"To serve the basic learning needs of all requires more than a commitment to education as it now exists. What is needed is an expanded vision that surpasses conventional delivery systems, whilst building on the best in current practices" (Declaration: 4).

Funding agencies and their partners are able to draw upon a range of instruments by which they can contribute to improving access to basic education and enhancing its quality. Some of the factors which influence the selection of instrument will be examined below, and reflect the importance of the particular country policy and institutional context, and levels of aid dependency. In broad terms, however, the late 1990s have witnessed a growing dissatisfaction among some agencies with more traditional project forms of support, and increasing enthusiasm for sector-wide approaches. It is difficult to track this shift in terms of aid flows, since sector budget support is not currently captured in the aid statistics of most agencies, but it is nonetheless apparent from recent agency documentation and correspondence. This section will look at the delivery mechanism options open to agencies, the factors influencing their choice, the reasons for any shift in approach, and some of the potential strengths and weaknesses of each instrument.

The principal delivery mechanisms used to deliver support to basic education include the financing and implementation of discrete projects, support to area wide programmes (often agglomerations of projects), technical co-operation within ministries or at local government level and, most recently, support for sector budget finance.

Agencies also provide general programme assistance, which may be defined as 'assistance made available to a developing country, without specific sector allocation, for general development purposes, i.e. balance of payments financing, general budget support and commodity assistance' (OECD 1991:5). Its lack of sectoral specificity is such that it is excluded from the analysis here, which naturally examines instruments in terms of their potential to contribute towards improving basic education.

Projects - "discrete time bound interventions with clearly specified objectives" - have been the dominant mode of channelling agency support to all sectors, including education and basic education, since at least the 1960s. Despite its well-documented limitations, it remains the dominant mode of aid delivery in education, though this is likely to change within the next few years. Among the familiar criticisms, three points stand out (OECD 1999):

the project mode easily leads to the construction of enclaves with few spread effects to surrounding areas;

it has usually been characterised by strong funding agency dominance during the design and implementation phases, resulting in low levels of financial and institutional sustainability;

project support often bypasses the Ministry of Finance (and sometimes the Ministry of Education) in partner countries, which can distort attempts to plan for a rational use of resources at the national (or regional) level.

A proliferation of projects can result in policy fragmentation and duplication, and management overload through servicing agency missions (Ratcliffe and Macrae, 1999). However, countries with low levels of aid dependency and a well-developed capacity to manage and co-ordinate funding agency contributions, have made successful use of project assistance (e.g. China). Even where sector approaches are dominant, there is still a need to finance and implement specific investment projects, and partner governments may still consider that funding agencies can play a useful role here. What is essential, however, is that any project assistance is consistent with the broad education policy and budget frameworks. It should not be used as a vehicle for circumventing weak government commitment to basic education, inadequate budget and management systems or an inability to co-ordinate agencies. These are core issues which have to be tackled head on (see Ratcliffe & Macrae 1999).

A growing recognition of the limitations of the project mode of delivery has led some agencies to seek other methods for aid delivery, with a particular accent on methods designed to increase local ownership and achieve a greater integration of agency and government effort. Central to this has been the growing emphasis on sectoral approaches.

The OECD defines sector programme assistance as 'programme assistance directed to a specific economic or social sector, such as agriculture, education, community development and transportation'. This can include grants or loans which combine a package of technical assistance and/or investment project financing and programme assistance, provided that the whole package is provided as a single transaction. This is a very broad definition, and more recently two tendencies, albeit overlapping and blurred, can be seen to have emerged.

One tendency is often named the Sector Investment Programme (SIP), which has for several years been dominated by the World Bank. However, several bilateral agencies, notably Denmark, Finland, Norway, Sweden and the UK, among others, participate in SIPs, but give preference to sector wide approaches (SWAp). Both have in common the fact that they are designed to create a supportive framework for the delivery of services to clients. This alternative tendency, the SWAp, hinges on improved co-ordination among funders, increased trust between partners, and sufficient implementation capacity of the partner government. The SIP approach, the most common until now, retains the identification of individual agencies with particular components of the programme, which remain in this sense projects. The process of putting together a SIP has tended to involve quite active brokerage by the World Bank, and an element of conditionality remains in terms of 'buying' policy reform, as well as some restrictions on the use of funds (see Oxford Policy Management 1997). The Bank's current support of sector programmes generally entails providing investment lending which is earmarked to specific expenditures and subject to World Bank project lending procedures (see Foster and Naschold 1999).

The SIP model presents risks, identified by Ratcliffe and Macrae (1999) as:

a danger that all the weaknesses of a project approach could be reinvented;

limited opportunities for continual strategic review and/or negotiation; and

little incentive for regular monitoring/review Moreover, one agency suggested that practice and principles diverged:

Characteristics in Principle

Sector-wide in scope

Based on a clear sector strategy and framework

(all) local stakeholders in charge; All main agencies participating in the funding

Common implementation

Reliance on local capacity

In practice

Restricted in scope

Collection of sub-sectoral strategies with little integration

Little non-governmental involvement

Collection of ear-marked projects

Separate arrangements

Excessive reliance on external consultants

One major difficulty is that bilateral agencies often require that 'their' aid should be identifiable, even though the principle was to combine it with government budgets, because agencies say that they need to be able to account for the use of their funds. In practice, as Gould, Takala and Kokkala (1998) say "the free merging of funds in a recipient controlled funding basket does not yet occur anywhere".

However, it is important to distinguish between divergences in the way in which financial support is provided (common pool versus narrowly attributed discrete projects, and everything in between), and the related policy dialogue and conditionality. There is a growing consensus among agencies on the importance of accompanying sector support, of whatever kind, with dialogue designed to promote a more favourable policy and institutional environment. It is recognised that an agency can "no longer push its own policy priorities through project support but will have to build their arguments and priorities into the policy dialogue" (Lexow, Fergus and Dalseng, 1998).

In principle, sector wide approaches can focus greater attention on education sector performance, intra-sectoral linkages, and outcomes and service quality by giving greater weight to improving the policy, budgetary and institutional context in which education services are delivered. SWAps are seen as providing a better mechanism for effective government-agency partnership, including enhanced national leadership and ownership of reform plans, and improved mechanisms for joint government-agency dialogue and performance review. Sweden has been a strong advocate for SWAps, and its views on the potential advantages offered by SWAps are indicated in Box 10.1.

Box 10.1: Potential Advantages of SWAps

encourage the recipient government to take the leadership and to use the foreign exchange provided in accordance with government priorities;

secure a realistic and constructive agency-government dialogue;

contribute to better agency co-ordination;

achieve a better connection between the financing of the sector development and the macroeconomic objectives;

make the use of resources spent in a sector more transparent and subject to dialogue;

facilitate long term financial sustainability in the sector (Sida 1995

To be successful, however, agencies need to give the government sufficient time to develop its own strategy, which may take two or three years (as in Ghana health sector). Government ownership and government ability to co-ordinate agencies is likely to be seriously undermined if agencies are too eager to start disbursements, or if too many agencies feel the need to highlight 'their' individual contributions (OECD 1999).

Other agency-side obstacles may prevent education SWAps from realising their full potential. Most agencies are, in practice, slow to agree on the harmonisation of procedures and reporting requirements, or having agreed to fund only activities that figure in government expenditure plans may undermine this by lobbying for the inclusion of their own projects in those plans.

However these difficulties and the emerging tensions between SIPs and SWAps are not unbridgeable. Possible ways forward include:

promoting an open exchange of views on concerns over government management systems through Joint Steering Committees and annual reviews;

the gradual dovetailing of existing project assistance within agreed policy frameworks;

agreeing a timetable for the phasing out of multi-track financing mechanisms, including agencies with the greatest difficulties (e.g. Germany (GTZ), Japan (JICA) and Canada)13

The extent to which education SWAps realise their potential depends on how far developing country authorities and their funding agency partners are able to avoid certain pitfalls. SWAps, including education SWAps, have tended to concentrate exclusively on the government's role, and without paying greater attention to the potential role of the private sector, NGOs and local communities SWAPs may reinforce centralisation and the power of government administrative structures - even though the intention is the reverse (see Ratcliffe and Macrae, 1999).

The risks can be reduced by involving the private sector, NGOs and community groups in the planning and implementation of reforms, and defining more clearly the respective roles and responsibilities of the private and public sectors. Relatedly, the incorporation of assessments of the institutional capacity-building needs of the private sector and NGOs should be incorporated into education sector programmes. To reduce the risk of over centralised and top-down approaches, SWAp design needs to address the need to stimulate bottom-up demand. This has occurred successfully in the Primary School Facilities programmes implemented in Cambodia, Lao PDR, Tanzania and Uganda (see Ratcliffe and Macrae 1999).

Additional 'risks' to the successful implementation that have been mentioned by some agencies are:

weak leadership, commitment, ownership and management capacity obtains at central and district levels;

education users are not consulted on the issues (especially with regard to equity and poverty);

there are not good links between the Ministry of Finance and the Ministry of Education;

information is not shared.

Although no longer 'new', the implementation of sector wide approaches largely remains in its infancy. There are a number of examples of education sector approaches in practice, including the reasonably positive example of Uganda. Ratcliffe and Macrae (1999) outline the main characteristics of the Ugandan sector investment programme as being:

clear links between education policies and broader development objectives (such as improved governance systems, decentralisation and poverty reduction);

development of agreed sectoral and sub-sectoral policies and strategies, including definition of public spending priorities and clear roles for the public and private sectors in education management and financing;

clear definition of the medium-term resource requirements, including linkage with forward budget plans and the broader macro-economic outlook;

strengthening of common management and audit arrangements for fund disbursement and procurement of services;

institutional reforms and development and broad decentralisation in line with civil service and Local Government reform policies and plans'

agreed joint Government/funding agency structures and processes for strategic negotiation and sector performance review.

an evolutionary way forward, through periodic review (twice yearly), rather than a blue print model. Thus ESIP is not 'written in stone'.

a strong partnership or compact between the government and a consortium of agencies, who agree to pool aid funds to support comprehensive plans developed by the government. The aim is to 'put governments in the driving seat.

This positive impression accords with the findings of Foster and Naschold (1999), who have analysed the positive steps taken to ensure financial accountability to parents, governments and agencies in Uganda's education sector programme. Nevertheless, there remain difficulties, which are likely to be encountered in other programmes. In particular, the monitoring of compliance with financial management and accounting requirements was judged to be weak and measures to strengthen financial management have been proposed. These suggestions may be of relevance to other SWAps, and highlight the kind of attention to practical issues that is required if education SWAps genuinely to benefit the population. Measures included:

enhanced Budget Framework Paper process, integrating recurrent and development planning, sectoral and local;

action plan and skills development for introducing outcome oriented budgeting;

mechanism for decentralised funding of development activities by local governments;

staff development plan in financial management at centre and local level, including non-salary measures to improve motivation and recruitment'

computerisation of local accounts;

improved audit and audit follow up, including supporting accountability to Parliament, supporting Parliamentary Accounts Committee;

performance monitoring support for statistics, economic and poverty data systems (within Ministry of Finance), support to delivery surveys.

Box 10.2 Sector Approaches: Some Early Lessons

SWAps may be better placed to internalise external factors, especially macro-economic and institutional concerns.

Flexible budgetary support needs to secure additionality to sector funding

A long term vision (not blueprint) is critical showing clear linkage between strategic targets and realistic resource envelopes.

SWAps for education needs to be set within broader civil service and local government reforms; sector ministries alone cannot lead from processes.

General attention needs to be given to outcomes and outputs, rather than simply resource shifts, linked to reliable data collection and monitoring systems.

Early commitment to SWAps can build up confidence, trust and capacity, while accepting the need for flexible adjustment to strategy and budgeting.

Agencies need to show more trust in Government systems, linked to reasonable accountability and audit mechanisms within Government.

Source: Foster 1999.

Success of PA or Non PA

Project Activities Strategies:

In that it aims to support the provision of services central to the education reform, such as planning, budgeting/accounting, teacher training and curriculum.

Focused on strengthening institutional capacity to create better functioning systems that deliver better services (e.g. curriculum budgeting and accounting, personnel, procurement).

Supportive, in that it provides for elements such as special studies, workshops and technical assistance - that are not conventionally or routinely covered by the education budget.

Non-Project Activities

Basic capacity and structure within the ministry of education to deliver education services.

Adequate host country management and accounting systems in place.

Commitment from the country to support the reforms, so that it is prepared to provide adequate public financing for the education sector reforms. Mutually agreed upon performance conditions between the government and agencies that reflect the country education policy framework.

When one or more of these conditions is absence, the use of NPA should be postponed, and PA can be used to create the requisite conditions.

Box 10.3 Commitment to Implement Reform

In a report on Guinea Bissau, Pehrsson, concludes that there remain fundamental questions about the role of school and say "In a situation where there is neither a Basic Law, nor a Curriculum Plan, which sets out the objectives of the educational systems and identifies the beneficiaries, it is difficult to interpret the role of the school. There does not appear to be a political will to implement a basic reform (although over the last twenty years, there has been a huge effort to redesign curricular and to produce new books)" ... It is not productive for either side to maintain an expensive sector programme ... when such fundamental questions remain undefined and the organisation of the systems blocks the necessary changes.

Pehrsson 1996

Agencies acting together may push SWAps on countries that are indeed not ready for a comprehensive sector reform programme, and where project and targeted technical Co-operation would be a better option.

The preoccupation with institutional reform, with system management, may result in insufficient attention being paid to the actual delivery of services and their impact on poor and marginalised groups.

"A serious challenge is to make sure that the process of negotiating support to [Education Sector Development Programmes] does not hijack the national democratic procedures of the host country. External partners have a tendency to forget the policies and budgets, i.e. long term programmes are supported to be debated and approved by national parliaments and their constituencies" (Sida 1995).

Allsop (1999) argues that, during their own development, no agency country would have tolerated the kind of intervention into their own economies and policies which are implied by the SWAps for current recipient countries. A poignant illustration was provided at a recent Education and Development Conference in Oxford, UK (September 1999), when the partnership aspects of SWAps were being explained by an agency official. A Minister of State of Education for Uganda said that the 'partnership' felt more like going naked into the Conference Chamber: the next day, he was asked to respond to an Anglo-American perspective on empowerment!

A further challenge concerns the possible need for an approach which bridges sub-sectors or, indeed, makes linkages with sectors outside of education. Many agencies argue for supporting secondary education in parallel to support for the primary sector, because education "operates as a system". A country's overall educational framework has to be capable of "absorbing" such a large impact at the primary level - which, according to some agencies, may itself also have a deleterious effect on quality. Some go further to argue that higher education is also essential and is complementary to (not incompatible with) the overall emphasis on basic education in development

A number of agencies identify the need for an intersectoral approach which makes linkages with other sectors, such as health. If, for example, one wants to understand the factors affecting the success rate of children, one has to take account of health status, cultural and social environment at home and socio-economic conditions in the community, as well as any data from the education system itself.

For the Swiss, sector policy for basic education is primarily an orientation and management tool to be effective, it must be geared to the local situation and main principles. "Basic education must not be allowed to become an isolated programme, it has applications in health demography, gender balanced development and vocational training" (SDC 1996:23).

The emphasis on an all-encompassing sectoral approach is sometimes negated by the exclusion of specific educational components. Whilst this is clearly relevant for capacity building and training included in the support to other sectors, it also has a particular implication for the focus here on basic education. Thus, for example, whilst early childhood development is naturally recognised as being part of the education sector by nearly all agencies, adult education is often taken up in other sectors both by the agency and by the recipient country (see Chapter 8).

For example, the Educational Sector Investment Programme in Uganda did not include discussion of the government's (and some agencies) rather substantial commitment to the Functional Adult Literacy programme, simply because the latter is organised by and under the responsibility of the Ministry of Gender, Labour and Community Development. On the agency side, the Belgian point to the difficulty of isolating their aid to basic education from aid to public health.

Agencies vary considerably in the extent to which they have embraced SWAps as an important approach, ultimately, their preferred instrument for delivering support to the education sector. Although many agencies have adopted or are supportive towards a pro-SWAp policy, some of the smaller agencies say that the restricted number of the staff mean that it cannot (yet) be implemented. Moreover, many agencies, even if they have adopted a SWAp approach, still have projects and programmes: Netherlands and Norway both say that over 50% of their current budget is committed via projects.

General tendencies are summarised in Table 10.2, though the reality is considerably more complex. Thus, for instance, Finland notes that SWAps represent a 'practical solution to the problem of translating political and economic dialogue into long-term development programmes' (see Ratcliffe and Macrae, 1999: 27). In addition, Finland in practice provides significant support to sector programmes in Ethiopia, Mozambique, Nepal and Zambia. There is also ongoing intensive preparation for implementation of SWAps. A similar situation prevails for Denmark. France and USAID have been less ready to support SWAps, but more recently have been shown to be increasingly willing to join SWAp partnerships, including in Uganda and Cambodia respectively (Ratcliffe and Macrae, 1999).

Generally Supportive

Denmark Finland France Netherlands Switzerland USAID CDB UNICEF

Less Supportive

Austria (restricted staff) Belgium CIDA (restricted staff) GTZ JICA (moving towards)

There are several possible implications of a move from projects to SWAps:

how do SWAps 'fit' with decentralisation

the kind of partnerships with NGOs and the ownership by governments.

equity and gender issues

link to the variety of possible delivery systems

what is the status of basic education within education and the status of education versus any other sectors.

The first issue is considered in the following two sub-sections; and the second in Section 10.8. Although the other issues are likely to be important, information is not available from the survey to cover them. As Ratcliffe and Macrae (1999) remark, there is little discussion in agency documentation of how communities or representatives of different groups should be included in discussions about SWAps. There are discussions about the appropriate curricula. The recent emphases on thematic approaches, although relatively new, does not appear to be logically linked to SWAps. Similarly, whilst it is emphasised that all approaches - however diverse can be brought within one envelope, experience from other sectors suggests that the minority concerns will tend to be squeezed out if their funds are not ring-fenced. Finally, some commentators have remarked that basic education might get lost in the Sector Approach; but again there is little concrete experience.

The extent to which development agency expertise and decision-making authority is concentrated in agency capitals or located in developing countries has implications for their capacity to focus effectively on basic education. There is clearly no simple correlation between decentralisation and effectiveness, and indeed potentially there are both advantages and disadvantages (see Table 10.2).

Some agencies consider that having empowered staff at the country level is an absolute requirement for effective partnerships in basic education; but this itself has implications with, for example, SIDA realising that neither their staff nor many of the consultants on their registers have the appropriate skills, and so need retraining. Such staff are seen as better placed to identify the opportunities for promoting education and specifically basic education agendas and engaging in country-level influencing of government, civil society groups, and other agencies (including the multilaterals).

Potential Advantages

helps understanding of political context and country specific knowledge; complexities of poverty

helps dialogue and partnership - and co-ordination

may make more responsive to emerging opportunities

may strengthen team-working across disciplines

involvement of civil society § strengthen government standard procedures

Potential Disadvantages

may dilute message from centre (bad if the message strongly supports basic education; good if not, and the flexibility allows BE proselytisers to flourish)

possible loss of objectivity

may be harder to recruit skills needed for basic education focus

higher running costs

hampers the exchange of experience and good practice, and dissemination of principles of developing basic education

The increased emphasis on sector-wide approaches is also seen by several agencies as requiring more staff at the country level, able to engage fully with the government and other agencies over the two-to-three-year development process. This period, when policies and institutional reform are designed, is seen as very intensive, requiring trust and support, which some felt that even advisers at the regional level could not supply. The potential disadvantages of decentralisation were felt by many to be less weighty, and could be mitigated by careful attention to systems of accountability and lesson-learning. However, some of, especially, the smaller agencies realise that they do not have the capacity to have 'mini-agencies' in each recipient country (e.g. Canada) indeed, that has used as a reason for not adopting the sectoral approach

Finally, whilst there has been considerable agency decentralisation from the metropolis to the recipient countries, the hugely increased need for 'policy advice' across the sector has spawned a new breed of TA policy advisors (e.g. ADB).

Cox (1999) provides a rough guide to the extent of decentralisation of many agencies - referring to the proportion of staff located in overseas offices and the degree to which these staff have decision-making authority. Only a minority can be classified as highly decentralised (Germany, Netherlands, Norway, Sweden, UK), with the remainder evenly split between the moderately decentralised and the mainly centralised.

There is considerable variation in practice and capacity. Denmark, for instance, has a substantial range of expertise in its embassies, as do the Dutch whose embassy formulates a sector plan on the basis of an analysis of the education sector, the sector policy and the main economic, political and social factors influencing its implementation. The World Bank has located some country directors overseas, but only a small number. Officials consider that this makes them more client-driven (although this is not necessarily consistent with an emphasis on basic education, with country directors able to call on a range of expertise as they need it. However, most Bank expertise remains located in Washington. UNDP traditionally has been more decentralised than many agencies, but has sought to increase decentralisation, while trying to improve accountability.

Historically in most partner countries where there are several agencies in the education sector, there have been regular inter-agency meetings to co-ordinate their aid programmes (or at least to know when they are stepping on each others toes). The Jomtien Conference moved this approach along in respect of basic education through the Round Tables sponsored by UNESCO in order to promote Education for All and the advent of SWAps here, of course, forced agencies to discuss a joint package of aid together.

According to UNESCO (1992), of the 121 countries from whom information was collected by the Secretariat to the International Consultation Forum on Education for All in 1992, sixty seven (53%) had already organised at least one post Jomtien national level EFA policy meeting; and the same number had organised some form of public information campaign. Most countries (87%) had announced EFA goals and nearly all (90%) had a strategy or plan of action to achieve EFA. However, only 12 countries (105) had reported a significant increase in the allocation for basic education in the national budget since Jomtien; and 34 countries (28%) had organised meetings with agency agencies (excluding the usual bilateral contacts) to seek external funding for basic education. JICA, in assessing whether or not they should move towards SWAps, reviewed the status of SWAps in 12 African countries, in terms of the strength of local institutional capacity both centrally and locally, the number of other major agencies and the extent of agency co-ordination between themselves and with the recipient government. They found that there is no apparent relationship between the strength of institutional capacity and the extent to which the recipient country's government is involved in co-ordination mechanisms; an interesting commentary on the practical importance of the country's institutional capacity and on the criteria advanced by, for example, USAID as to when a SWAp is appropriate.

Agency co-ordination ought to be driven by the governments of the partner countries; but agency co-ordination is complicated by various philosophies, agendas, modalities and bureaucratic procedures (e.g. disbursement mechanisms, technical assistance recruitment, and reporting practices) of the different agencies and funding agencies (USAID 1998).

The process of implementing SWAps should be carefully assessed to ensure that the outcome is that the nature of the burden on government evolves from managing relations with a plethora of agencies to managing the core business of government.

The formal institutions for co-ordination outside the country context are well known.

OECD Development Assistance committees with regular three monthly meetings although education is only occasionally on the agenda.

IWGE (International Working Group in Education) sponsored by UNDP, UNESCO, World Bank with the secretariat at IIEP (International Institute of Education Planning) with regular 6 monthly meetings. It brings together representatives of the main agencies and ministries of education from Sub-Saharan Africa.

ADEA (Association for Development of Education in Africa) with regular (6 monthly) meetings about progress in respect of the problems and opportunities for the development of education in Africa. It has ten working groups focusing on different areas of policy.

Horizon 2000 Meetings of Education experts organised between the EU Commission and Member States which has produced examples of best practice and the Code of Conduct for agencies working within the education sector.

The latter three groups have grown up in the last decade and have been substantially influenced by the Jomtien agenda. The secretariat of the IWGE has published three reviews of trends in aid based on these meetings and they include statement of each agency's position. ADEA is an even more open forum in that it invites (academic) researchers to present and discuss at its meetings. In addition to these formal fora, there are often informal meetings:

EU-Africa/Africa-Africa inter university co-operation (Benin April 1999; Anglophile, Africa)

New Zealand recommended that regular trilateral meetings between the Cook Islands Government, New Zealand and ADB be formalised according to an established procedure and timetable.

The problems with purely project support to one aspect or another of the education system or to the basic education sub-sector are almost universally recognised. The late 1980s and 1990s have seen a shift, partially borne out in spending, in favour of sectoral 'packages' of support. These potentially offer significant opportunities to deliver support to education that is institutionally and financially more sustainable, and reflecting an increased recognition of the importance of ownership by country partners.

However, these advantages will only be delivered in practice if key challenges are overcome. These include ensuring the involvement of private sector, NGO and community actors in the process of design and implementation, in order to ensure solutions meet needs and the dangers of excessive centralisation are mitigated. Attention to stimulating bottom-up demand is likewise required. Further elements include the extent of partner country "readiness" for comprehensive sector support through the budget (including appropriate financial systems, such as a medium term expenditure framework). Effectiveness will also hinge on the ability of agencies to subsume their individual ambitions within a nationally developed framework of priorities and actions.

From the partner government's point of view, sectoral approaches involve massive reform challenges which usually appear on the agenda simultaneously. The need for financial accountability in sector wide approaches places considerable demands on national financial and budgetary systems. A step-by-step approach allowing space for partner governments is required, since SWAps may require the implementation of a comprehensive cross-sectoral and sometimes inter-sectoral change. Capacity building and institutional development measures need very careful consideration to ensure that they do not add to the strain by taking away key officials for training.

The move towards SWAps, whilst welcomed in principle by many agencies, has posed problems for them: firstly in terms of accountability; and second in terms of the resources required to implement their aid programmes. The more intensive policy dialogue in-country has meant that the in-country embassy staff of each agency have to be aware and conversant with the sectoral wide approach in general and those adopted by their bilateral colleagues working in the same country in particular. In turn, this has led to many agencies themselves going through a process of decentralisation of their central office staff. One smaller agency commented:

"The move towards SWAps is causing increasing difficulty to the smaller agencies in terms of the increased amount of in-country co-ordination required. The proliferation of fora for external co-ordination has not necessarily helped this problem".

Further challenges include the danger that sector-wide approaches imply an emphasis on the education sector as a whole, with the attendant risk that this may lead to a downgrading of basic education within that. Several agencies have mentioned re-introducing support to tertiary education as part of the package of support to basic education. However, this may equally be an opportunity as well as a risk. An effective approach might involve working to improve the policy and institutional environment of the education sector as a whole, but underlining the importance of allocating very significant (and usually increased) resources to basic education within that.