'As different as ground and sky'* - Involving children and communities - A case study from Ethiopia

* Adapted from a comment of a school committee member during the evaluation, 'This year is as different from last year as the ground is different from the sky.'

Figure

analysis: Elizabeth Mekonnen, Abiti Tadele, Dereje Wordofa, Ababaw Zelleke, Rachel Lambert, Camilla Croso Silva, Kimberly Ogadhoh
writing/editor: Kimberly Ogadhoh
contributors: Marion Molteno

What are the problems for children?

Life in Somali Region 5

About 3.5 million ethnic Somalis form the second-largest geographical region (called Somali National Regional State, or Somali Region 5) in the new federal Ethiopia. Somali speakers live throughout the Horn of Africa: in the late 19th and early 20th centuries these areas were divided by colonial powers, dividing Somalis among five East African countries. A large number of Somali-speaking people now live in the eastern lowlands of Ethiopia. Most Somalis are herders whose economy depends on grazing and on finding water for livestock. Because they have a shared ancestry and a shared way of life, this area has strong cultural and economic ties with Somalia and Somaliland, and cross-border trade and migration are common.

Life in Somali Region 5 is challenging. For the past twenty years the region has suffered from the effects of conflict, drought, floods, and from large scale population movements. From 1987, the area was treated as a war zone, during which the Somali government fought to unite ethnic Somalis within its own borders, but the situation has stabilised since 1993 and the change in government. In addition to the Ethio-Somalia war, many parts of the region have been affected by clan conflict and unrest, typically over the use of natural resources and the control of trade links. As a result, the area has received only minimal development investment, and there is little social infrastructure outside of the relatively more settled Jijiga zone. As one author writes, there is '...hardly a road, a telephone, a school, or a clinic...' in this sparse region of the country¹.

'I was born here in Turanod, but I left in 1991. I lost all of my livestock, my house was burnt down, I had nothing. I had heard about the refugee camps so I went there. I stayed there for a few months and then rumours of guerrilla fighting forced the government to take us to a camp near Djibouti. But then there was conflict there, the government stores were looted, and we were forced to flee on to Boroma.' 35 year old woman².

Education is a particularly neglected area. When the Transitional Government of Ethiopia was installed in 1991, Region 5 had the lowest primary enrolment rates per capita³. There were few formal schools, and the majority had been destroyed or abandoned. Doors and tin roofs had been taken off their hinges. Bricks had been dismantled and used to build other structures. In one school not far from Jijiga town, primary classrooms were in such bad condition (broken desks, no blackboard, crumbling walls) that they were being used as toilets and as shelters for livestock.

'Before all of the trouble there used to be a school here.... One of my sons, and my daughter used to attend it. When we were away, the children did not get the chance to get to school, as there were none. There is no longer a school here: it was destroyed in the fighting⁴.' 35 year old woman

What Children experience in School

Only about 20% of children of school age in Ethiopia actually attend school, and the percentage is even lower for children living in Region 5 (For some of the reasons, see Box 1). Inside the few schools that are functioning, children struggle to learn. Teaching materials and text books are in short supply, and most are still written in Amharic - the former national language of instruction - even though the language of the majority of children in the region is Somali. Children often complain that they have to share one textbook among a large group of students, and spend significant time each day just looking for the few that are available. According to Ministry of Education (MoE) reports, only 40% of schools in Ethiopia had textbooks in 1992, and the national student-book ratio was 4:1⁵. In one of the schools visited for this review, students said they had not seen any textbooks the entire school year.

Teachers also face difficulties trying to teach. At the start of the regionalisation process, teachers had little or no formal training: the Regional Education Bureau (REB) estimated that at least 75% of primary school teachers in Region 5 had no formal qualifications⁶. As a result, many teachers understandably could not teach well: they didn't prepare lesson plans, they didn't use teaching aids, and they didn't check to see that students understood the lesson. Outside of Jijiga town, teachers receive salaries irregularly and are poorly motivated. Many do not come to school, or if they do, they come late. Some come to school, but do not show up to class. In one of the schools visited, students reported that the behaviour of their teachers was so poor - they smoked in class, insulted and fought amongst one another - that outsiders wouldn't be able to differentiate between them and the students!

National Policy and Decentralisation

Following the overthrow of the Dergue regime in 1991, a new Federal system was created which grants autonomy and substantial administrative authority to nine regional states, constituted on the basis of ethnic and linguistic criteria. The creation of a federal system has uge implications for the delivery of services, which historically have been highly centralised.

Box 1: Schools - what are the problems? A Community Management Workshop in Jijiga Zone in December 1997, listed the following as major problems: · poor condition of schools · lack of school facilities such as water tanks, latrines and fencing · shortage of educational materials such as desks, windows, doors and textbooks · shortage of teachers in rural areas · shortage of classrooms in bigger towns, and lack of secondary schools · lack of rooms for pedagogical centres in rural schools · need for material supply and follow-up of school pedagogical centre activities · poor students' attendance · high work load at home for children · early marriage for girls · poor relationships between teachers, parents committee, administration, head teacher and REB

In the education sector, major policy changes under regionalisation include:

· using the majority language of the region as the language of instruction, enabling the majority of children to learn in their mother tongue

· decentralisation of central ministry services (e.g., curriculum development, materials production, educational radio production, adult education)

· reorganisation and decentralisation of systems of management and administration⁷.

Box 2: The Central State - Changing concepts of schooling The history of education provision in Ethiopia is one of sweeping changes and reforms. Ethiopia has a long history of religious education (both Christian and Muslim) but modern education was only introduced as recently as the early 1900s. The emergence of central state authority, the arrival of diplomats and missions from abroad, and the growing demand for foreign languages increased the demand for modern education. Emperor Menelik II launched the first official government policy for the expansion of the education sector. From 1935-45, the Italian occupation resulted in the massacre of 3,000 educated Ethiopians and the introduction of the Italian education system. After 1941, Emperor Haile Selassie took forward the expansion of education, introducing secondary and tertiary education and establishing the first long-term plan to deliver primary education to all children. The imperial regime was overthrown in 1974 and replaced by the Dergue, a socialist, military regime. Under the Dergue the education system was reformed to produce 'socialist citizens with all-round personalities', emphasising education for production, education for socialist consciousness, and education for scientific enquiryi. This Marxist approach to education gave priority to mass education and communities in rural areas were mobilised to finance and construct schools. This campaign resulted in a rapid expansion of enrolment, particularly in primary education in rural areas. i. Initially, fourteen new regional states were created However, the increase in pupil enrolment was not matched by an increase in quality, which has declined steadily over the past two decades. Falling quality has been linked primarily to a decline in per student expenditure, scarcity of instructional materials and facilities, and inappropriate curriculum and teaching methodsi. According to a recent research report: The curriculum lacked relevance with no clearly defined objectives, and instruction concentrated more on theoretical knowledge with little connection to daily life. The approach also had a high tendency towards rote learning which did not prepare young people for living in the community'.

Table: Regionalisation - who does what?

Area of responsibility	Ministry of Education	Regional Education Bureau	Zonal Education Department
Policy	Proposes and contributes to national policy	Contributes to national policy, & makes plans for region on basis of national policy. Formulates regional policy.	Proposes plans to the REB
Standard setting	Sets standards	Implements standards	Implements standards
Examinations	Prepares national examinations	Implements and supervises national examinations	Implements and supervises national examinations
Curriculum	Sets curriculum for secondary and higher education; assists in preparation of other school curricula	Prepares primary and junior secondary curriculum	Provides feedback and implements curriculum
Inspection		Inspects schools
Teachers	Sets standards and requires qualifications (above); posts secondary teachers to regions	Pays teachers; recruits teachers and trains primary teachers	Pays primary teachers; provides in-service training
Instructional materials	Bulk procurement	Provides text books and materials	Distributes materials
School establishment	Establishes higher education institutions; licenses private higher institutions; sets standards for institutions (above)	Establishes schools and junior colleges; licenses private schools	Establishes schools and vocational training centres
Data	Collates national school census data & assists in system development	Collates regional data	Compiles zonal data

Regionalisation also provides for an additional level of decentralised authority; the woreda (or district) level of administration. Hence, the new administrative structure will comprise the centre, the regions, the zones, and the woredasⁱⁱ. The plan is or the administrative centre of each of these, as well as each school, to have a Pedagogical Centre to support the development of educational materials. There is provision in principle for budgets to be devolved to the school level, but this has as yet not happened.

^ii. There are currently 9 regions in Ethiopia, with about 55 zones and over 650 woredas. Each region has an Education Bureau (REB), each zone has a Zonal Education Department, and each woreda has a Woreda Education Office.

The central feature of the new policy is that the responsibilities and power of the Federal MoE have been greatly reduced. Under regionalisation, the main role for the MoE will be to determine national standards, while the other three levels will be responsible for their implementation. For example, under the new policy the MoE will be responsible for setting standards and required qualifications for teachers in all regions, while the regions will be responsible for recruiting, training and paying them⁸.

The limits of Regional Capacity

Regionalisation has undoubtedly opened up new opportunities. But it happened suddenly, with little preparation to build up local skills. There was little support or infrastructure at local level to enable the new authorities to put the new policies into practice. In a previously highly centralised system, the people who were suddenly given the authority to make important decisions or implement aspects of centrally determined education policy had themselves never had experience of that level of responsibility. To take advantage of the new arrangements required a sophisticated understanding of systems and budgetary processes. Budgets that were officially available to the regions were not forthcoming, and to extract them would require an ability to manoeuvre through a rapidly changing political context where relations between central and regional authority were often complicated by political/ethnic tensions.

These features took more extreme forms in the more remote regions. In Somali Region 5, they were further compounded by security problems and a history of border conflicts which had left much of the infrastructure (including many schools) destroyed and large concentrations of refugees and displaced people requiring resettlement. Given the poor legacy of investment, ensuring even the most basic provision of education has been difficult. The capacity of the Regional Education Bureau to deliver any kind of educational service is extremely weak. Management of education is affected by the paucity of local, skilled personnel and the high turnover of government officials. Offices in the more remote areas do not function. Budgets are low, or non-existent. Schools and villages are widely scattered, yet there is little money in the budget for transport to enable officials to visit the schools for which they are responsible. For all these reasons, Region 5 lags behind others in many aspects of implementation of the new education policy and strategy, such as developing a new primary school curriculum and opening a Regional Teacher Training Institute.

In adapting to the new policy there have been particular problems associated with teacher employment and language of instruction. Under the Mengistu regime, primary education was in Amharic and secondary and tertiary education in English, with English taught as a subject from primary grade 3. The new policy allows for a diversity of languages of instruction at the primary level and the introduction of English as a subject from grade one. For the first time, the medium of instruction in primary schools is the language of the majority ethnic group. When the Somali language was chosen as the medium of instruction for primary schools in Region 5, more than 300 Somali speakers were recruited to teach in local schools. Unfortunately, this was done in haste, without properly screening for suitable candidates. Many of the teachers who were subsequently employed had never taught before, and did not know how to do basic things, like prepare a lesson plan, use teaching aids, or evaluate student's learning. Some of the teachers had little more than a primary education themselves.

The complexities of divided authority are shown up by a bizarre situation that has arisen out of the transition to a new school language. The issue is, what should become of the primary school teachers who were already in post but who do not know Somali? In contrast with the newly recruited Somali speaking teachers, the established teachers are Teacher Training Institute (TTI) trained, but are now only able to teach the few lessons a week of English or Amharic. Most do not live in the villages where they are allocated to teach and so (theoretically) journey out daily from Jijiga; in practice, they often do not turn up or are seriously late, their motivation dulled by their low workload. Yet as qualified teachers they are on full pay, whereas the unqualified Somali teachers, who now do the bulk of the teaching, have an ambiguous status. The initial solution to pay the Somali teachers at the level of new TTI graduates was not acceptable to the Ministry of Finance and ways of formalising their qualifications are being explored. Viewed from the point of making a rational use of resources, some decision to resolve this situation is clearly needed. But the REB does not have the power to resolve the issue since the role of setting standards (e.g. on teacher qualifications) is retained by the central Ministry. Nor do the regional officials (mainly Somalis) feel it politically advisable to raise the issue, because it could be interpreted as an attempt to oust non-Somali teachers. The situation thus remains: in a state of extreme scarcity of resources, there is an inefficient use of those little available.

The response

Working between Ground and Sky

Save the Children was already working in region 5 at the time when these far-reaching changes were introduced or expected to happen. Save the Children's experience here began in the early 1970sⁱⁱⁱ. From 1988 it has been working in five refugee camps in Jijiga zone with a mixed population of returnee refugees from Somali camps or those internally displaced due to civil strife. The agency's involvement in education began in 1994 when Save the Children commissioned a study of the camp populations to identify the constraints to their successful return home, and investigate how to support people who chose to settle in the areas surrounding the camps⁹. Lack of education was identified as a major constraint for those who wanted to return to their areas of origin, as educational facilities were much better in the camps than in the rural areas of Jijiga.

^iii. Save the Children collaborated with the government to assess the severity of the 1973/74 famine. This led to the establishment of a nutrition surveillance programme, which developed into a long-term presence in the area when Save the Children intervened in the 1988 refugee emergency, setting up health and nutrition programmes.

Towards the end of that year, Save the Children and the Zonal Education Department jointly carried out a needs assessment of the educational sector, from which developed the current programme of work. The needs assessment highlighted a number of problems:

· The regional education structures were newly formed and scarcely functioning: there were few trained staff and a limited budget.

· More than 75% of teachers had no formal qualifications or training¹⁰. They had been recruited to respond to the urgent need to provide teaching in the Somali language, but recruitment had been done without adequate technical screening.

· Communities had not been involved in decisions affecting the delivery of educational services. As a result, during the chaos surrounding the fall of the Dergue regime, lack of ownership in the school system had led to the widespread looting and destruction of schools.

It became evident that one of the main stumbling blocks was the lack of communication between the new authorities and people in the community. In trying to help get the newly decentralised education systems in place, Save the Children has chosen to work at the interface between the 'Ground' - what children actually experience - and the 'Sky' - education policies and provision. The programme is not conceived as a set of 'Save the Children' activities, but as a series of supports to encourage more people to be actively involved in learning what children and communities want from schools, what actually goes on in schools, solving the problems, and influencing the planning of school provision to be more appropriate to the needs of local children.

From participatory assessments of the issues facing schools, the following were selected as areas where support from Save the Children could contribute towards such processes:

· Providing support to regional officials on planning and supervision;
· Involving the community in what goes on in the school;
· Improving the school environment;
· Developing a training course for teachers to upgrade their teaching skills.

Two other priorities emerged after a review of the initial two-year phase of the programme and were incorporated into the next phase:

· Reaching children who aren't in school: participatory research on adapting schools to be more inclusive

· Supporting the Region to develop the Somali curriculum and Somali language textbooks.

What has been learnt, and how has it been used?

Through the years of Save the Children's involvement, children, parents, teachers, and education authorities have met together, expressed their problems and needs, discussed issues, and argued about possibilities. Linking all these activities is the assumption that one mechanism for improving schools is to encourage more sets of people to be involved; and specifically, that if adults in the community are given a framework for taking a more active role in their children's schooling, and can learn about the kinds of problems that children are experiencing, they will be in a position to use that information to improve what happens in schools. This is an assumption that could be said to apply anywhere, but its implications are particularly far-reaching in situations of severe resource constraints.

In April 1998 field research was conducted to gain a better understanding of these processes. The researchers aimed to learn:

· To what extent have Save the Children-supported activities created opportunities for more people to learn about and be involved in what goes on in schools?

· Whether there has been an improvement in the quality of schooling, from

- better planning and supervisory capacity among officials
- more parental involvement
- the skills-upgrading course for teachers

· What children and parents think about the changes that are still scheduled to happen under regionalisation, including the new curriculum.

Data was collected through semi-structured interviews from six primary schools over a period of two weeks, conducted by small teams who met periodically to record and classify results. A detailed checklist of questions was developed to guide interviewers, and responses were analysed to show:

· what people had learned about the problems children were experiencing in school;
· how this information is shared;
· what (if anything) they were doing to help address these needs.

The team of interviewers included regional and zonal officials, members of parents' committees, and students, making a total of eleven people external to Save the Children. In addition there were eight representatives from Save the Children and one representative from Radd Barna (Norwegian Save the Children.)^iv

^iv. For further details of the research methodology, see appendix A.

An important part of the methodology - new to many participants - was the central place given to asking children what they thought:

'We never used to ask children. We just talked to other adults and drew our own conclusions. When you ask the children, they tell you and you see it in a new way. For instance they say clearly, I don't want to be beaten'. Elizabeth Mekonnen, Save the Children Programme Officer

Findings from the review suggest that school children in Region 5 do benefit from more people knowing about, and being in a position to respond to, the problems that they face learning in school. Specifically, the findings suggest that children benefit by:

· having more people know about the problems they face;

· having more people (and more people at different levels in the system) share information about these problems;

· having more people available to try to address these problems;

· having more people available to check-up on what other people are doing to address these problems.

Because people with different roles in relation to the schools (such as officials, parents, students, and teachers) are learning more about what children are experiencing in school, there is more potential for action to be taken at different levels. This is not to say that everyone does take action: in some cases it was found that people were in a position to help solve problems that arose, but chose not to help, or were blocked by others in their attempts to assist.

The following sections highlight what has been learnt about the problems faced by children in the schools visited; and describe the mechanisms that have been created to encourage wider local involvement and sharing of information, so that these problems can be addressed.

Regional Officials: learning from users

A major component of Save the Children's activities has been to support the Regional Education Bureau (REB) to learn firsthand about the problems of schools in the region, as experienced by children, parents and teachers, so that this understanding can inform the Region's planning. Because of the severe resource constraints facing the REB, there is not the budget locally to cover the cost of visiting the schools within their area. One important mechanism adopted by Save the Children has been to provide the means in terms of vehicle and per diem to get regional officials out to visit the schools they manage. These have often had an immediate practical impact, enabling the officials to sort out specific problems (e.g. the practical difficulties for teachers of where and when they get paid), as well as serving the more general purpose of letting them talk with students and other members of the community to better understand what hampers the effectiveness of school provision.

Training and awareness activities for a core team of school supervisors have aimed at developing their planning, supervision and management skills so that they can genuinely support school heads and classroom teachers. Topics have included:

· how to get community members involved in planning discussions;
· how to improve communication between teachers and head teachers;
· how to collect base-line information about the schools for planning purposes.

Trips have also been organised to other regions for regional officials to see how the process of regionalisation is being implemented elsewhere and to get a comparative sense of provision in their region in relation to others. Through the combination of all these activities regional officials have been stimulated to consider:

· what features of life in Region 5 might inform the needs of the schools, and the broader school system;

· which national policies and procedures seem to work well at local level, and which do not.

The possibilities opened up by these activities have been enthusiastically received. As one official told the research team,

'We can only identify needs across schools if we go to the schools and see what the needs are.'

But the findings from the review also make it clear that lack of resources often stops people from acting once they become aware of a problem. One regional official admitted:

'I don't have the means to follow-up.'

If action is to be taken on some of the issues that have come up, donor funding will be required, though not necessarily large amounts. One outcome of the programme is that it has put regional officials in a stronger position to define for themselves and negotiate for the kinds of funding that they prioritise, based on their own analysis of the problems. For instance, REB officials pressed Save the Children to include in the second phase of the programme support for developing the new Somali curriculum, which had not initially been one of the areas Save the Children expected to get involved in, but which regional officials insisted was a priority, even though they recognised that they did not have the resources (human or financial) to implement it.

There is also evidence that being exposed to the views of children and parents has given regional officials the confidence to tailor national policies to local situations. For example, the REB issued a statement in response to guidance from the MoE that schools must not take in more than 45 students in Grade One. But in some zones in Region 5 this meant that so many children would be left out of school that parents went to the region and complained - so despite the fact that reversing this decision would result in crowded classrooms the REB retracted this statement and suggested that all children be enrolled.

The Community and School Management

· 'Sometimes parents come to class, but they don't ask us anything' (student) Prior to 1994, the government ran and managed schools. Although parent forums had an official place within the school structure, parents were not actively involved in educational planning, and did not generally feel as though the schools 'belonged' to them or to their community¹¹. Under the new education policy, roles and responsibilities of parents have been clarified and given new emphasis. Specific structures have been developed to allow community members to participate more broadly in school management and administration, including the formation of 'parents committees' (selected by other parents at the beginning of each academic year), 'school committees' (composed of parents and other interested members of the community) and 'education guidance committees' (composed of head teachers, parents, teachers, students and a representative from the local government administration).

Save the Children has been involved in helping to raise awareness about the role of parents in school planning and administration through organising workshops at community level. Efforts have focused on helping to activate a community management structure that had been set up by the national government. Save the Children has helped to formalise a forum for parents, elders, and local authorities to share information about - and work together to improve - different aspects of the school environment. A series of workshops has been held involving students, teachers, community leaders, parents and education officials to raise awareness about the importance of participating in educational planning and management, and to help clarify and better understand their roles and responsibilities in running the school. Follow-up activities have included helping parents set up parents' committees, identifying projects to rebuild or rehabilitate the school facilities, and helping parents to mobilise other members of the community to get these projects underway.

Improving the Learning Environment

· 'The classrooms were in complete disarray' (parent) During transition, one of the first problems to be noted by members of the community was the poor state of the local schools. Consequently, a key component of Save the Children's education programme in Jijiga has been working to rehabilitate primary schools in the region, through joint efforts initiated by local communities. One of the functions of the parent's committees is to mobilise the community to contribute towards rebuilding the school. In all of the schools we visited the community had provided contributions like labour and raw materials (such as sand and wooden poles) to build latrines, water tanks, new classrooms, and fences. In some schools the committees had been able to raise additional contributions, including funds for staff salaries and learning materials/teaching aids. In one community the local authorities were even persuaded to donate ten cents from every kilo sold of 'qat'^v to the school to ensure ongoing material support. Both students and parents in the schools we visited said that renovating the schools had helped to improve the learning environment.

^v. A local herb widely used as a stimulant

Part of this strategy involves raising the profile of the schools - both within and outside the local community - so that the conditions of the schools are more widely known. This involves documenting and sharing basic information about the schools, and creating opportunities for officials to visit the area to actually witness the environment and poor conditions for themselves. Seeing the situation often stimulates action. For example, when the Minister of Education visited the region a few years ago, she also visited one of the schools where the classrooms were being used as toilets. 'Everyone was embarrassed because the classrooms were in complete disarray' one parent remarked. This collective embarrassment led to change: soon after the Minister's visit the community elected a parents' committee and began to raise funds to help renovate the school. In an unprecedented move, the zonal office donated several bags of cement.

What does Parental involvement achieve?

The committees are the main vehicle through which people not directly involved in the school system can learn more about and influence what goes on inside the school. 'Now parents feel comfortable with school activities' students at one school said. All of the schools we visited had recently formed parents' committees, although participation in school activities varied widely between schools; from infrequent visits to handle major complaints to daily meetings with teachers and students to monitor classroom learning. In most of the schools parents played a key role in helping to resolve problems facing teachers and students, including problems that arose between these two groups. In our review, typical problems reported to parents by students included: late or absent teachers, wrongdoings by teachers, and bullying by other students. Typical problems raised by teachers were: lack of or late salary payments, and lack of teaching materials. Problems that could not be resolved by parents were usually reported to the school supervisor, who commonly visited the school once or twice each year.

*

TEACHERS AND PUPILS

Do Teachers come more regularly?

· Why are we the ones going after the teachers?' (student) Teachers can't teach if they don't come to school. Before Save the Children began the skill upgrading programme, getting teachers to come to school - and to class on time - was reportedly a common problem for children in Region 5. 'The teachers used to come when they wanted' a student at one school told us candidly. Five out of the six schools we surveyed reported that teachers' attendance rates had generally improved over the past year, but a few schools still had a few problems with a few teachers, and getting teachers to come on time was a nagging issue in four schools. In another school, students said that they 'struggled' to get one of their teacher's to come to class, while students at a different school had even gone to one of their teacher's home to 'beg' her to return.

Findings suggest that teachers may not have been inspired to come to school regularly because they weren't always paid promptly for being there. However, the main reason why teachers didn't come to school consistently appears to be lack of awareness by people outside the schools that teachers weren't showing up, and lack of enforcement^vi. One strategy now being used in several of the schools we visited is establishing a system of joint reporting in which students and parents act as 'class monitors'. The role of the student monitor varied in each of the schools we visited, from simply taking attendance (of both teachers and students), to following up on cleanliness, reporting student illnesses, and monitoring the morning processional to raise the flag. Schools that had elected class monitors kept two kinds of records: daily attendance records that were maintained by the head teacher, and records for each period that were maintained by students, to help ensure that all periods were occupied every day.

^vi This problem was probably also related to weaknesses in school supervision, and lack of community involvement in school administration and management.

Schools that had the fewest problems with absent teachers actively monitored this information. More importantly, in those schools with the fewest problems, information about attendance was also monitored by people who did not work directly within the school system. For example, in the only school that did not report problems with absenteeism there was an active, three-way reporting system: students reported problems to members of the parents committee; committee members constantly asked students whether things had improved; and if they hadn't they reported this news back to the head teacher. In other schools, the more problems there were with teachers, the fewer the links' in the reporting chain.

When teachers don't attend school regularly, students are less inclined to show up. In one school where students said their teachers were frequently absent, students admitted that they lacked discipline themselves, and usually 'disappeared to town' if their teacher didn't show up. In most of the schools we visited, the same mechanisms that were being used to monitor teachers' attendance were also being used to monitor student's attendance. In this school, however, there were no reporting systems in place, so parents were not always aware that their children or the children's teachers were not in school.

Can they teach better?

· Not all teachers are good - some get off the topic' (student) Teachers now come to school more regularly, but can they teach? According to the students we surveyed, many of their teachers used to teach poorly because few of them had received any training. 'Our sports teacher used to give us a ball and remain behind, without giving us instructions' students at one school said. Students in another school said that their teachers used to be 'uninterested in teaching'. And our findings suggest that these teachers continued to teach poorly because few people (other than the students) knew that they were teaching poorly.

Tests were subsequently given to ensure that teachers had at least a 12th grade education. To help upgrade the teaching skills of those who had passed this qualification, Save the Children has supported a series of short-term training courses, focused on basic teaching methodologies and subject areas. Training courses in school management have also been developed for head teachers. Save the Children's contribution has been largely organisational and financial; the courses themselves have been run by Ethiopian educationalists, and the approach to both content and methodology have therefore been fairly traditional. Save the Children has also facilitated awareness-raising sessions on issues where they could draw on the organisation's wider experience, such as strategies for including children with disabilities.

Students in five schools say that there has been a 'great change' in classroom teaching since their teachers attended the training courses, and these improvements were only noted for teachers who had participated in the training programme. A central component of the training has been supporting teachers and administrators to learn more about the process of teaching and learning. Students say that their teachers now use a range of techniques to involve them in classroom activities, such as organising group discussions and school debates. They also prepare and integrate teaching aid materials into the lesson plans, and encourage students to choose and make their own learning materials. These changes were highly appreciated by the students. 'I understand when I make maps myself' one student said. There is some evidence to suggest that these new methods have helped to improve the student's grades. Teachers in three schools also say that these methods have improved their own teaching skills.

With the advent of the parents committees, people outside the school now have the authority to visit the classrooms to monitor the teachers' performance. This gives parents and other community members the opportunity to witness problems for themselves. 'I sit in classes and I see' one parent told us. 'I can tell if they are learning. If not, I follow it up.' One of the things this parent found was that many of the Somali speaking students had difficulty understanding instructions given to them by their Amharic-speaking teacher. Since he understood both languages, he was able to act as a translator. 'Now the teacher asks if he doesn't know how to explain in Somali', the parent said. Sitting in on classes also gives students the opportunity to talk to parents and supervisors about problems they experience with individual teachers. For example, students in one school told a visiting supervisor that 'Some teachers just write on the blackboard and sit down without any explanation'. Unfortunately, however, not all of the parents committees take the opportunity to sit in on classes or talk with students, and - according to the students - not all of their complaints have been addressed: in the case described above, the students said that there had been 'no reonse' from the supervisor about this issue.

Another component of the skill upgrading programme involved providing information about how to identify and work with children who have disabilities. Responses from teachers suggest that this information was illuminating, helping them to be more responsive to the needs of their students. For example, one teacher told us that - at first - he thought that one of his students who was deaf was slow. 'I saw that his attention was poor', the teacher said. 'I went to the back of the class and repeated the lesson and he understood, so I realised that the problem was his hearing. He wasn't slow. Now I put him up front near me during classes.' Since the training, teachers in most of the schools now take different measures to assist students with disabilities: such as rearranging seating plans so that children with poor hearing and eyesight can sit up front; allowing children with physical disabilities more time to get to class; speaking louder, or writing larger on the blackboard; tutoring slower learners; and instructing other students to assist children with disabilities.

Teachers and the new School Language

· 'Some teachers are guests in our school' (student) In 1994, the Somali language was chosen as the medium of instruction in schools in Region 5. For the majority of the groups we surveyed this was a very welcome change, because it meant that most children in the region could now 'easily understand what is learned'. Students and parents were particularly vocal about the advantages of learning in their mother tongue: in two schools Somali-speaking teachers were even described as being the 'best' teachers in the school, simply because the children could understand them. However, in those schools and communities that were more ethnically diverse, reactions to the new choice of language were more subdued. Parents from a mixed community near Jijiga town mentioned that the perception of the school among Somali speakers had improved as a result of this change, while it had declined for those who spoke other languages.

The decision to adopt Somali in schools affects many aspects of the school system, including staffing requirements (both in schools and local government), curriculum development, management and administration. Teachers who speak Amharic now have to learn to teach in a different language, and face being replaced by teachers who speak Somali: those who already do are in very short supply as described in the opening section. New textbooks need to be developed, and other materials need to be translated from Amharic. Because this decision has widespread implications, implementing the policy has been slow, and erratic. This has caused unique kinds of complications for students that are only beginning to be recognised and addressed. For example, national examinations were developed in Somali before the Somali textbooks were completed, so in one of schools we visited students had been taught in one language and set examinations in another. Teacher's guides are not likely to be ready in time to be distributed with the new texts. Staff hired to develop the radio programmes to supplement the primary curriculum do not have sufficient resources to complete them, so students may be examined on material they haven't yet learned.

In the majority of schools we visited, problems related to language were being raised and addressed through the parents' committees. The earlier example of a parent fluent in both languages volunteering time to help in lesson translation is a good illustration of the practical benefits of parental involvement. In another school, parents responded to student complaints about the arrival of examinations in Somali by requesting zonal officials to send the exams back in Amharic 'so that children's performance wouldn't suffer'.

Teachers and Discipline

· 'Teach our teachers not to punish us' (student) In one of the schools we visited, students reported that the behaviour of their teachers was so poor - they smoked in class, insulted and fought amongst one another - that outsiders wouldn't be able to differentiate between them and the students! Our findings suggest that disciplinary problems among teachers (and in two cases with head teachers) had been an issue for most of the schools we visited, but had largely been resolved by removing the principal offenders. In most cases these problems were well known to people both inside and outside the schools, but were seen to be too great to be resolved internally. The majority of cases were subsequently resolved by local education officials, but in one instance the offenders had to be disciplined by the local police. In the majority of schools we visited parents and head teachers said that teachers were now more self-disciplined in school. A representative from the zonal bureau even said that he had noticed that teachers were better behaved outside the school: for instance, they paid more attention to personal hygiene (combed their hair, wore smart clothes), and no longer chewed qat in public.

A more troubling issue for children is how teachers discipline them. Students in half of the schools reported that at least some of their teachers punish them physically, although this was not confirmed by parents or teachers in one of these schools. According to some of the students, some teachers routinely 'slapped and kicked' them, made them sit in painful positions 'holding our ears', and beat them over the head with books or sticks. There was substantial evidence from the review that children did not prefer this style of discipline. Children in one school labelled this kind of punishment 'abuse', and children in another school said that the way they were punished was 'severe'. Eight out of eleven students we surveyed in one school said that they preferred to be given advice rather than be beaten. However, statements from regional and zonal supervisors suggest that physical punishment is still widely accepted and practised in schools. 'How can a child learn without a stick?' one teacher asked a school supervisor during a routine visit.

Findings suggest that the use of physical punishment has decreased since the beginning of the training course, and that participants on this course have begun to discover how children can learn without being beaten. In one school children said that there was a great difference between teachers who had and had not attended the course, in terms of the way they were treated. Similar statements were made by parents in another school. Both groups said that the teachers who had attended the course were 'better teachers' because they didn't beat their students. Teachers in half the schools we surveyed said that they used other methods to discipline children, such as talking with and advising them, giving them extra chores on the school compound, speaking with their parents, and expelling them from school. Neither head teachers nor parents in these schools reported significant behavioural problems with students, suggesting that students did in fact respond to these methods

One of the mechanisms suggested during the training course to address discipline problems in school was to create school disciplinary committees composed of students, teachers, head teachers and parents. Two of the schools reported that they had established such committees, and two others indicated that they had some sort of reporting system in place. These structures appear to be very valuable channels through which children can keep tabs on each other's behaviour - both inside and outside the school. In one school, just having a committee appears to have raised students confidence: 'we know we have the right to be heard' a student told us. Students also said that because they were involved in the committees, they could report quarrels that occurred outside the school grounds. Perhaps not surprisingly, there was a strong negative correlation between the use of physical punishment and the use of disciplinary committees: in schools where teachers did not beat students, active reporting mechanisms or formal discipline committees were in place.

Teacher-pupil relationships

· Some teachers leave and go out without saying anything' (student) Although the majority of teachers do not beat their students, students in a few schools still felt that some of their teachers 'lacked respect' for them, by making rude remarks, 'mocking' them, or otherwise treating them unfairly. In one school a student reported that this behaviour was more pronounced in interactions between teachers and children with disabilities: 'Some teachers treat children with disabilities differently' he told the research team. Another student in this school said that teachers had more respect for the girls than they did for the boys, because the girls were 'quieter'.

Findings suggest that students want better relationships with their teachers. 'Teachers shouldn't see us as their enemies.' a female student at this same school told us. 'They should try to understand us.' Some students say they try to be better understood by reporting incidents to the school disciplinary committee, the parents' committee, their head teacher, or the school supervisor. But in most cases this behaviour was not reported.

Relationships appear to have improved when teachers learned more about the nature of childhood and adolescence provided through the summer training course. 'We learned how to handle students better because of the child psychology' one teacher said. The head teacher at this school remarked that after this course the teachers in his school were more aware of the 'life situation' of the students because they were more sensitive to the children's needs. They paid more attention to what was bothering them, tried to find out if they were sick, hungry, or having problems at home. One teacher described how this information helped him to assist a student at his school:

'There is a slow learner in Grade 3. I tried to provide additional assistance, but she was not improving. Then I asked about her family background and found that she lives with her elderly grandmother and she is very poor. I brought her case to the attention of the head teacher and the school agreed not to collect money from the family.'

Students in one school also said that those teachers who had gone on the training course treated them with greater respect than teachers who had not been on the course. 'The trained teachers thank us at the end of class, so we know the class is over' a student explained. Other teachers simply walk out, leaving the students confused. 'We even think he may come back, but he doesn't.'

*

A RESPONSIVE SCHOOL SYSTEM?

The New Curriculum

· 'It is not our culture' (student) There is significant anticipation about the new primary curriculum currently being developed for Region 5, which will be written in the Somali language and seeks to reflect Somali lifestyle and culture. Each region is responsible for developing primary school textbooks, teacher's guides, and supplementary materials in a majority language for children in grades one through six. The MoE sets national standards for the curriculum by providing an outline of the syllabus and objectives in each subject area. These outlines are then used by the regions to emphasise local situations, culture and lifestyle. For practical reasons, the books are being prepared in stages: in Region 5 textbooks have been developed for Grades 1 and 5, and those for Grades 2 and 6 are currently being written. Save the Children has helped to train curriculum developers in technical production and provide funds for their salaries.

Most of the people surveyed - including the majority of students - were aware that a new curriculum was in the process of being developed, and had similar ideas about what they hoped it would include. Their comments also illustrated what they didn't like about the current curriculum. Their primary objection was that the old curriculum didn't reflect Somali culture and religion. This was expressed in different ways. Several groups described items in the textbook as being 'unfamiliar' to people in the region, such as different kinds of food, household articles, and different styles of architecture. Other things were described as being 'inappropriate' culturally, like illustrations of a woman wearing shorts, and a picture of someone washing a dog. The absence of Somali images - pastoral scenes of herders, pictures of mosques, and the use of Somali names and stories - was mentioned by groups in almost every school. While groups in two of the schools said that it was equally important to learn about things outside Somali culture, respondents clearly wished to see more of their own history and culture represented in the texts, and learn more about their surrounding environment.

Community responses to new concepts

· 'We only see pictures of girls in home economies' (student) The involvement of outside agencies in discussions about curriculum raises new issues about responsiveness. UNICEF has had a role in framing the centrally designed curriculum, Save the Children in framing questions to ask communities in Region 5. Both have a child rights perspective, which leads them to raise issues that are in some senses age-old (e.g. gender), but are raised in new ways that may not be seen as relevant by local communities. How have the communities responded to such questions?

On issues of gender and disability students' statements indicated that they were open to new ideas about how the curriculum could reflect all children's potential involvement in schools. They wished to see a balanced representation of males and females in the textbooks, and positive portrayals of people with disabilities. The current lack of a 'gender balance' was reported in five out of the six schools we surveyed (in the sixth school children said that they didn't have any books, so therefore couldn't comment about what they contained). Two schools reported that their texts contained more pictures of men than women, and teachers in one school reported that male names were used more frequently than female names. Groups in two schools also expressed concern about the way in which females were portrayed: for example, girls in one school said that pictures of females were only found in the home economics subject. Similarly, teachers in four schools observed that people with disabilities had not been adequately portrayed in the old curriculum and said that there should be illustrations of disabled children playing and learning alongside children who do not have disabilities.

· 'We have nothing to do with AIDS' (parent)

The question of HIV/AIDS was much more contentious, touching as it does on areas that most people do not speak about even outside school. While information about AIDS will be included in the new national syllabus (and is already the topic of a national radio programme), teachers and parents we spoke to in Region 5 were divided as to what aspects of this subject should be taught, and by whom, and there is concern about the way in which information will be presented in the classroom. Teachers in four of the schools we visited reported that this is a subject they now teach following Save the Children supported training in HIV/AIDS; usually alongside other health issues, and with the assistance of local health personnel. Teachers in another school said they did not apply what they learned in the training in the classroom, because they felt it would encourage children to have sexual intercourse, and they felt that primary school children were too young to learn about the issue (however, the head teacher in this school reported that children do learn about HIV through the national radio programme!).

While there was general agreement among the adults we interviewed that children should be informed about the effects of the disease in school, overall, parents were more likely to say that information about transmission and methods of prevention should be left to them or to religious educators, or not taught at all. 'We have nothing to do with AIDS' one parent remarked 'we have the Holy Quran'. A parents' committee member in another school said, 'Learning about condoms in school is not good. Religion allows us to teach this and how to prevent it'. In contrast, children generally appeared to be comfortable with the level and type of information that they were getting about HIV in school, particularly in those schools where students were also actively involved in disseminating messages about the disease.

Since the textbook that will include this subject has not yet been written, Save the Children and the REB are currently exploring ways to bring people together to discuss how this can be done sensitively. Overall, the findings suggest that issues about what children are or should be learning in school have not yet been actively raised in general forums and discussion groups. During the review it became clear that members of the community had not had a formal opportunity to influence what would go into the new curriculum, but wished to help determine what children would learn in school.

Why are children not in school?

Save the Children's initial activities have aimed at improving the quality of what goes on in schools; this is therefore what the review questions also emphasised. But the majority of children in Region 5 do not attend school; because they have to work, their parents won't allow them to go, school is too far away, or a multitude of other reasons. [Box 3 gives the reasons that people gave for this].

More recent Save the Children activities have been directed towards trying to understand the kinds of changes in the pattern of school provision that might increase access possibilities for children. One group of issues relates to children in families with a pastoral life style. Participatory research in pastoral areas has highlighted the significant fact that while relatively few children of pastoralists attend formal school, almost all children (and a large percentage of girls) attend at least a few years of Koranic school. There are at least three reasons why this is so: 1) Koranic schools are mobile, and move with communities at certain times of the year; 2) the school day is flexible, and organised around children's work day (Koranic schools typically open in the early evening, for example, when children have returned from grazing animals); and 3) the community values the kind of education that Koranic schools provide (such as knowledge of the Quran, and basic literacy in Arabic) so they help make it possible for children to attend, for example by constructing ponds for water use by students.

In some places, basic literacy and numeracy have been incorporated into the Koranic school curriculum. These initiatives suggest an alternative model of schooling that bridges the gap between, and tries to maintain the best qualities of, both Koranic and formal schools. Save the Children is now working with the REB in Region 5 to see whether some of the features of Koranic schools can be more widely introduced so that girls and children of pastoralists have a greater chance to attend school.

Do improved schools attract more children?

· 'We are waiting for you and Allah' (child, not in school) Another approach to the question of access is to consider the effects of school improvements on enrolment patterns. There is a common perception that more parents would encourage their children to be in school if they felt school was useful to them. The review sought to discover whether there had been a perceived change of this kind in Region 5, and also the extent to which members of the community were helping to raise awareness about the plight of children who were not in school.

Four of the schools we visited reported increases in enrolment over the past year. Groups attributed increases in enrolment to two main factors: efforts by parents to raise awareness about the value of education, and overall improvements in school facilities. A central role of the parents' committees is to motivate other parents to send their children to school. 'We tell them to bring their children' one parent said. 'We tell them that education can bring them out of darkness.' In most schools parents said they went about this by talking with other parents, and in two schools parents said they also helped other students financially. Students reported that these efforts had helped to change parent's attitudes towards schooling; particularly their attitudes toward the value of schooling for girls. 'Parents know that girls may marry and divorce and come back to the family uneducated, so they had better have their own skills and education' one student said. Physical renovations appeared to have both direct and indirect effects on enrolment. Students in one school said that enrolment had increased because there were better facilities within the school like latrines and water tanks. Teachers said that renovations had improved the learning environment by providing better protection from harsh weather, and causing fewer distractions on the school compound. Two schools reported that enrolment had increased for specific groups of children who had particular problems getting to school: girls; children of pastoralists; children with disabilities; children who are very poor; and over-age children. 'Students who weren't here before have come back' one student said. One girl mentioned that it had been possible for more children to come to school because of the recent federal exemption from paying school fees^vii. Another girl said that the reason for the increase in female enrolment in her school was that girls were now able to 'plan work around school' due to introduction of shift systems: one month they went in the afternoons, one month they went in the mornings.

^vii. Under regionalisation, responsibility for financing primary and junior secondary education has been devolved to the regions. While official school fees have been banned by the central MoE, schools and communities will still be expected to supplement operational budgets for infra- structural improvements through income generation activities (USAID, 1993)

While enrolment increased in the majority of schools we surveyed, two schools reported that certain groups of children (girls and students in higher grades) were more likely to drop out earlier than other children. Girls had a tendency to drop out when they reached maturity - around the age of 13 or 14. One school had attempted to address this by establishing a 'girls-only' class for Grade 2, but found that this wasn't very effective. Parents in two schools reported that the lack of a junior school in the area was a disincentive for children who made it through to the higher grades: since there was nowhere to go when they graduated, finishing the last few years of school was less important. Parents in one of these schools said they are now lobbying for a new school to be built.

Box 3: What keeps children out of school? Only about 20% of children of school age in Ethiopia actually attend school, and the percentage is even lower for children living in Region 5. One factor is the long distance to school, particularly in the more rural areas, where classes for children in upper primary grades are often non-existent. Some children in these areas have access to schools for grades 1-3, but then have to walk long distances to attend higher grades, and travel even farther to attend high school. In the town of Heregale, for example, once children complete grade three the closest school available to them is 12 km (or 2 hours walking distance) away. The only high school available is located in the regional capitol, Jijiga: because of the long distance to this city (42 kms) most children will be forced to drop out at this stage: others who are more fortunate will have to spend the weekday or semester with relatives in town and then return to their villages over the weekend or when school closes. Another factor is time. Because children are closely involved in the agro-pastoral activities of their families, many are unable to combine learning at formal schools with their household responsibilities. A primary feature of pastoral life is mobility: moving with the seasons to find water and grassland for herds. Depending on the degree of pastoralism, most children of school age are expected to accompany elders during periods of migration. The main migratory period, called the 'jilaal' dry season, typically occurs between October and March each year. Since children also assist in agricultural activities in April, their ability to attend formal primary schools is limited from May to September. However, during this five month period children have a different set of obligations that also limit the amount of time they have for school. Both boys and girls, for example, are expected to herd animals, sow crops, and collect firewood, while girls are also expected to collect water, prepare food and sell milk. Some children are also involved in wage labour, either as domestic labourers in towns, or as paid herders. 'My youngest son does go to... school here. He is ten years old, and in addition to his schooling he is looking after the cattle of my son-in-law. My older son is too busy to go to school, he helps my son-in-law with the farming.'i 35 year old woman Girls are particularly disadvantaged in the current system, with few attending - even in the early years -and high drop out rates throughout the primary cycle. In Jijiga zone, available statistics show that 28% of children enrolled in primary school are girls, falling to 17% at secondary leveli. The reasons for low girl enrolment are complex and cultural, closely connected to the heavy domestic responsibilities of girls and their future role as wives. 'There is no school in this village, if there was one then I would like to go. There is a Koranic school, but my parents don't let me go, I don't know why. Everyone can go to the school, but as they get older, the girls are too busy to carry on'i 9 year old girl Children with disabilities face similar challenges. One consequence of the ensuing conflict and of poor health conditions in the region is the large number of children with traumatic disorders and physical disabilities. The majority of poor families cannot afford the mobility aids that might make it easier for children with disabilities to attend school, so many are simply kept at home.

What has been learnt?

The effect of local involvement

The benefits of involving a wide range of people in the design and development of local education are complementary, with effects at different levels multiplying together to bring greater improvements for children. Information and ideas about problems and solutions are shared across the system; people with power to act at different levels of the system begin working together to achieve change, and others bring in different kinds of resources from their various backgrounds. Finally, it is important to have a variety of people at different levels with an interest in checking up on what others are doing to resolve the problems.

Making a reality of the possibilities of decentralisation

The new national policy of decentralisation provides a framework for community involvement, by giving people the authority and leverage to respond to challenges at different levels. However, it does not automatically lead to more community-responsive education: because of resource and capacity constraints, regionalisation created significant new problems, alongside new opportunities. In particular, officials at regional and zonal levels, as well as teachers themselves, lacked the experience to take on new roles expected of them.

Participatory frameworks need to be created at all levels within the school system - and connections made between them - if benefits are to reach children in schools. Unless responsive mechanisms are created all the way down to classroom level - and children allowed to be part of those mechanisms by providing their own version of events - people responsible for managing schools won't have the proper information on which to act.

Linking providers and users

Seeing what children experience allows officials at regional level to tailor national policies to local situations. In Region 5 regional officials are using their broadened view of the education process and a base of information informed by local realities to begin to shape the school system around the needs of the local community.

The importance of children's own perspectives

Children's views have had a significant effect in challenging adult's perspectives on the purpose of education and on methods used in schools. Perhaps because children are not traditionally consulted, their ideas have had a greater impact through providing a fresh vision: for example, many adults in the community were simply unaware that there were problems with teaching methods and teachers' behaviour in schools. Children's own descriptions of their experiences in school made this rapidly clear. Children were also more much more open to contentious innovations (such as education on HIV/AIDS) than adults had expected.

A facilitating role for external agencies

External agencies can help to create responsive frameworks by helping people get the information on which to act, and by helping to provide the resources for people to act on what they find. Though lack of resources often stops people from acting once they become aware of a problem, meaningful contributions do not have to be large: in Region 5, Save the Children began with small-scale, well-targeted and relatively low-cost interventions (like funding the cost of transport, exchange visits and teacher training) that could foreseeably be within the reach of the government's education budget.

The style of international NGO work

The review concentrated on processes that had been stimulated among local people, and did not ask participants to comment on the role played by the international NGO in facilitating these processes. But there are some pointers from the experience in Region 5 as to what factors may have contributed to the generally positive results.

· Before the start of the programme Save the Children had considerable knowledge of the area and culture, built up through work in other sectors.

· Management of the programme rested in the hands of local Save the Children staff, who built up relations of trust with both communities and government.

· Save the Children staff was genuinely concerned to support local processes rather than control them, and did not see themselves as experts on education, but rather as facilitators of a process whereby local groups would become increasingly proficient. Communities and officials were centrally involved in planning, prioritising, implementation and evaluation.

· The programme offered a medium-term involvement in the development of better schooling rather than short discreet inputs, and developed organically in response to what was continually being learnt about the specific nature of the problems to be tackled.

Editors' Conclusions · Decentralisation of authority within the school system needed to be accompanied by support to build the capacity of officials and teachers to take on new roles. · Save the Children provided support for planning and supervision, curriculum development, adapting teacher training courses, sharing experience across regions, and facilitating systems for community involvement. These low cost interventions could be undertaken within resource-constrained government budgets, with a disproportionately high impact on the quality of schooling. · Mother-tongue teaching was one of the most important factors in making school worthwhile for children. But simply recruiting Somali-speaking teachers (to meet policy requirements), without addressing their lack of understanding of basic teaching methods, failed to solve the problems children experienced in schools. · The participatory evaluation of the programme involved a wide variety of groups with an interest in education. Their ideas and perspectives were fundamental to the schools' success. · Before training, many teachers were unaware of children's diverse needs; similarly, before asking the children their opinions, many adults had no idea that there were problems with teachers' behaviour and teaching methods. · Capacity-building has an impact not only locally, but also puts regional education officials in a stronger position to define for themselves the needs of their region, and to negotiate for the kinds of funding they prioritise based on their own analysis.

Notes

¹ Parker, B., 1995. Ethiopia: Breaking New Ground, Oxfam Country Profile, Oxfam, Oxford

² Save the Children, 1997. 'Evaluation of Community Resettlement and Reintegration Programme Activities', internal report, Save the Children, Ethiopia

³ Penrose, P. 1996. 'Budgeting in the Education Sector in Ethiopia', unpublished report commissioned by Department for International Development

⁴ Save the Children 1997

⁵ USAID, 1993. DeStefano, J. (et al), 'Ethiopia Education Sector Review, Part II', Addis Abeba, Ethiopia

⁶ Save the Children, 1998. 'Proposal for Banyan Tree Foundation: Basic Education Support Programme in Somali National Regional Sate, Ethiopia', internal report, Save the Children

⁷ USAID 1993

⁸ Penrose 1996

⁹ Farah, 1994. Internal report, Save the Children

¹⁰ Save the Children 1998

¹¹ Save the Children 1998