Education Programme Document 2014-2016, Somalia Federal Republic, South Central Somalia, 2014 89 Interview with a representative of GTEC NON-STATE EDUCATION PROVIDERS Altai Consulting 48 June 2018 child’s educational career, the child is more like ly to be successful in school.” – Teacher in Kismayo Finally, teachers in Somalia tend to be poorly remunerated – arguably the greatest challenge faced by the profession. Low salaries take a toll on motivation, contribute to absenteeism, and reduce the appeal of becoming teacher – all likely having profound effects on teaching quality. “The compensation teachers receive is appalling. […] Teaching comes with no additional benefits, unlike most careers. Teachers are NOT entitled to sick days and may go for three months without pay.” – Teacher in Mogadishu EDUCATIONAL CONTENT AND OUTCOMES Educational Content Nearly all headmasters (99%) indicated that they follow the unified national curriculum. However, the situation as presented by key informants was less clear, as they indicated that fewer secondary schools tend to follow the national curriculum , whereas primary schools use a mix of curricula including the Kenyan curriculum, the pre-war Somali curriculum, Egyptian curriculum, and combinations thereof. A teacher in Bosasso commented on the curriculum: “The surrounding schools are taught in the same way though there are differences in the curriculum. For example, some use the Arabic curriculum, others the Somali curriculum developed by UNESCO, and some use the Kenyan Curriculum. These differences in curriculum leave an impact on students’ educations since they cannot transfer from one school to another.” Nonetheless, schools consistently offered the same set of core subjects ,90 defined as being offered at more than 80% of schools overall.  Core primary subjects: Arabic, math, English, Somali, Islamic Studies, and science.  Core secondary subjects: math, biology, chemistry, Arabic, history and geography, physics, Somali, and Islamic Studies. Although most schools have computers, the small number of schools teaching information technology (IT) indicates that they are not being used for educational purposes. Only 17% of primary schools and 37% of secondary schools offer IT courses. Most schools offer Islamic studies, although the method of delivery is different than at the IQSs. Whereas formal schools offer Islamic studies as a regular subject, IQSs teach Islam in a more intensive manner. Often, half the school day is dedicated to studying the Quran while the other is for regular courses. Parents reported that access to religious education is important regardless of whether it was at a religious or formal school. Other less commonly taught subjects at the primary level include environmental education (6% of primaries/25% of secondaries), physical education (3%/16%), and business studies (3%/44%). There are concerns that the curriculum does not focus enough on employable skills. 91 Similarly, several stakeholders stated that more attention should be given to technical and vocational training. In the pre-colonial era, Arabic was most commonly used as a language of instruction. During the colonialization of Somalia, both English and Italian were used. English was then endorsed after 90 According to a recent report the new national curriculum will be implemented in August 2018, comprising 7 core subjects: mathematics, science, social studies, Islamic studies, Arabic, English and Somali (UNSOM, 2018). 91 Interview with representative from FPENS NON-STATE EDUCATION PROVIDERS Altai Consulting 49 June 2018 the war in order to strengthen the English skills and internationality of the students. Finally, the liberation of Somalia led to the desire to move towards Somali as the main language of instruction, especially for the primary school level. This is significant because, according to UNESCO, learning in one’s mother tongue is more effective, meaning that receiving education in English may compromise knowledge transfer.92 Starting in August 2018, the government aims to regulate and standardize the language of instruction in classrooms, prescribing Somali as the language of instruction in primary while secondary courses will be taught in English and Arabic.93 Currently, primary courses are taught in a mixture of Somali and English, depending on the school and the teacher's English skills. In secondary, courses are often taught in English, especially science subjects (math, biology, chemistry, physics). The exceptions are history and geography, and Islamic studies, which are usually taught in Arabic. Additionally, Somali textbooks are being developed by a government-led working group of stakeholders. However, most textbooks are still in English, and some are in Arabic; the language of the textbook and the language of instruction are not always the same. Often, the textbooks are in English and the teachers explain the content in Somali. Educational outcomes The rate of students passing their final exams, both primary and secondary, was reported as fairly high by the school headmasters at 88% across all school types. This figure is lowest for mixed- management schools (78%) and highest for independent schools (91%). However, of 27,600 secondary students taking the government exam in 2018, only 97 were from government schools according to an interviewed MoECHE official.94 A government official working for MoECHE explained this as a consequence of the poor quality of education at government schools. The rate of students repeating a grade is low, ranging from 2 to 4% for all school types with an overall average of 3%. Dropout rates are slightly higher than the repeat rate, at 6% overall. This rate is slightly higher at government schools (9%) and lowest at mixed management schools (4%). As government schools are less likely to charge school fees, it could indicate that education offered at these schools is less valued by the students. According to parents and teachers, common reasons for dropping out are the inability to pay school fees, the need to help at home or to work to sustain the family, and students voluntarily choosing not to attend school. Delayed entry to primary is a common issue in Somalia; 50% of students enrolled in primary are over the age of 13.95 Figure 16 compares the exam pass rate, dropout rate and repeat rate across the school types as reported by the headmasters at surveyed schools. 92 http://unesdoc.unesco.org/images/0016/001611/161121e.pdf 93 UNSOM, 2018. A new way to education in Somali schools starts in August. 94 This low number of government students participating in the secondary graduation exams could, at least partially, also reflect the relatively low number of government-managed secondary schools. Out of the 170 schools in the sample, 12 were government- managed schools that combined primary and secondary instruction and only one was a government secondary school with no primary component. It is unlikely that a significant number of government secondary schools would be present in rural areas not covered by the survey. 95 Situation analysis in National Education Sector Policy draft, 2016 NON-STATE EDUCATION PROVIDERS Altai Consulting 50 June 2018 Figure 16. Selected educational outcomes by school type 100 90 80 70 % of students 60 50 Exam pass rate (n = 156) 40 Dropout rate (n = 128) 30 Repeat rate (n = 127) 20 10 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools AVAILABILITY AND CONDITION OF PHYSICAL INFRASTRUCTURE AND LEARNING MATERIALS In both the qualitative and quantitative portions of the survey, infrastructure and learning materials were often mentioned as key indicators of education quality . The presence or absence of certain infrastructure and equipment, as well as their relative condition, determine the perceived quality of a school. Infrastructure As shown in Figure 17, nearly all surveyed schools (99%) had access to water, and in most cases (97%) to an improved source – most commonly piped water.96 The only two schools that did not have access to water were both located in Garowe. The availability of water may be related to the fact that surveyed schools are located in urban and peri-urban areas and therefore have access to more developed infrastructure. Although improved water sources are generally present, there are still concerns over water quality, and distribution systems such as taps may be inadequate. 97 96 WHO/UNICEF Joint Monitoring Programme for Water Supply, Sanitation and Hygiene (JMP) defines improved water sources as “those that have the potential to deliver safe water by nature of their design and construction, and include: piped water, boreholes or tube wells, protected dug wells, protected springs, rainwater, and packaged or delivered water”. 97 Discussions with parents in Bosasso and Kismayo NON-STATE EDUCATION PROVIDERS Altai Consulting 51 June 2018 Figure 17. Access to water by source and school type (n = 170) 100 90 80 70 Government schools % of schools 60 Mixed-management schools 50 Umbrella schools 40 Independent schools 30 All schools 20 10 0 Piped water Borehole Water tank Other source No water As seen in Table 9, most schools have basic facilities such as electricity, a wall for security, and a recreation area. On the other hand, canteens and libraries are less common. Additionally, DCOs were asked to observe and rate the condition of various school facilities on a scale of 1 to 5 (1 indicating very poor condition, 5 indicating excellent condition) during their interview visit. Facilities appear to be in acceptable condition as the average rating was above 3.5 for all facilities with no significant differences between school types.98 However, parents and students during FGDs complained about the poor state of infrastructure. According to a parent in Mogadishu: “This school needs better infrastructure because when it rains it’s affected by the floods and it is dangerous for the lives of the students”. Table 9. Percentage of schools that have select facilities by school type (n = 170) The majority of schools have basic sanitation facilities, with all but one school having at least one latrine. Although most schools have separate latrines for girls, the absence of separate latrines in some schools is especially detrimental to girls who may feel uncomfortable using shared facilities .99 Furthermore, schools generally have too few latrines with a median student-latrine ratio of 97:1.100 This ratio varies between school types as shown in Table 10. The student to latrine ratio is 98 p > 0.05 for all one-way ANOVAs on condition of facilities 99 Discussion with secondary school students in Bosasso 100 Water, Sanitation, and Hygiene (Wash) Programmes recommend a student latrine ratio of 25:1 for girls and 30:1 for boys. See Wash in schools, URL: http://washinschoolsmapping.com/projects/kenya.html [18.5.2018] NON-STATE EDUCATION PROVIDERS Altai Consulting 52 June 2018 significantly higher at umbrella schools 101 and lowest at mixed management schools. Although larger schools tend to have more latrines to accommodate the greater number of students 102 but the number of latrines does not scale to the number of students, 103 indicating that larger schools may have more unhygienic conditions. Table 10. Student-latrine and student-classroom ratios by school type Overall statistics Median by school type Mixed- Government Umbrella Independent Mean Minimum Maximum Median management schools schools schools schools Students per latrine (n = 138 13 800 97 74 78 130 80 159) Students per classroom (n 44 10 233 33 40 30 31 35 = 155) Access to electricity contributes to higher quality education by providing a more effective learning environment through lighting, ventilation, and the use of technology. Although electricity is usually available, the hot Somali climate can still be an issue. According to a student in Mogadishu: “Classrooms are very hot, so we come outside for air, and classes are very boring due to the heat ”. All umbrella schools have access to electricity, possibly due to their concentration in Mogadishu, while only 84% of government schools do. The presence of a security feature (wall or fence) is valued for both students and parents who reported that it is important to feel safe at school. However, many students and parents stated that the greatest danger is in getting to school, as road traffic and explosions are tangible threats. Additionally, students indicated that walls can help eliminate distracting noise from outside the school. The absence of libraries at most schools can negatively impact quality since they provide students opportunities to learn independently and develop literacy skills. Most schools (90%) have a recreation area, which can be an important center for social interaction. Their presence or size may be limited in part by land issues at some schools.104 Although basic facilities are available, the needs of the student body are often inadequately met. As seen in Table 9, students must often share chairs when the school cannot provide enough. At some schools, up to 5 students can share one desk.105 There is a significant difference between school types:106 government and umbrella schools are more likely to have one seat per student than other schools. Similarly, the number of students per classroom was often reported to be too high. Although the median value of 31 is in line with other countries in the region, students, parents, and teachers all listed large class sizes as a major challenge to learning. Smaller class sizes allow more opportunities for 101 One-way ANOVA (F3, 160 = 4.73, p < 0.01); post-hoc Tukey test (p < 0.05 for all pairwise comparisons except with government schools) 102 Linear regression (F1, 161 = 30.64, p < 0.001, R2 = 0.16) 103 Linear regression (F1, 162 = 94.27, p < 0.001, R2 = 0.37) 104 Discussions with parents and students 105 Discussion with teachers at a secondary school in Kismayo 106 X2 (3, N = 170) = 8.90, p < 0.05 NON-STATE EDUCATION PROVIDERS Altai Consulting 53 June 2018 engagement and personal interaction with teachers. Furthermore, classrooms are often physically too small to accommodate the large number of students:  Teacher in Kismayo: “The biggest challenge I face in my job is overcrowded classrooms”.  Student in Kismayo: “Classrooms are very small and congested because of the lack of space”.  Parent in Garowe: “The inadequate number of classrooms available in the school causes congestion”. Overall, there is no significant difference107 between school types. Regionally, the median ratio is highest in Kismayo (61:1) and lowest in Mogadishu (30:1). Learning materials and equipment: Access to computers and Internet Figure 18 shows the availability of the Internet and computers across school types. The majority of schools have an Internet connection (61%) and at least one operational computer (73%), which is mostly used for administrative tasks. The percentage of schools having both is lower at 55%. Parents, students and teachers often expressed a need to expand the use of technology in schools:  Teacher in Mogadishu: “There is a need for more equipment in order to help teachers teach. For example, computers, projectors, and other tools that help the students for practicing the theories that they have learnt”.  Parent in Garowe: “The school doesn’t have a computer laboratory with an Internet connection. This is an essential facility especially in this information age to help the students enhance their learning in subjects they are interested in.” Figure 18. Availability of computers and Internet by school type 100 90 80 70 Internet connection (n = 150) % of schools 60 50 Computer (n = 158) 40 Computer with 30 Internet connection 20 (n = 139) 10 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools Learning materials and equipment: Textbooks Textbooks are in circulation for most subjects taught but are not universally available. They are available at most schools for commonly taught subjects including math, sciences and languages. Umbrella schools have fewer textbooks available than other school types; 34% of umbrella schools do not work with any textbooks. This is notable given that umbrella schools often have other facilities and equipment, but they likely prefer to work with their own materials, substituting textbooks for 107 One-way ANOVA (F3,161 = 0.47, p = 0.70) NON-STATE EDUCATION PROVIDERS Altai Consulting 54 June 2018 handouts. Schools that do have textbooks tend to have them for many different subjects: schools that work with textbooks have the texts for no fewer than four major subjects. Although textbooks exist, there are not enough copies – students must share textbooks in most schools (65%). Usually there are fewer than five students sharing a book but this ranges to over 250, indicating that there may be only one book per grade level. Textbook availability appears to be highest at independent schools, where there is one book per student at 80% of schools, and lowest at mixed- management schools where this the case at 29% of schools. Parents, students, and teachers often complained about Figure 19. Textbooks at two the shortage of textbooks, which can also add to the cost of independent schools in Bosasso education if they must be purchased from the market. The condition of textbooks, which could be an indicator of funding availability and education quality, varies greatly. Some schools have new editions while others have old photocopies. Using a five-point scale (1 indicating poor condition, 5 indicating excellent condition), DCOs rated the condition of textbooks during the quantitative survey. Overall, textbook condition was highest at government schools (3.9) and lowest at independent schools (2.9), with an average rating of 3.3 across all schools. The photographs in Figure 19 give an impression of the different levels of availability and condition of textbooks at two of the surveyed schools. Basic educational materials such as chalkboards, chalk, and stationary are available at most schools. Teachers and students therefore generally have the fundamental resources necessary to deliver and receive education. However, there is still a need to increase the availability of learning materials. For instance, science lessons can be limited solely to theory due to the lack of laboratories and equipment. 108 Figure 20 below presents a word cloud regarding infrastructure and learning material components and how they are perceived by parents, teachers, and students.109 Additional data on infrastructure and learning materials can be found in Annex 4 ‘Additional Information’. 108 Discussions with parents, students and teachers 109 This word cloud was generated from words mentioned when FGDs explored the topic of infrastructure. The relative size of each word is based on how frequently it was mentioned. NON-STATE EDUCATION PROVIDERS Altai Consulting 55 June 2018 Figure 20. Quality of infrastructure mentioned during FGDs EVALUATING QUALITY BEYOND THE PROPOSED TYPOLOGY As described in the previous section, the typology of schools used in this report is based on management and therefore does not capture other important factors that might differentiate schools. These are briefly explored in Focus Box 5 below. Although the highlighted schools are similar in terms of number of students, teachers, and gender balance, some differences do appear to exist. Focus Box 5. Variation in education quality beyond the proposed typology Umbrella-affiliated schools Integrated Quranic schools Rural/peri-urban schools As previously stated, schools not Within the survey sample, there Although the study focused on directly managed by umbrella are 12 integrated Quranic schools urban centers, 14 rural and peri- associations may still be (IQSs). These appear to mostly urban schools were surveyed in associated with one. This is the offer primary education and were all locations except Kismayo. As case for two thirds of the sample – evenly split between Garowe and with IQSs, the small sample size including 46% of schools not Bosasso. Using the indicators means that findings require managed by an umbrella used throughout this section, IQSs cautious interpretation. Rural association. The main observed appear to be similar to other formal schools appear to have fewer differences are that umbrella- schools in terms of education resources, as they are less likely affiliated schools are more likely to quality and learning outcomes. to have electricity and textbooks charge school fees and appear to The only significant differences than urban schools.112 Although have better facilities than other found are that IQSs are less likely not significantly so, teacher schools.110 This may be explained to have computers and Internet salaries appear to be lower at by the fact that government access.111 This could be related to rural schools. These findings may schools, which are less likely to the fact that IQSs tend to be indicate lower quality and explain 110 School fees: X2 (3, N = 170) = 45.1, p < 0.001; Electricity: X2 (1, N = 170) = 4.8, p < 0.05; Computer: X2 (1, N = 158) = 6.7, p < 0.01; Internet: X2 (1, N = 150) = 31.7, p < 0.001 111 Computer: X2 (1, N = 158) = 7.9, p < 0.01; Internet: X2 (1, N = 150) = 5.8, p < 0.05 112 Electricity: X2 (1, N = 170) = 5.6, p < 0.05; Primary textbooks: t (154) = 2.5, 2 tailed p < 0.05; Secondary textbooks: t (115) = 3.0, 2 tailed p < 0.01 NON-STATE EDUCATION PROVIDERS Altai Consulting 56 June 2018 charge school fees, are also less primaries, which are generally less some of the rural/urban disparities likely to be associated with an likely to have computers, Internet, highlighted earlier in this report. umbrella. The difference in and IT classes than secondaries Interestingly, dropout and repeat resources may therefore be linked and combined schools. Given the rates are lower in the rural to the greater availability of funding small number of IQSs included in schools, perhaps a sign that resulting from the payment of fees, the sample, these findings should students that can overcome the and the fact that umbrellas may be interpreted with caution and initial barriers to accessing assist their affiliates by providing subject to further investigation. education are more likely to take equipment. their studies seriously.113 In terms of the region comparison, there is an inherent difficulty in attempting to compare the quality of education delivery in Somalia to regional developments. Although the education sector is visibly underfunded and under-developed, relative lack of directly comparable and up-to-date data only allows for a limited comparison.114 What appears clear is that the gender gap is wider in Somalia than anywhere else in the region. According to the results of this research, there are on average 198 male students per 100 female students in the four towns covered by the survey. In comparison, in the Republic of the Sudan the GER is 76.74% for male students at the primary level and 70.35 for women.115 In Kenya, a host country for the significant portion of Somali diaspora, the reported gender gap is non-existent at primary level with 105.48% female and 105.13% male GER.116 Comparison is less conclusive on other quality indicators . For example, the average student- teacher ratio of 37:1 is somewhat more than 31:1 in Kenya 117 and significantly less than 55:1 in Ethiopia,118 however without additional information such as median values of the distribution or results for urban areas, the implication of this comparison remains unclear. 3.6. ACCESSIBILITY & INCLUSIVITY Key Takeaways ▪ The reliance of schools on fees as their principal source of funding creates financial barriers to education. ▪ Average annual school fees are USD 99 for primary and USD 149 for secondary but there are often additional costs such as registration, uniform, and examination fees. ▪ Although most students travel less than 30 minutes to get to school, distance can be a driver of exclusion – especially in rural areas. ▪ Gender is a major driver of exclusion as financial, institutional and cultural barriers keep many girls out of school and contribute to higher female drop-out rates. 113 Note that these differences are only significant when using a one-tailed distribution. Repeat rate: t (125) = 1.9, 2 tailed p = 0.056; Dropout rate: t (126) = 1.8, 2 tailed p = 0.077 114 Additional comparison could be made on levels of enrolment, on which data is generally available throughout the region. However, as comprehensive enrolment data was not collected under this assignment, comparison of enrolment is left out of the analysis. 115 UNESCO Institute for Statistics, http://uis.unesco.org/en/country/sd, data from 2015 116 Ibid, http://uis.unesco.org/en/country/ke, data from 2016 117 Ibid 118 Ibid, http://uis.unesco.org/en/country/et, data from 2011 NON-STATE EDUCATION PROVIDERS Altai Consulting 57 June 2018 FINANCIAL DRIVERS OF EXCLUSION As seen in the Section 3.4 on funding, the financial burden of funding education falls to parents, meaning that cost is the main driver of exclusion. Government officials, non-government stakeholders, teachers, and parents were all in agreement that costs can be prohibitively high. Students from low-income backgrounds and ethnic minorities are especially disadvantaged since they are more likely to be required to help at home.119 Additionally, their parents may be less invested in their education, leading to a greater chance of dropping out. As described by a teacher in Mogadishu: “There are many children who would like to attend the schools but cannot - every person in the world loves to learn and achieve a higher level in life. The school supports those who are in need by providing scholarships and supporting education. But we cannot provide everything because there are other fees they cannot afford such as transportation and food.” In addition to school fees, there are several additional costs associated with obtaining education . Table 11 presents the annual costs of education borne by parents. Table 11. Average cost of education to the parent per surveyed school in USD (n = 170) Average yearly cost % of schools Average yearly cost Cost category (only when fee is where fee is (across all schools) charged) charged School fee 99 114 88% (primary) School fee 149 163 88% (secondary) Registration fee 4 7 65% Uniform 9 15 58% Textbooks 6 18 33% Transportation 11 79 15% Examination fee 6 8 70% Extracurriculars 3 32 11% Meals 6 135 5% Some parents also reported that their children received private tutoring several days a week after school. Tutors are often hired to help students with Math and English, with tuition fees ranging between USD 10 per month up to USD 15 for a single session. As seen in Table 12, the average total cost to parents is USD 132 per year for primary education and USD 197 for secondary. The cost to attend government schools is lower – largely related to the fact that they are less likely to charge school fees. However, when charged, fees at government schools are similar to those at other school types. A full breakdown of average school fees by school type can be found in Annex 4 ‘Additional Information’. Table 12. Average annual cost of education per surveyed school by school type in USD All Mixed- Government Umbrella Independent schools management School fees 99 57 94 106 111 119 This was mentioned as main drop-out reason besides lack of academic skills in teacher discussions NON-STATE EDUCATION PROVIDERS Altai Consulting 58 June 2018 Primary Additional fees 33 14 40 31 40 (n = 156) Total annual cost 132 71 134 137 150 School fees 149 83 140 164 153 Secondary (n = 95) Additional fees 48 19 82 46 51 Total annual cost 197 102 222 210 204 Although an analysis120 comparing cost to education quality did not uncover a clear trend, additional insights from the qualitative portion of the survey suggested a relationship, or at least the perception of one. As described by a parent in Bosasso: “Those independent schools that charge high fees generally offer comprehensive education beyond only academics. They provide greater academic, cultural, and sporting opportunities, as well as highly-qualified services such as academic support and career guidance. ” Since quality does not appear to be the clear driver of cost, other factors are likely to be involved (Focus Box 6). Focus Box 6. Major drivers of cost Major drivers of cost According to the interviewed government and umbrella officials, headmasters, teachers and parents, the following factors drive up the cost of education :  Public funds are insufficient to offer free or subsidized education in all locations.  The lack of external funding sources forces schools to generate their own revenue by charging school fees.  Insufficient funding limits the ability of schools to offer scholarships to students from low-income backgrounds, causing demand to exceed supply.  More qualified teachers demand higher salaries, necessitating higher school fees to cover these costs.  Families that do not live near a school must pay additional transport fees.  External factors, such as drought and conflict, have hindered the education system’s development and contributed to high costs for parents. To assist students with the cost of attending school and to overcome the challenge of the parents having to carry the major financial burden, schools may offer scholarships. Overall, only 38% of schools offer scholarships to students (Table 13), and the eligibility criteria for these scholarships are not always clear, especially for need-based scholarships.121 Scholarships based on academic achievement are usually awarded to the student at the top of her/his class. Scholarships specifically for girls are also available but are less common. This is relevant insofar as the gender gap is large, and some parents with low income specified that they then chose to only send their sons to school. 120 Details of this analysis can be found in Annex 4 ‘Additional Information’. 121 Teachers and umbrella association representatives indicated that the headmaster and teachers select students of which they think they are most 'needy' from the neighbourhood. It remained unclear based on what criteria this is taking place, and whether this reaches families that are not on the headmasters' or teachers' radar. NON-STATE EDUCATION PROVIDERS Altai Consulting 59 June 2018 Table 13. Availability of scholarships by school type (n = 168) Mixed- Government Umbrella Independent All schools management schools schools schools schools Need-based 15% 24% 12% 10% 18% Performance-based 15% 16% 12% 10% 19% For girls 16% 24% 12% 10% 19% Any scholarship 38% 44% 41% 33% 38% Representatives of umbrella associations reported that for umbrella schools, every school has to accept 15 students from poor neighborhoods for free. However, this regulation does not consider the size of the school, nor the criteria according to which students are eligible for such a scholarship. Additionally, families that have more than five children are not required to pay school fees for two of their children. Orphans and children of Internally Displaced Persons (IDP) are also more likely to receive scholarships. DISTANCE TO SCHOOL Most students live in the relative vicinity of the school, as reported by the headmasters and shown in Figure 21. Figure 21. Average time spent travelling to school by school type (n = 170) 100% 90% 80% 70% 60% More than 1 hour 50% 40% Up to 1 hour 30% 10-30 minutes 20% Less than 10 minutes 10% 0% Government Mixed- Umbrella Independent All schools schools management schools schools schools According to the headmasters, the average student across all school types travels 10-30 minutes to school. However, travel time can still be long for some students (students travel more than 30 minutes to 8% of the surveyed schools, and more than an hour to 2% of the surveyed schools). Additionally, some headmasters indicated that geographic distance prevented certain children from attending school. According to government data,122 most students walk to school, although other forms of transport, including school buses, may be available in some cases. Students tend to travel the shortest distance to independent schools, and the longest distance to government schools. The willingness to travel longer distances may be driven by the desire to avoid 122 MoECHE, 2017. Education Sector Analysis 2018-2022. NON-STATE EDUCATION PROVIDERS Altai Consulting 60 June 2018 school fees. Although most students travel less than 30 minutes to get to school, distance can be a major barrier to students' attendance if they need to pay for transportation, as reported by the headmasters and teachers. Long journeys to school may negatively impact academic performance and increase the likelihood of dropping out as described by a teacher in Bosasso: “The long distances children must walk to school disrupt their concentration in class. Some of them arrive at school sweaty, stressed, and exhausted both physically and psychologically, which compromises their performance to the point that they finally give up and drop out.” In the graph presented below (Figure 22), the distance to school is disaggregated by region. Students travel the shortest distance to school in Bosasso, and the longest in Kismayo. Figure 22. Average time spent travelling to school by location (n = 170) 100% 90% 80% 70% 60% More than 1 hour 50% Up to 1 hour 10-30 minutes 40% Less than 10 minutes 30% 20% 10% 0% Bosasso Garowe Mogadishu Kismayo All Schools GENDER BARRIERS Gender is a major driver of exclusion as evidenced by the large gender gap presented earlier in Section 3.1. Low availability of sanitation facilities and gender-sensitive learning environments, a lack of female teachers, a lack of role models outside of school, safety concerns, and social norms are mentioned as factors preventing parents from enrolling their daughters in school or causing girls to drop out of school. With Somalia being a highly patriarchal society,123 social norms lead to the undervaluing of girls’ education as the role of women is traditionally to remain home. Parents indicated that they would choose to only send their sons to school if they could not afford education for all children. According to teachers, girls being married is also a major reason for girls to drop out in secondary school . These findings are substantiated by Somalia's high illiteracy rate of girls and women. 124 The overall literacy rate of adult women is 36% compared to 44% for men. 123 Gardner, J. and Musse, F., A Gender Profile of Somalia, EU Somalia Mission, 2013; Gardner, J. and El-Bushra, J., Somalia: A State of Male Power, Insecurity and Inequality, Rift Valley Institute, 2017. 124 MoECHE, 2017. Education Sector Analysis 2018-2022. NON-STATE EDUCATION PROVIDERS Altai Consulting 61 June 2018 OTHER DRIVERS OF EXCLUSION Although cost, geographic distance to school, and gender have been identified as the major divers of exclusion, there are other factors that limit access to education. Details on these additional drivers of exclusion are provided in Table 14. Table 14. Drivers of exclusion Drivers of exclusion Explanation Safety impacts access since parents stated they would not send their Safety children to schools in unsafe areas. Lifestyle can be an exclusion criterion for nomads as schools cannot always accommodate their movements if the children cannot stay with relatives living near the school, as education experts stated. IDPs Lifestyle/life situation among the parents also brought up that they face difficulties from having to move often, being confronted with different curricula, and sometimes even a different language of instruction. Lack of parental education and engagement with the children's Lack of parental education can be a driver of exclusion if education is not considered education relevant, and no efforts to facilitate the child's education are made, as teachers stated. Students may be taken out of school in order to have more time to Religion study the Quran, as stated by the teachers. Additional views on the inclusivity of Somali schools are presented in Focus Box 7 below. Focus Box 7. Stakeholder perspectives on inclusivity Stakeholder perspectives on inclusivity During group discussions and interviews, stakeholders agreed that schools do not actively exclude certain groups. “The school focuses on the whole child — the social, emotional, mental, physical, and cognitive development of each student is valued regardless of gender, race, ethnicity, socio- economic status, or geographic location. It prepares the child for life, not just for testing” – Parent in Garowe Instead, access to education is largely dependent on the ability to pay school fees. “No discrimination in this school; the only discrimination is the income, the excluded group is the low-income community” – Puntland government official Parents and teachers stated that schools often try to accommodate low-income families by negotiating reduced fees and offering scholarships. In other cases, parents may send their children to distant schools to avoid paying fees. “If we cannot pay the full amount, we talk with the management and discuss how the student can continue his studies. We make a deal on how the money will be paid ” – Parent in Mogadishu “Other children go to very far public schools because they cannot pay the school fees ” – Parent in Kismayo NON-STATE EDUCATION PROVIDERS Altai Consulting 62 June 2018 In addition to socio-economic status, parents and teachers both stressed the importance of parents valuing education. The lack of support at home was identified as a major reason why many children are out of school. “There are two types of people who don’t send their children to school: those who are economically not able to pay the school fees, and those who don’t understand the importance of education and instead encourage their children to work.” – Parent in Mogadishu Although many parents were understanding of the realities of Somalia’s education system, others felt let down by the government and exploited by non-state providers. Some teachers shared this opinion. “Since the government has failed in the provision of quality education to its younger generation, parents have no option but to seek education from privately managed schools whatever the cost might be.” – Parent in Garowe “High-quality private education remains outside the reach of the majority, and the unfettered drive for profit in the private sector has led to virtual cartelization where a handful of elite schools exploit parents whose collective bargaining position is already weak .” – Teacher in Garowe IMPLICATIONS FOR SCHOOL CHOICE For many families, cost is the principal determinant of which school their children attend. The number of nearby schools, reputation, and perceived education quality may also have a role. The factors influencing school choice are explored in Focus Box 8. Focus Box 8: Factors influencing school choice Factors influencing school choice If families can overcome the initial barriers to education, the next step is to choose a school to attend. During FGDs, stakeholders identified three main levels of choice that families have in selecting a school: 1. No real choice exists 2. The decision is influenced by external factors 3. The decision is based on what is best for the student At the most basic level, school choice is made out of necessity. In such cases, no decision between schools is made - there is only one that could be realistically attended. For instance, there may be only one school in the area. Another possibility is that low-income families cannot afford to pay school fees so have no choice but to send children to schools that offer free education. At the next level, despite having the choice of several schools, factors beyond the control of the family determine which school is chosen. For instance, the security situation may lead parents to send children to nearby schools to avoid travelling through potentially dangerous neighborhoods. The amount of school fees charged, as well as additional costs to bear, such as for uniforms, textbooks, transportation, or exam fees, are external factors that can determine the choice of school. Finally, at the highest level, the school chosen is the one that is most suitable based on individual needs and expectations. The opportunity to select a school purely based on what is best for the student is only available to the most fortunate Somalis. They must be in a comfortable financial NON-STATE EDUCATION PROVIDERS Altai Consulting 63 June 2018 situation to be able to bear the costs of education and live in an area with many schools to choose from. Most commonly, students simply attend the school that is closest to where they live to minimize travel time. When selecting based on quality, teacher qualifications and academic reputation are the main factors taken into consideration. Additionally, language of instruction and the availability of specialized programs or facilities can influence the decision. The quality and stability of non- government schools is perceived to be higher – largely a function of the lack of consistent funding and qualified teachers presumed of government schools. 3.7. ACCOUNTABILITY Key Takeaways ▪ Schools in Somalia are held accountable through both the ‘long route’ and the ‘short route’. ▪ Through the long route, the government uses school inspections and EMIS data to monitor education quality. However, the efficacy of these systems is limited by institutional capacity and funding shortages. ▪ Through the short route, communities interact with school management via formal and informal mechanisms to voice their concerns and improve quality. ▪ Umbrella associations also participate in the short route by setting and monitoring quality standards in response to market pressures. LONG ROUTE OF ACCOUNTABILITY Education service providers can be held accountable by either the ‘long route’ or the ‘short route’. Accountability through the long route takes place when an elected government exercises oversight over the education system on behalf of the electorate. The short route, on the other hand, is taken when citizens interact directly with service providers.125 Federal Government Accountability Mechanisms As set out in Chapter 2 ‘Education Sector Background’, the prolonged absence of a stable government in Somalia has resulted in a lack of a nationwide and centrally imposed framework for education standards. Two main mechanisms for enforcing accountability currently available to the MoECHE are the EMIS, the collection of education data collected by the government, and the REOs, regional representatives of the MoECHE. Both are described in detail in Section 2.3. EMIS is central to MoECHE's capacity to plan, monitor, and evaluate the education sector. Data is collected from government schools directly and through intermediaries like the umbrella associations for other schools.126 This approach has so far resulted in incomplete and out-of-date datasets. According to the interviewed government officials, a change to a bottom-up approach with education providers having the obligation to deliver the information to the Ministry is in discussion. Additionally, EMIS is fully donor-dependent.127 125 The World Bank, World Development Report 2004. Making Services Work for Poor People, 2003 126 EMIS data has been collected since 2011. Part of this evidence base is collecting data on the number of schools, student enrolment rates, number of teachers, and number of students taking national exams. 127 See Appraisal report of the Federal Government of Somalia Education Sector Strategic Plan 2018-2020 NON-STATE EDUCATION PROVIDERS Altai Consulting 64 June 2018 REOs help create accountability by providing a point of contact between the MoECHE and individual schools. They are responsible for carrying out government inspections and providing guidance to schools on how to improve quality. Government Inspections Inspections are the most commonly implemented quality assurance and supervision mechanism used by the government. Scheduled and unannounced inspections are carried out by the different agencies, who are then expected to report to the MoECHE. Figure 23 shows the proportion of schools inspected, disaggregated by school type and the inspecting body. Figure 23. Government inspections by school type as reported by headmasters (n = 170) 100 90 80 % of schools inspected 70 60 50 Government Agency 40 District Education Officer 30 Regional Education Officer 20 10 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools Just over half of the total number of surveyed schools (54%) had been inspected by someone other than a DEO or REO, described in the survey as a ‘government agency’. This may refer to a representative of a state-level MoE or another unidentified part of the state or federal government. Government (96%) and mixed-management schools (88%) are commonly inspected by a government agency, unlike the umbrella schools (16%). Inspections are also less common in Mogadishu (27% of schools were inspected), where most umbrella schools are located, compared to the other locations (85%). Figure 24 shows the frequency at which school inspections take place across the different school types. Government schools tend to be inspected more frequently than other schools (68% of government- managed schools are inspected quarterly while this is the case for only 16% of other schools). NON-STATE EDUCATION PROVIDERS Altai Consulting 65 June 2018 Figure 24. Frequency of government inspections by school type (n = 92) 100 90 80 70 % of schools 60 Monthly 50 Quaterly 40 Twice a year 30 Annually 20 10 0 Government Mixed- Umbrella Independent All schools schools management schools schools schools Survey data indicate that inspections are fairly consistent in their components, most commonly including:  Inspection of facilities (98% of inspections)  Evaluation of curriculum (98%)  Classroom observation (97%)  Assessment of teachers (90%)  Interviews with stakeholders (87%)  Assessment of student learning (86%) Beyond these components, according to an interviewed DEO in Kismayo, the inspections are also supposed to include: classroom size and its adequacy for the number of students, the availability and adequacy of learning materials, and the use of child-centered teaching methodology. In most cases (63%) the visits include teaching feedback for teachers, according to the headmasters. However, in the FGDs, teachers raised concerns that they did not receive feedback or guidance on their teaching. Additionally, the inspection system faces many challenges as the lack of funding makes it impossible to regularly visit every school. 128 Although systems are in place to use the outcomes of inspections to improve quality through workshops, they are limited by the lack of resources, coordination between government bodies, and the inspectors’ limited authority over non -state schools. SHORT ROUTE OF ACCOUNTABILITY Umbrellas As described in Section 2.3, umbrella associations participate in the short route of accountability by adopting a quasi-governmental role towards member schools in the network. Umbrellas set standards members schools must adhere to and maintain their own data collection and reporting mechanisms. These mechanisms tend to be transparent and shared with the MoECHE. FPENS, for example, via its database has access to all schools’ enrolment numbers, the teachers and their qualifications, students’ names, identity cards, their marks, and whether they have transferred from another school. By setting standards of quality, umbrella associations contribute to their reputation and can attract more students. 128 MoECHE, 2017. Education Sector Analysis 2018-2022. NON-STATE EDUCATION PROVIDERS Altai Consulting 66 June 2018 Since schools are dependent on school fees, parents and students may hold umbrella associations accountable through their client power, possibly leading to higher quality overall. Perhaps the most important role of umbrella associations in terms of accountability is their function as a platform for individual schools to voice and discuss their concerns. This ability to collectively communicate with the government agencies is not available to independent schools outside of the umbrella system. Community Based Organizations (CBOs) As described in Section 2.3 both PTAs and CECs in practice have a limited role in holding the school accountable. For instance, CECs inspected 51% of the surveyed schools. This number rises to 68% when umbrella schools are excluded. As with government inspections, CEC visits are most common at government and mixed-management schools. Most parents state that they are very satisfied with their communication with the school management, and consider good communication between parents, teachers, and school management crucial in order to be informed about their children's progress or absence. Focus Box 9 below gives insight into parents' and students' views on the short route of accountability. Focus Box 9. Perspectives on the short route of accountability Perspectives on the short route of accountability In the Somali education system, short accountability is actualized either through formal or informal mechanisms. Parents, students, teachers, and school managements are the principal stakeholders involved in such pathways. CECs and PTAs are the main formal avenues of communication between schools and the community, especially students and parents. The roles of PTAs are varied, but generally, PTAs provide a forum for parents to voice their concerns. “We have monthly regular meetings where we discuss school developments. If we have complaints about either the school administration or teachers, we bring them up during the meeting.” – Parent in Kismayo At some schools, PTAs can also be involved in decision-making, working closely with the school administration. “Parents and school management work together, and parents are involved in the decisions. […] For example, when the school management wants to increase the school fees, the administration cannot just decide to increase without consulting the parents.” – Parent in Bosasso Students may also organize themselves to collectively address issues with school management and teachers, although this appears to be less common than PTAs or CECs. Student associations and the designation of prefects are among the ways that students may liaise with school management. “We complain when teachers are not doing their job well. School management deals with our complaint and advises the teacher to change their behavior. ” – Student in Bosasso Besides these formalized processes, schools often interact with parents on an ad hoc basis. Open lines of communication at many schools allow parents to be informed of their child’s progress, and for schools to better understand individual families' needs. “Whenever there is an issue the school calls us. For example, if the student did not go to the school or if he/she is not doing well in the exams.” – Parent in Mogadishu NON-STATE EDUCATION PROVIDERS Altai Consulting 67 June 2018 4. RESEARCH IMPLICATIONS 4.1. POLICY IMPLICATIONS As research findings presented in this report have shown, education delivery in Somalia in its current state shows the following fundamental weaknesses: • Fragmentation of the education sector among a diverse set of state and non-state bodies. Research captured a wide variety of ownership and management models, in which government tends to play only a limited role. Only 15% of the surveyed schools were described as managed and owned by the government, with the remainder managed and owned by a combination of actors including umbrella associations, local community bodies and private local and diaspora owners. In many cases, two or three of these owners and managers are involved in the same school. To further add to the fragmentation, there is a limited degree of cooperation between the federal MoECHE and state-level MoEs, while the designation of ‘government’ may apply to either of the two institutions. • Lack of government capacity for policy planning, standards setting, financing, collaboration with private providers and data collection. This is in evidence across the sector but namely in the variation in the education quality index constructed for this report: on a scale from one to 100, schools ranged from 44 to 89, irrespective of the school type. There is therefore a wide variation in availability and state of infrastructure, resources and qualified teaching staff, with government agencies having little to no involvement in monitoring or addressing these shortcomings. The lack of consolidated data on education sector became apparent during the survey methodology design stage, when the team was unable to find a consolidated list of schools in major urban areas and had to compile one through a physical mapping process. The MoECHE also has a very limited access to feedback from the individual schools. In Mogadishu, only 27% of the surveyed schools have been inspected by a government body. • Inadequate financing. Almost all (89%) schools charge school fees, which cover, on average, 74% of a school budget. According to the survey findings, an average school is able to access 1.3 sources of funding. This means that a number of schools rely solely on school fees for funding and operates in the state of a chronic deficit. On school level, chronic underfunding is evidenced by delays in teacher salary payments and lack of investment in infrastructure and teaching equipment. Financial support to students from poor families is unstructured and ad- hoc, even as inability to pay school fees was highlighted as a major obstacle to accessing education in interviews. • Lack of teaching capacity. Government does not regulate teacher qualifications and teacher training is offered by a diverse set of state and non-state institutions. As a result, 12% of teachers do not have post-secondary education and the situation varies among school types: 30% of teachers at mixed-management schools have no post-secondary education, while 75% of teachers at independent schools have a university degree. Even though over 90% of interviewed headmasters said that they offered guidance to teachers at their school, none of the teachers who participated in the FGDs confirmed this. • Gender gap. Severe gender imbalance was found in all aspects of school operations. The surveyed schools served a median value of 156 males per 100 females and had on average 6.8 times more male teachers compared to female teachers. Only 3 of the 170 headmasters interviewed during the survey were women. To respond to weaknesses of the education sector identified in this report, Somali and international stakeholders are encouraged to explore the following areas of support and intervention: NON-STATE EDUCATION PROVIDERS Altai Consulting 68 June 2018 1. Induce cooperation between the government institutions (MoECHE and MoEs) and umbrella associations to encourage the spread of common standards and build the FGS capacity. The major umbrella associations have a longer and more extensive history of self-regulation than the FGS institutions in their current incarnation. For the cooperation to be to the mutual benefit of all parties, it should feature the following: a. Collaborative process. Umbrella associations are independent of government in their financial, management, administrative and other procedures, reaching over thousand schools across the country. Regulations unilaterally imposed by the government will therefore likely find limited reception, while mutually beneficial collaboration will extend the reach of the MoECHE and the MoEs across the country. b. Formal forum for structured interaction. This research found no formal or structured interaction between the FGS and umbrellas. A formal forum that regularly convenes major stakeholders would constitute the first step towards a collaborative dialogue, allowing the umbrellas to speak to the government in one voice. c. Incentives to collaborate. Although umbrellas have become accustomed to operating independently of government, data shows that umbrella-managed schools are plagued by the same shortcomings as the rest of the sector: 30% of umbrella-managed schools do not have a chair for every student, 34% do not use textbooks and over 10% of their teachers do not have post-secondary education. These are sector wide challenges and umbrellas are more likely to respond positively to initiatives that will contribute to their solution. d. Extended network through direct engagement of school leadership. Although umbrellas dominate the education sector (by directly managing schools or through a looser association), one third of the surveyed schools fell outside of the networks. As the FGS resources are likely to remain limited, umbrella networks represent the most efficient vehicle for engaging individual school leadership. 2. Clarify the roles and responsibilities of the MoECHE and the state-level MoEs to reduce stakeholder fragmentation within the education landscape. As with a number of other ministries, the MoECHE continues to adapt to the federal governance system and to forge relationships with state-level institutions. This process of clarification of roles and responsibilities can be centered around specific issues such as: a. School inspections. The data shows that inspections are being carried out in parallel by the REOs and the DEOs, who report (mostly) to the MoECHE , and ‘other government agencies’ on the state level. There appears to be no system to allocate responsibility for inspections, to define their frequency or duration, or to process the outcomes of inspections. In addition, there is a lack of inspections, especially in Mogadishu where 73% of surveyed schools have never been inspected. Inspections are therefore a pertinent issue for discussion among the MoECHE and the MoEs on which federal and state-level collaboration could be trialed. 3. Expand, reinvent and refine the business model for education to improve quality and stimulate expansion of the school network. Most schools struggle to cover teacher payroll while relying predominantly on school fees. Some of the directions that could be explored include: a. Grant-delivery mechanisms. In South Sudan, the Girls Education South Sudan (GESS, http://girlseducationsouthsudan.org) encourages schools to fulfil a set of conditions in exchange for a yearly grant that can be used to invest in school infrastructure, supplies of equipment. The program is an example of public-private partnership, whereby the program was administered jointly by an external implementer and government agencies, using government structures to disburse funds. NON-STATE EDUCATION PROVIDERS Altai Consulting 69 June 2018 b. Advise schools on sustainable business models. The majority of budgeting decisions are currently being made on the level of individual schools, by the school leadership with occasional input from communities. Best practices on designing and maintaining a school budget in the Somali context should be developed and shared to better equip headmasters to deal with these challenges. 4. Extend access to education for women and vulnerable children through systemic change. As described above, based on the available data, Somalia has one of the worst gender gaps in the region. In addition, cost of education reduces access to schooling. However, the relative lack of development in Somalia also provides opportunities for change at the systemic level: a. Introduce gender mainstreaming into teacher training curriculum. A development of a national teacher training curriculum would present an opportunity to insert the principles of gender mainstreaming into the core of the education system. b. Prioritize training of female teachers. Quotas for female students at teacher training institutions could improve the gender balance among teachers. In addition, if women are to progress in their education, it will be necessary to encourage female teachers to qualify to teach secondary grade classes. c. Provide targeted financial support. A number of schools have an unstructured approach to supporting vulnerable children with scholarships and fee waivers. Targeted scholarships or subsidies could also be delivered through a grant-delivery mechanism such as the one described in 3 a) above. Such centrally distributed support would benefit vulnerable students and facilitate involvement of MoECHE. 5. Invest in systems that support teacher training and education. Given Somalia’s security and accessibility issues, it will take a significant amount of time and investment to improve infrastructure and increase availability of equipment and textbooks. In this context, investing in teachers is the fastest way to increase education quality: a. Construct a performance baseline. With no regulation on teacher qualification, there is no available data on teacher performance. A baseline will inform the development of a teacher training curriculum and any subsequent regulation. b. Teacher Management and Information System (TMIS) should be used to store and manage data relating to teacher performance and professional development needs. c. Develop national teacher training curriculum to standardize teacher education and qualification across the country. d. Explore the introduction of ‘mobile’ in-service training. This could take the form of ICT solutions taking advantage of smartphones and tablets, but also of ‘mobile’ tutors using motorbikes to visit individual schools, as is practiced in Uganda. 129 e. Subsidize teachers’ wages. The current wage range of USD 50 – 350 corresponds to that of a day laborer. Unless the compensation increases, there is a risk that poor staff retention will undo any effort to improve teachers’ capacity. Such subsidy could be disbursed through a kind of grant mechanism described above. It would nevertheless require a developed TMIS. 6. Encourage and stimulate data collection and reporting within umbrella associations and on the level of individual schools. The first step towards successful engagement with the education sector are reliable records of the overall number and basic characteristics of schools and student enrolment. 129 Hardman, F., Ackers, J., Abrishamian, N., and O’Sullivan, M., 2011. Developing a systemic approach to teacher education in sub-Saharan Africa: emerging lessons from Kenya, Tanzania and Uganda. NON-STATE EDUCATION PROVIDERS Altai Consulting 70 June 2018 a. Strengthen EMIS. The EMIS database is currently incomplete. Alongside strengthening the data processing capacity within the unit itself, MoECHE should focus on more productive collaboration with umbrella associations and on providing incentives to individual schools to submit data. Focus Box 10 offers examples of EMIS data collection in fragile contexts, adopting decentralized approach and taking advantage of ICT technology. Focus Box 10. Country Examples: Afghanistan and South Sudan Afghanistan provides an example of how an EMIS can be used in fragile states to strengthen the education system.130 In 2004, the country’s EMIS was rudimentary and data were collected with pen and paper. Since then, technological advancements have been made to allow the collection and management of data from several components of the education system including schools, teachers, and student achievement. For example, tablets are being used to conduct student learning assessments and monitor results in real-time. Although many challenges remain, the new information management systems have improved accountability and transparency while providing policymakers with reliable data to inform new strategies. South Sudan is another fragile state that has developed an EMIS, initiating its program in 2006.131 One of the main challenges to data collection is the country’s large rural population distributed across remote and inaccessible areas. This made it impractical for the federal MoE to be solely responsible for gathering EMIS data. Instead, a decentralized approach has been adopted with EMIS focal points in each state. So far, the initiative appears to be a success as the EMIS has completed annual education censuses nearly every year since 2007, reaching 95% of schools. Despite the initial cost, Somalia could benefit from the expanded use of technology to monitor progress and promote information-sharing. If implemented correctly, long-term gains would justify the investment. Like South Sudan, Somalia has a large rural population, making decentralized data collection more practical. Leveraging the existing capacity of state-level MoEs would facilitate this approach. 4.2. SUGGESTIONS FOR FURTHER RESEARCH There are a number of research topics that were not covered in this report, due to sample and operational limitations, or because their relative importance only emerged out of the study findings. For future research assignments, the following area of interest should be considered: 1. RURAL SCHOOLS: Since this study focused on urban schools, limited information was collected from those in rural areas. A study specific to rural schools would be required to assess their quality, evaluate the student body, and understand the factors contributing to the education inequalities that exist between urban and rural populations. 2. NON-STATE ACTORS: Among non-state actors, umbrella associations were revealed to be the most prominent as important providers and regulators of education. Since the study mostly relied on data collected at the school-level, organizational information on umbrella associations was limited. An in-depth study of their organizational structure, operations, and resources would be helpful in determining how to effectively structure public-private partnerships to improve educational 130 de Silva, S. and Valsangkar, P., 2015. The Impact of Education Management Information Systems: The Case of Afghanistan, URL: http://blogs.worldbank.org/education/impact-education-management-information-systems-case-afghanistan 131 Ministry of Education, Science and Technology (South Sudan), 2015. Education for All 2015 National Review. NON-STATE EDUCATION PROVIDERS Altai Consulting 71 June 2018 outcomes. In addition to umbrellas, other non-state actors also play key roles in the education system. The structure and function of CECs should be subject to further study in order to assess their efficacy and strengthen the short route of accountability. The specific roles of the diaspora and private individuals remain unclear and should therefore also be investigated. 3. SCHOOL LEADERSHIP: There is a need to further study the role of school leadership (including ownership and management) as they have important roles in determining quality through resource allocation, hiring practices, and establishing school guidelines. The study should pay particular attention to the school-level decision-making process on fundraising, resource allocation, management and training of teaching staff and curriculum. 4. TEACHER QUALITY: Almost no information is currently available on the teaching quality at primary and secondary schools in Somalia. A baseline assessment of teacher capacity and performance across a range of school types would inform the development of teacher training curriculum as well as any future interventions in support of teaching staff. 5. FUNDING: The lack of resources was identified as one the biggest challenges faced by schools. The reliance on school fees places the financial burden on parents and contributes to exclusion. Future research should look more specifically at how much funding schools receive and generate themselves, as well as how the funds are allocated. This would help identify alternative means of financing education and improve the efficiency of spending. NON-STATE EDUCATION PROVIDERS Altai Consulting 72 June 2018