SPECIAL REPORT No 7/98 in respect of the European Community Development Aid Programme regarding South Africa (1986-1996) accompanied by the replies of the Commission pursuant to the second sub-paragraph of Article 188 C (4) of the EC Treaty (98/C 241/03)

CONTENTS

>TABLE>

>START OF GRAPHIC>

Summary

Between 1986 and 1997 the European Community provided 830,2 Mio ECU in its aid programme to South Africa. Initially, and notably from 1986 to 1991, the programme funded projects to assist the victims of apartheid and was carried out in the face of active opposition from the South African authorities. It succeeded, nonetheless, in making a positive contribution to peace and democracy. As the political situation in South Africa evolved, the programme continued to support democracy and human rights, and started to promote social and economic development. In 1994 the democratic government was established, the Commission Delegation was opened in Pretoria and a cooperation agreement between South Africa and the Community was signed. The European Programme for Reconstruction and Development (1994-99) was started.

This report examines how the programme was implemented and managed by the Commission and other organisations involved. It focuses mainly on the period 1993-97.

Many of the key findings in this report concern weaknesses on which the Court has previously commented in other reports on development cooperation programmes. They are, therefore, of general relevance, and the Commission should, without delay, take steps to overcome them.

As in other parts of the world, the implementation of the programme has been adversely affected by there being too many layers of decision making and administration, with too many matters having to come up to Commission headquarters in Brussels for final decision. This was one of the main causes of delays and deficiencies in project implementation and monitoring. In particular during the first phases of the European Community and South African cooperation, it was found that projects were often poorly formulated, timetables were unrealistic, and some conditions in contracts and financial agreements were inappropriate. Treasury management by the Commission was poor. Reporting procedures were inadequate, and monitoring by the Commission was weak. Apart from the global evaluation of the European Special Programme carried out in 1996, there was insufficient evaluation of the effectiveness of the actions financed through non-governmental channels.

The Commission underestimated the institutional strengthening that was needed by its South African partners to enable them to implement the programme more effectively and efficiently. When the Commission began the programme with the new government in 1994, like other donors it was unrealistic about the volume of funds that could be efficiently and effectively absorbed in the short term by the South African administration. It would also have been preferable to concentrate resources on fewer sectors initially.

The European Programme for Reconstruction and Development is delayed in its implementation, and disbursements are much lower than planned. The Bursary Programme 1993-95, which was one of the main Community actions, was less effective than it could have been because of poor management and monitoring. A backlog of more than 750 completed or inactive contracts going back to 1986 have not been properly closed by the Commission.

Throughout the period the resources allocated by the Commission for the management of the programme, and their organisation, were inadequate. There was a lack of appropriate management information systems and procedures were not clearly defined.

The Commission should establish a more decentralised management approach that places decision making authority at the most appropriate level in the system and utilizes the limited resources of the Commission for managing and controlling the use of Community funds in an optimal manner. Operational and financial responsibilities should be decentralised away from the central services in Brussels to the field, and management resources should be reallocated accordingly. The Commission should simplify procedures and reduce delays.

The Commission should address more realistically the constraints in its own services which affect programme implementation and the absorption capacity of its South African partners. It should pay greater attention to strengthening the administrative capacity of its South African partners, whether governmental or non-governmental.

Donor coordination should be strengthened, especially with, and between, the Member States. In future trade and cooperation agreements, the Commission and Member States should ensure that the Union's commercial policy is supportive of their strategy for development cooperation.>END OF GRAPHIC>

INTRODUCTION

Background on the European Community Development Aid Programme regarding South Africa

1. In September 1985, in response to the intensified political repression in South Africa, the Foreign Ministers of the European Community (EC) adopted a twin-track policy of restrictive and positive measures to foster peaceful transition to democracy (1). The restrictive measures involved a trade embargo, and economic and political sanctions. The positive measures launched in 1986 involved support financed from the general budget of the European Community to victims of apartheid and disadvantaged communities (2).

2. From 1986 to 1997, a total of 830,2 Mio ECU were committed to the EC aid programme, for projects in a wide variety of sectors. The programme passed through different phases taking into account changes in the political environment, and the introduction of new objectives and arrangements for implementation. Three different conceptual phases can be identified: 1986-1990 (Special Programme to Assist the Victims of Apartheid), 1991-November 1994 (Special Programme for South Africa) and November 1994-1999 (European Programme for Reconstruction and Development in South Africa). Clear cut-off lines between these phases cannot be drawn: project execution timeframes overlapped these periods and projects were integrated into successive programmes. Figure 1 sets out chronologically the political events that occurred, the measures undertaken by the European Community in response to these events, and the legal and operational arrangements that were established. Figure 2 provides further information on each phase.

>REFERENCE TO A GRAPHIC>

>REFERENCE TO A GRAPHIC>

>REFERENCE TO A GRAPHIC>

FIGURE 2

>START OF GRAPHIC>

EC development aid scheme regarding South Africa

Phase I - 1986-1990

Special Programme for the Victims of Apartheid (SPVA)

Aim:To foster transition to a peaceful, stable, democratic and non racial South Africa by assisting the victims of political repression.Main areas targeted:- Humanitarian & social initiatives;

- Training & education;

- Legal assistance.Implementing organisations:- European NGOs.

- South African Channels: SACC, SACBC, KT & ICFTU.Organisational arrangements:The Commission was not involved in the definition, appraisal, monitoring & implementation of the projects.

Legal responsibility laid on European NGOs.

Appropriations committed108,6 Mio ECUPhase II - 1991-Nov 1994

Special Programme to South Africa (SPSA)

Aim:To foster transition to a democratic and prosperous South Africa by integrating development objectives and supporting disadvantage communitiesMain areas targeted:- Tertiary & adult education;

- Vocational training;

- Primary health care;

- Community & rural development;

- Human rights & democracy.Implementing organisations:- South African Channels: SACC, SACBC, KT & ICFTU

- No intervention of European NGOs.Organisational arrangements:The Commission was involved through the Project Coordination Office (PCO) in the identification. assessment and monitoring of projects.Appropriations committed:230,7 Mio ECUPhase III - Nov 1994-1999

European Programme for Reconstruction and Development (EPRD)

Aim:To foster transition to a democratic and prosperous South Africa by supporting policies and reforms undertaken by the South African Government under the Reconstruction and Development Programme 1994-1999Main areas targeted:- Education & training;

- Health;

- Rural & Urban development;

- Economic co-operation, trade & investment promotion, support for small, medium and micro enterprises;

- The promotion of good governance and democratisation.Implementing organisations:- Central, provincial en local government structures;

- Transitional National Development Trust (TNDT) & local NGOs.Organisational arrangements:The Commission is directly involved in project management and monitoring through the EC Delegation in Pretoria.Appropriations committed:479,5 Mio ECU (at 31.12.1997)Source: Court of Auditors. Evaluation of the European SPSA (SPM, 1996), accounts of the Commission 1986-1996.>END OF GRAPHIC>

3. The first phase covered the period 1986-90. A total of 108,6 Mio ECU was committed to over 400 projects in areas of humanitarian and legal assistance and education. The vast majority of these projects were promoted by three local organisations called the 'Channels` (3): the Kagiso Trust (KT) an interdenominational body formed by the churches and trade unions in South Africa, the South African Catholic Bishops' Conference (SACBC), the South African Council of Churches (SACC) and, to a lesser extent, by the European-based International Conference of Trade Unions (ICFTU).

4. During this phase there was no direct link between the Commission and the Channels: contractual responsibility towards the Commission for the projects lay with European non-governmental organisations (NGOs) which were responsible for forwarding the funds to the Channels in South Africa and for financial and narrative reports. The programme was carried out in the face of active opposition from the South African Government and the beneficiaries were normally opponents of the regime and subject to harassment, banning orders and imprisonment. The Commission adapted its normal financial control and reporting requirements to the realities of the situation and treated the projects with a high degree of confidentiality.

5. In 1990, the political situation in South Africa evolved positively and the trade embargo and economic and political sanctions were gradually lifted. In 1991, the EC aid programme entered a second phase and was reorientated to take into account development objectives and new sectors. The involvement of European NGOs came to an end and the Commission concluded project contracts directly with the Channels in South Africa. The political situation did not yet permit the establishment of an EC Delegation in Pretoria. The Commission, however, opened a Programme Coordination Office (PCO) in February 1991 staffed by technical assistants under European Association for Cooperation (ECA) contracts, in order to provide support to partners and to monitor projects.

6. The third phase of the EC programme of assistance to South Africa began in November 1994 following the democratic elections, the establishment of the Government of National Unity and the conclusion of the Cooperation Agreement between the EC and the Republic of South Africa (4). The aim of this phase, called the European Programme for Reconstruction and Development (EPRD), is to contribute to South Africa's harmonious and sustainable development and to consolidate democratic society. The South African Government is now the interlocutor of the EC in the development policy dialogue. Project implementation has been opened as a priority to central, regional and local government departments; further, instead of the traditional Channels, local NGOs are financed directly for sectoral programmes and a Transitional National Development Trust (TNDT) (5) has been established to direct funds to local NGOs for small projects.

7. In December 1993, the agreement for the establishment of the Delegation of the EC in South Africa was signed. The Delegation offices were opened in Pretoria in May 1994 and the PCO was closed in December 1994. Finally, in November 1996, the Council adopted a regulation on development cooperation with South Africa (6), which set out the legal basis for the EPRD operations; 500 Mio ECU was earmarked to finance the EPRD during the period 1996-99. The National Indicative programme was signed by the South African Government and the Commission in May 1997.

8. Since November 1994, EC-South Africa relationships have also developed in other political, commercial and economic areas. In June 1995 negotiations of a bilateral trade agreement for the establishment of a free-trade area were opened. Such an agreement is of major importance for the development of the South African economy, progress has been slow and discussions were ongoing at the date of adoption of this report. In April 1997, South Africa became a member of the Lomé Convention concluded between the EC and the Africa, Caribbean and Pacific (ACP) countries (7).

The Court's audit policy towards EC's South Africa Aid Programme

9. In 1991 the Court reviewed the Commission's administration of the programme to date and identified, after taking into account the management and control constraints under which the Commission and its partners were working, a number of weaknesses: incomplete files; lack of progress reports; the weak role of the European NGOs in project implementation, and poor financial management (over-funding of projects, large outstanding balances and non-recovery of unused amounts).

10. In 1993 and 1994, the Court examined the financial accounts regarding the South African aid programme. The Court pointed out to the Commission that financial management was weak: the Commission did not have up to date information on the volume of funds in the bank accounts of the Channels; unnecessarily large amounts were being held in these accounts and there was no policy for the treatment of interest earned. In addition, there was a backlog of more than 750 projects dating back to 1986 which had not been closed and, in some cases, final payments were still outstanding. The services of the Commission appeared not to have the capacity to deal with this.

11. The findings of the above audits were communicated to the Commission but were not published in Court reports. In order to follow up these findings, and the Commission's response to them, in 1996 the Court started a comprehensive audit of programme to South Africa.

12. A mission to South Africa was carried out in 1997, involving meetings and documentary checks at the EC Delegation, at the Government Departments, at the Channels organisations, at the NGOs and at other organisations. A sample of projects were visited on the spot, final beneficiaries were interviewed and physical achievements were checked.

13. The Court's enquiry covered the following elements:

(a) the procedures and systems applied by the Commission and by the Channels for managing projects approved during the period 1993-94 which were still being implemented in 1997 (paragraphs 20 - 52). Particular attention was paid to the Bursary Programme implemented by Kagiso Trust, which was the largest EC operation carried out in South Africa (Annex);

(b) the procedures and systems applied by the Commission, South African Government departments, NGOs and other organisations for projects financed under the EPRD (paragraphs 53 - 85);

(c) the organisation of the Commission services (paragraphs 86 - 92) and how the co-ordination between the Community and the Member States in the sphere of development is working in South Africa (paragraphs 93 - 97).

14. In the following pages, a number of shortcomings in the financial management and control of the EC's aid programme are set out. This notwithstanding, it should be emphasised as a general conclusion that, in spite of the difficult conditions in which the programme was carried out, the EC was able, during the first two phases, to support projects which contributed to the democratic transition in South Africa and which relieved victims of apartheid and disadvantaged communities (8). With the start of the EPRD, a new era for development cooperation with South Africa is beginning, an era in which basic principles of cost-efficiency, financial control and accountability must be observed by both the Commission and the South African partners.

IMPLEMENTATION OF THE BUDGET

15. The impact on the budget of aid to South Africa for the whole of the 1986 to 1997 period is shown in Figure 3. Out of the total of 818,9 Mio ECU committed between 1986 and 1997, only 456,0 Mio ECU (56 %) was disbursed. At the end of 1997 the commitments still to be settled, consisting mainly of projects under the EPRD, totalled 352,6 Mio ECU (9).

>TABLE>

>REFERENCE TO A GRAPHIC>

16. Over the last three years the amount of commitments remaining to be settled rose markedly. This highlights a number of interrelated points which are examined further in this report. The increase with the introduction of the EPRD in the availability of commitment appropriations has run ahead of the capability of the Commission and the South African authorities to spend the funds. The absolute level of payments declined considerably from 1994 to 1996, despite the continued increase in payment appropriations available. The rate of utilisation of payment appropriations in 1995 and 1996 was very low, reflecting the problems encountered in implementing the EPRD. The budgeted levels of commitment and payment appropriations indicate that the Commission did not anticipate these problems.

17. Figure 4 shows the evolution of the EC commitment and payment appropriations granted to the Channels between 1993 and 1996; this figure shows the drastic decline in 1995. Figure 5 sets out the payment appropriations by Channel; KT was the largest beneficiary of the EC funds (82 % of payment appropriations), followed by the SACBC (14 %). Figure 6 presents the sectoral breakdown of the projects; the most important sectors, which totalled 77 % of the commitments, were the Bursaries for tertiary education, Community development and Education and vocational training.

>REFERENCE TO A GRAPHIC>

>REFERENCE TO A GRAPHIC>

>REFERENCE TO A GRAPHIC>

>REFERENCE TO A GRAPHIC>

18. The variation in Community commitments and payments for the new programme (EPRD) between 1994 and 1996 is shown in Figure 7. This shows how the growth since 1994 of unspent commitments is almost entirely explained by the situation of the Government projects. Of the 16 Government projects funded during the period, only five had actually got underway by the beginning of 1997. Figure 8 shows how, following the normalisation of development cooperation with South Africa, the financing of NGO projects assumed a more modest role, falling from 38 % in 1994 to only 11 % in 1996. However, it should be noted that implementation of the projects that are directly cofinanced with the government is, in fact, partially delegated to NGOs. In the indicative programme signed in May 1997 (10) the amount earmarked for decentralised cooperation accounts for 25 % of the EPRD annual financial package.

19. The sectoral allocation of the EPRD programmes funded between 1994 and 1996 is shown in Figure 9. Projects in the 'Good governance and Democratisation` and 'Education and Training` sectors are the most significant.

>TABLE>

SYSTEM FOR IMPLEMENTING PROJECTS THROUGH THE CHANNELS

Introduction

20. The Channels system which operated during the first two Phases of the EC aid programme has been replaced by the arrangements under the EPRD. During 1997, however, there remained some projects started in 1994 and 1995 which were still being implemented. The key observations on the Channels system are presented here in order to highlight those weaknesses in the system which should be avoided in future Community programmes, whether in relation to South Africa or elsewhere.

21. The observations presented below deal mainly with procedures, systems and management arrangements. There is a need also to consider the achievements and the impact of the programme run through the Channels system. There is, however, little firm information on which to base an assessment, because few evaluations have been carried out. Apart from one overall evaluation covering the period 1986-94 carried out for the Commission by consultants in 1995 (11), out of more than 700 projects, only four have been evaluated including the Bursary Programme (12). No sectoral evaluations have been carried out.

22. The consultants who carried out the overall evaluation in 1995 reported that, in general, the Special Programme up to 1995 was a success, relevant to the needs of the target population, had had a significant positive impact on political developments in the country, and contributed to the demise of apartheid. They considered that the Programme was particularly successful during the early years (before 1991), when projects were initiated from below and there was a high degree of involvement in and 'ownership` of projects by beneficiaries, with high levels of creativity and enthusiasm. It was noted, however, that, notably after 1991 and as far as operational procedures were concerned, there was much irritation and frustration stemming from the involvement of too many parties in the management chain, a lack of common understanding of who had authority over what, unclear and misunderstood procedures, and a shortage of staff. The Commission was seen as a slow, bureaucratic organisation, which took too long to approve projects and process payments. Reporting requirements were unclear, involved duplication of work, and generated little substantive feedback.

23. Figures 10 and 11 sketch the responsibilities of the different parties for approving, paying and implementing projects funded through Channels. These procedures developed in an ad-hoc manner, according to the clauses in the standard project contracts concluded between the Commission and the Channels.

>REFERENCE TO A GRAPHIC>

>REFERENCE TO A GRAPHIC>

24. The distinctive feature of the system was that there was no direct link between the European Commission and the project holders; all relations passed through the Channels which were the sole interlocutors with the Commission. This entailed filtering funding applications, concluding contracts, managing and disbursing EC funds, monitoring and assisting holders and reporting progress.

Observations concerning the Channels system (13)

Design weaknesses

25. The Channels system had key weaknesses which hindered adequate control and sound financial management and which caused shortcomings in the implementation of the projects. Firstly, there were too many organisations and levels of the Commission services involved in the administration of the system (project holders, the Channels, the Commission Delegation and Central Services) with overlapping responsibilities and insufficient information flows and coordination. Decision making was slow, with the result that both payments and the implementation of actions were delayed.

26. Secondly, the Channels and the Commission over a short period had to identify, appraise, execute and monitor a large number of small projects (14). Most of them were managed by NGOs and Community Based Organisations (CBOs) with limited administrative capacity. As a result, the number of decisions to be taken, checks and financial transactions to be handled was very high. The Commission did not have adequate resources to deal with this workload.

27. Finally, the results of the examination of the Channels system highlight some of the endemic weaknesses in the Commission's management of development cooperation programmes on which the Court has commented repeatedly in its reports.

Formulation of projects

28. From the outset projects were poorly formulated: essential elements (such as the budget, total donors' contributions and a timetable for the activities) were not clearly defined or were completely absent, terms of reference were incomplete and confused. Despite attempts by the Commission to improve matters by introducing in 1993 the use of logical frameworks (15) for project design and implementation, these weaknesses remained up to the end of the Channels system.

29. Unrealistic implementation timetables in the contracts were the source of problems. Almost all projects requested the extension of the original timetable on several occasions. The Commission approved these extensions without analysing what had been achieved, the plan to complete the projects or the real possibilities to respect the new deadlines. From the documentation contained in the Commission's files it was not always possible to check that the contracts were extended in accordance with the proper procedures and that deadlines fixed for the financial commitments had been respected. Projects were identified for which the timetable was extended after the initial one had expired.

30. Extending contract deadlines placed a considerable administrative workload on the holders, the Channels and the Commission. For example, in 1996 192 extension operations were processed through the system, with each one having to be approved in Brussels. Similarly, the contracts did not provide for any procedure for modifying the project budget; minor changes had to be approved by the Commission's central services in Brussels and be the subject of contract riders. This led to payment and operational delays.

Financial and treasury management

31. In 1994 and 1995 the Court pointed out in letters to the Commission that it had not monitored adequately the management of the funds transferred to the Channels. It had failed to ensure that the Channels submitted financial reports reconciling the funds transferred by the Commission to their EU project bank account, the amounts disbursed to the project holders, and the interest accrued. Excessive balances had built up notably in the KT project account, and considerable interest earned, about which the Commission, both in Brussels and in South Africa, had wholly inadequate information. Following the intervention of the Court, the Commission took some action. By 1997 the situation had improved in the case of the largest Channel, KT, but not in those of the other Channels.

32. The contracts with the project holders provided that payments were to be made in ECU and converted into local currency (ZAR) applying the monthly rates published in the EC Official Journal to the expenditure incurred in the same month. This provision was both inappropriate and impractical since it would have obliged the Channels to keep a bank account in ECU (not permitted by the South African monetary authorities), and to identify the monthly payments for each project, on which to base disbursements. In practice, the Channels converted funds received from the Commission immediately into ZAR and deposited the proceeds in a local currency bank account. Although the Commission was aware of this, it did not modify the terms of the contracts accordingly.

Payment and reporting

33. Project payment and reporting procedures in the contracts were deficient and did not permit adequate monitoring by the Channels and the Commission.

34. The contracts with the project holders neither specified the period to which reports should refer, nor fixed deadlines for submission. Reporting was exclusively linked to payments; as a result, the project holders only provided reports when they requested disbursements. If an organisation did not request a payment, no report was provided. In addition, the contracts did not set out sufficiently what the narrative and financial reports should contain.

35. The majority of the Channels' projects involved the funding of the recurrent costs of the beneficiary organisations. In many cases, however, reporting problems arose because the contract period did not correspond to the financial year of the organisation, which is the normal period for accounting for and reporting on the activities financed by the contract. Further, when in effect the Commission was providing an annual grant to an organisation, it signed a new contract before the previous ones had been reported on and cleared. For some organisations, the Commission had open at the same time separate contracts funding different years.

36. Reporting required cumbersome and complicated reconciliations. The difficulty of financial reporting meant that the Commission's services often rejected the reports submitted. This caused frequent interruptions and delays in payments, leading to uncertainty for the holders, jeopardizing the execution of activities and hindering effective monitoring.

37. An analysis of project execution at 31 December 1996 shows that:

(a) of the 189 projects approved between 1993 and 1995, around 75 had the last instalment outstanding for several years. This could be an indication that the projects were overfunded, and that there are unused balances at the project holder - it is part of the closure exercise (paragraphs 40 - 52) to identify this. On the other hand it could indicate that in some cases holders preferred to lose the final 10 % payment rather than produce a final report. From the point of view of the holder this may be seen to be a reasonable option, taking into account the small payment due and the trouble of producing and getting the reports approved;

(b) more than 50 projects had not submitted or had not had the interim report approved; these were projects which had received an advance of 50 % or 60 % of the project amount on the signature of the contract but for which the second instalment had not been paid.

38. Most of the project reports submitted were accompanied by audit reports on the accounts. These reports were, however, of limited value, since they did not tackle the key problems relating to the EC funding; for example, the period to be covered by the report, other grants and double financing. In the case of the Bursary Programme, the audit exercise, the key control mechanism, was of limited value because of the lack of adequate terms of reference to the auditors and the absence of follow-up of findings in the audit reports (see Annex, paragraph 4).

Project monitoring

39. The Commission did not monitor the projects adequately. This was because of the weaknesses of the reporting procedures mentioned above, the lack of project tracking computer tools and failures of the Commission information system.

40. The monitoring of SACC and SACBC projects, for which very little information is available, was particularly weak. Of 14 SACC and SACBC projects approved in 1993 and 1994 and totalling 7,66 Mio ECU, the documentation held in the Commission's files was insufficient to assess the current status of the projects. Indeed the Commission did not know whether 8 projects of the 14 were still active. In only 2 cases, progress project reports had been submitted. In spite of the lack of information, the Commission had extended project implementation time frames on several occasions.

41. The Commission's monitoring of KT projects was also inadequate, although some improvements were made after 1995.

42. As a consequence of weak monitoring, the Commission was unable to react promptly to holders' and Channels' queries. Numerous projects suffered delays, were in a stand-by situation for long periods, had been definitively halted or were unable to meet their objectives. As far as the Bursary Programme was concerned, lack of supervision and of reliable figures reduced effectiveness and sustainability. Although around 34 000 bursaries were granted, the opportunity to grant 8 700 more was missed because of inefficient management, and more than 72 Mio ZAR (approximately 12,5 Mio ECU at December 1996 exchange rates) in reimbursements were not paid into the SA national bursary scheme (see Annex, paragraph 17).

43. The ability of the Commission to monitor adequately the implementation of projects was weakened by the poor state of its information systems. It did not have a satisfactory computerized database containing details of project implementation and reporting (16). Documentation in project files was not up to date, and reports and correspondence were sometimes missing. Information between the Channels, the Commission Delegation and Central Services was not always passed on. The Commission did not have a copy of all the contracts signed between the Channels and the project holders, and thus was not aware of those cases where the project scope and financial conditions differed from its contract with the Channel.

44. A financial consequence of unsatisfactory monitoring was that projects which were approved but which never started, or projects which started but became inactive, still remain unnecessarily opened in the Commission accounts. Thus about 10 Mio ECU in respect of projects committed between 1993 and 1995 but which never started, or became inactive, remained in the accounts at 31 December 1996, overstating outstanding liabilities.

Closure of Contracts

45. Following audits in 1994 and 1995, the Court pointed out to the Commission that it should give priority to closing a large backlog of more than 750 contracts, in respect of which a total of 441,7 Mio ECU had been committed, going back to 1986. The Commission informed the Court at the end of 1995 that it aimed to close some 350 outstanding contracts by July 1996, and that a technical assistant had been employed to reinforce the capacity of its services to carry out this task.

46. In January 1996 the deadline for closure of the 350 projects was extended to end-December 1996. The overall target was then to close 604 projects out of the 664 projects dating from 1986-93 by the end of 1997.

47. By the end of June 1997, however, none of the contracts in the backlog had been closed. The Commission encountered serious procedural difficulties during 1996, largely because of inadequate planning and definition of the task, and the failure to devote sufficient resources to it. A review by the Court of a sample of 250 closure sheets that were prepared in 1996 showed that there were numerous unexplained anomalies, and that the supporting information presented was quite inadequate for decision-making purposes.

48. In order to increase the capacity of the services to close the old projects, the Commission decided in March 1996 to engage private sector audit firms. It took the Commission, however, more than a year to engage these auditors, contracts with two separate firms finally being signed in May and July 1997 respectively.

49. During 1997 the services concerned established revised procedures requiring more justification for the actions proposed, which include recovery or write-off of outstanding balances held by implementing organisations, and the cancellation of unneeded commitment balances. New targets were fixed, with final closure of all non-active projects from the period 1986-1994 to be achieved by the end of 1998.

50. A corollary of the lack of progress in closing contracts is that little progress was made during the closure exercise in recovering surplus funds held by NGOs in Europe or in South Africa. During missions to European NGOs in 1995 and 1996 such surpluses were found in a number of cases. Contrary to the Commission's own internal rules, however, which require that recovery orders be established as soon as a debt is identified, with a 'pm` mention if the amount cannot be precisely evaluated, recovery orders were only issued in respect of one NGO.

51. The problems encountered by the Commission in closing this large backlog of projects reflect the weaknesses of financial management of the South African programme. Consistently, and despite the size and political importance of the programme with South Africa, the Commission did not allocate sufficient resources and did not have sufficiently robust and effective management procedures so as to enable the new and expanding programme to be prepared and implemented, at the same time as the closure exercise. The priority accorded to the new programme always meant that the limited resources were drawn away from the closure exercise.

52. Further, the closure exercise was badly prepared, with inadequate objectives and milestones, procedures were unclear, roles and responsibilities of those involved were not sufficiently defined and agreed, and there was no common understanding on what was needed to enable projects to be closed. As a result, the limited resources available were not used efficiently or effectively.

THE EUROPEAN PROGRAMME FOR RECONSTRUCTION AND DEVELOPMENT (EPRD) 1994-1996

The EPRD system

53. Following the democratisation of the country a new approach to cooperation became necessary. At that point it became a matter of funding a programme of direct cooperation with the South African government.

54. In a climate of considerable optimism all the donors, including the European Community (EC) and the Member States, were eager to launch large-scale development programmes in association with the government. The EC, the largest donor (17), had supported the South African forces for democracy throughout the apartheid era and now wanted to increase its support for the new government in a legitimate way. Within a relatively short period the Commission thus approved several programmes for substantial amounts in several areas of activity.

55. This was why, in 1994, the EC, like other donors, joined the reconstruction and development programme launched by the South African government. More than 5 000 million ZAR (approximately 1 150 Mio ECU) annually from the national budget were pooled with international aid to form the RDP Fund, with a specially-created central administrative structure, the RDP Office. However, the latter soon proved to be unsuitable and in March 1996, following criticism of the low level of programme implementation, the RDP Office was closed and its responsibilities for programme implementation transferred to the Ministry of Finance and the sectoral ministries.

56. At that time South Africa's central administration underwent major restructuring, with important changes at the top levels. The government departments which initiated the execution of development programmes had only limited experience of working with international aid organisations. The South African administration faced serious difficulties in bringing its systems and procedures into line with the various donor requirements. New structures had only just been created at provincial and local level and others were still at the planning stage.

57. In this context, the EC - and other donor organisations - should have been more realistic about the volume of funds that could be efficiently and effectively absorbed by the South African administration. In the case of individual programmes, such as the health sector and the Land Reform Pilot Project, the Commission should have recognised that until the provincial and district structures through which the programme was to be implemented were established, the funds for the programme could not be absorbed. Further, in light of the limited resources and sectoral capabilities of the staff of the Commission Delegation, it would have been preferable for the Commission, with the agreement of the South African authorities, to have concentrated fewer resources on a more limited number of sectors.

58. At the time when the programme was launched in 1994 the Commission was, in fact, also facing serious problems. The headquarters department in Brussels was required to simultaneously monitor and close the projects funded via the Channels (paragraphs 20 - 52), programme the EPRD and examine the related files and negotiate the legal framework for the new relationship of cooperation with South Africa (18). There was a similar lack of resources at the Delegation (paragraph 90).

59. The problems of managing such a broad programme with the limited resources available led the Commission to change its approach to the financing of NGO projects and activities. The NGOs were encouraged to get together and to combine their applications for finance for action in a particular sector - health, urban development, rural development, etc - in programme form.

60. In adopting this approach the Commission should, however, have paid particular attention to the organisational structures responsible for implementing it and ought to have ensured that they had the necessary skills and resources. Of the three sectoral programmes audited by the Court two had substantial weaknesses at the level of the structures responsible for the overall programme:

(a) in the case of the 'Urban sector network` programme, the Coordination Office (PCO) proposed a coordinating structure in the form of a NGO which had neither internal rules nor adequate resources and was unable to monitor the programme components regularly;

(b) in the case of the health sector programme, the activities of three NGOs operating in the same region were combined under one contract. Responsibility for managing and monitoring the whole programme was unilaterally assigned to one of the NGOs by the Commission. It was difficult for the former to produce the required progress reports, as the other partners had different management systems and structured their accounts differently. Moreover, the NGO did not have the resources required to monitor the operations.

61. In the case of the smaller projects carried out by the NGOs and the community-based organisations (CBOs), the Minister responsible for the EPRD was instructed, in 1994, to find a solution that would allow these projects to be funded within a coordinated framework and proposed the creation of an independent body, TNDT. However, the establishment of TNDT was slow and the funding of NGOs and CBOs was severely disrupted in 1995 and 1996. Some of them were obliged to cut back on staff and activities.

Government projects

62. In the health part of the first series of programmes cofinanced with the Government, in 1994, the Court audited the National HIV/AIDS Programme, the District Health Systems Support Programme and the Technical assistance programme for restructuring the national health system, which amounted to 19,4 Mio ECU in total. At the time of the audit none of the projects had progressed very far. The programmes had, in fact, been started before the structures that were essential for their implementation had been put in place. For example, the districts on which the new health system was to be based were not fully established until the second half of 1996, which delayed the actual introduction of the measures at the province level. It also resulted in the South African authorities asking for the execution phase of the three programmes to be extended to the end of 1997.

63. The Human rights programme was approved by the Commission in December 1994 on the basis of a study carried out by external consultants. The financing agreement with the South African government was for an amount of 15 Mio ECU and was not signed until May 1996. This government programme is being carried out by NGOs and a special foundation was set up to administer it. The programme was launched after a very long delay and in March 1997, at the time of the Court's audits, the project had barely got under way.

64. The aims of the project are somewhat vague. The 1996/1997 action plan was based on the results of the study mentioned above, which was carried out in 1994; the amount of expenditure required was very much an estimate and was not supported by a breakdown of the calculations. The situation in the country changed between 1994 and 1996 and updating of the objectives and resources to be employed was found to be necessary. This assessment was acknowledged by the people responsible for implementing the programme, who had also included an initial exploratory phase of clarifying needs and objectives in their action plan. However, they thought it important for the project to start, and proposed to make the necessary adjustments at a later stage rather than wait for the results of this first stage. The Commission should monitor developments on this programme carefully.

65. The Parliamentary Support Programme (total cost 36 Mio ECU) is cofinanced by three partners. The actual start of the programme was planned for the end of 1996. In March 1997 it had still not started because the staff required to manage it had still not been fully recruited. Moreover the actual arrangements for implementing this jointly-funded programme were not satisfactory. At the time of the Court audits the three donors' contributions had not yet been combined within a single budget. It was, therefore, not possible to understand and verify how the South African government's contribution (19,35 Mio ECU) was to be incorporated and allocated to the programme. The work programme which had recently been tabled by the South African authorities covered only the Community contribution (16 Mio ECU). It is important that the Commission verifies the reality of the joint funding for this programme.

NGO projects

66. For the NGOs these were the first projects to be funded directly by the Commission, without the intermediary Channels. Their lack of experience of Community procedures, the difficulty of obtaining the necessary advice from the Delegation and the problems of communicating with Commission headquarters produced start-up delays for all the projects visited by the Court. NGOs must present an annual work programme in order to obtain payment of Community funds, but they had not been given adequate instructions (notably in the financing agreements) regarding the nature and level of breakdown of the information required. In most cases this resulted in the Commission's referring the files submitted back to the NGOs. In some cases programmes were accepted by the Delegation after lengthy exchanges of information with the NGOs, only to be rejected subsequently by Brussels, so that they then had to be resubmitted. This two-tier approval procedure is not an efficient use of the Commission's limited project-management resources and it also caused excessive delays in implementation.

67. The Participatory Health Care Programme (1,12 Mio ECU) covered the essential needs of the most disadvantaged people in the townships around Cape Town. The initial stages of the programme were particularly affected by the weaknesses mentioned in the previous paragraph. In addition, the viability of the programme, which was in the final stage of implementation in March 1997, was also at risk. The contract that was signed with the Commission in fact stipulated that the hospitals for which finance was provided were to be incorporated into the South African government health-care structures, but this was not done. Further funding by the international donors then became essential, so as not to imperil the continuation of the services that had been available till then.

68. The provision of low-cost housing for people with low incomes was one of the main objectives of the Urban Sector Network programme (5,9 Mio ECU). Of the 4 350 houses that were to be provided in the first year only 879 could be handed over, because the Commission and its partner had under-estimated the difficulties of implementing a housing loans policy throughout the country. The progress of the programme was also disturbed by the slowness of the community payments. At the time of the Court's audit certain NGOs in the network had been obliged to suspend the activities associated with the programme pending the disbursement of funds.

69. The start and execution of the micro-projects programme financed jointly with the Microprojects Programme Trust in 1995 (12,6 Mio ECU) were also disrupted by the excessive length of time required for the Commission to sign the financing agreement and disburse the funds. As of March 1997 68 micro-projects had been approved, but the level of progress was very low. The project managers had been obliged to make use of overdraft facilities pending the Commission payment, so as not to hold up execution of the financed measures.

Weaknesses in the tools and procedures for establishing the EPRD

70. Weaknesses in the tools and procedures used by the Commission were the root cause of the substantial delays that were noted at every stage of EPRD implementation and also had a negative impact on execution of the programmes.

Payment delays

71. Most of the projects visited by the Court had been affected by the endemic slowness of payments: in the view of those concerned the Commission was the slowest of the donors. These delays were essentially due to the excessive time taken by the Commission to approve interim reports and updated work plans. Where the Commission is the principal or sole source of finance, partners may experience serious financial difficulties. In the case of the Micro-project and Urban Sector Network programmes the managers were obliged to resort to overdraft facilities in order to avoid suspending project activities.

Exchange losses

72. For several NGOs the amounts received in ZAR were, in general, lower than the amounts due under the terms of the contracts, where amounts are specified in ecus. Analysis of some twelve project payments showed that these differences were the result of a private South African bank applying extremely unfavourable exchange rates (19). On a total of 10,8 Mio ECU of disbursements, the equivalent of 74,9 Mio ZAR at the INFORECU rate, the NGOs received only 72,8 Mio ZAR after conversion at the bank's own rate. The difference of 2,1 Mio ZAR is the equivalent of 0,4 Mio ECU at the INFORECU rate, 3,7 % of the total disbursal. In some cases the difference was as much as 7 %. Comparative analysis of some of the payments for government projects showed that the rates applied by the Reserve Bank were much more favourable. Since these are very significant amounts for the NGOs, the Commission should examine the possibility of adopting alternative procedures, so that NGOs receive the amount specified in the contracts (20).

Financing agreements

73. Poor drafting in the financing agreements (NGO and government) also disrupted the start and execution of the EPRD. The clauses covering, inter alia, payments, reports and project accounts appear in various parts of the agreements and in many cases lack both clarity and consistency. The inconsistency is particularly great in the case of the December 1994 financing agreements, which, although they apply to direct funding of NGOs, included the contract clauses used under the old Channels system. The wording of the financing agreements was not properly checked and material errors were not detected during the approval procedure (21). The financing agreements concluded with the NGOs in 1995 and 1996 only provided for payment of advances of 80 % of the annual budgets and did not cover the procedures to be observed for payment of the year's balance of 20 %. The Court identified at least two payments of balances, for a total of 746 171 ECU (22), which had been effected without the Commission's having identified the omission, and suggested amending the agreements in question.

74. On none of the projects reviewed by the Court had the NGOs provided the bank guarantees required by the financing agreements. The South African banks refuse to give such guarantees for NGOs, because the latter do not have the requisite level of own funds. Instead of bringing the terms of the agreements into line with reality, the Commission asked a bank to issue 'certificates` which were in no wise equivalent to guarantees.

75. As for the projects adopted during the two phases of the Special programme, partners are required by the terms of the financing agreements to convert the month's project expenditure into ecu by means of the monthly rates published in the Official Journal of the European Communities (paragraph 32). This provision is inappropriate and was not followed for any of the projects audited. The Commission should modify these contract conditions with something that is practical and appropriate.

76. The Delegation certainly tried very hard to make up for the drafting weaknesses of the contracts, but it was unable to remedy all the shortcomings. The texts of financing agreements have improved significantly since October 1996, but effort is still required in order to rationalise the provisions of the contracts and, in particular, to resolve the difficulties to which the payment procedures are prone.

77. Major amendments of financing agreements were adopted by the Delegation during meetings with the partners or were authorised by means of a simple exchange of notes (23), without any formal amendment of the text of the agreement. Brussels headquarters, too, often ratified amendments without formally amending the text of the agreement.

78. The Court has already highlighted weaknesses in the procedures for amending provisions of financing agreements, in some of its earlier reports (24). This situation is indicative of weakness in the Commission's control system and is mainly due to the lack of clear instructions regarding the procedure to be followed and, in particular, the division of responsibilities to be observed as between Brussels headquarters and the delegations.

Tendering procedures

79. The Community rules and procedures that apply to project procurement and tendering are different from those applicable to central and local administrations under South African law. A major difficulty is that the South African rules give preference to South African firms, and particularly to those from 'previously disadvantaged communities`, whereas Community rules require open tendering for European Community companies and, under certain conditions, also for other third countries. This was a source of difficulties among the government projects which the Court audited on the spot and, in the particular case of the National HIV AIDS programme, it led to considerable delays and severely disrupted the execution of the programme. There were also problems with the implementation of the Parliamentary Support Programme.

Technical assistance

80. EU funded technical assistance is both well perceived and remains an important component of the overall EU aid programme to South Africa. Nevertheless, it is a high cost resource, particularly where expatriate technical assistants are recruited from Europe. At present the Government Departments receiving such assistance remain largely unaware of its cost because it is not financed directly from project funds.

81. In South Africa there is a significant amount of skilled expertise available in many fields and there is considerable scope to recruit technical assistance locally, applying local terms and conditions which are more cost-effective than those for European technical assistance. In general terms, the Commission should first determine whether, for particular assignments, local expertise is available and appropriate for the task and, only where it is not, turn to the recruitment of expatriate technical assistance.

82. The precise duties and responsibilities of EU financed technical assistants seconded to work in various Government Departments have not been sufficiently defined by the Commission, especially as regards their responsibilities towards the Delegation and the Government Departments.

83. In the case of a Government Health Project the intended role and function of a technical assistant to develop and implement a management accounting system, was diverted for 14 months to that of coordinator, with the result that the original objective had not been achieved at the time of the audit. Also, for a period of nearly 3 months, this technical assistant acted in a senior line management role within the Health Department, without the agreement of the Commission.

84. Technical assistance has not always been provided at the right time. Normally, the contract time period of technical assistance should be in line with the project timetable. In the case of the Land Reform Project, for example, the administrator provided for under a technical support programme was only appointed 12 months after the signing of the Financing Agreement. The South African authorities had to start the project without his assistance.

85. More attention needs to be paid to ensuring that sufficient technical assistance is provided at a local level, and not only at central level. For example, regional beneficiaries involved in the Land Reform Project would have obtained easier direct access to advice, instead of having to route their queries through elaborate committee processes.

COMMISSION ORGANISATION AND RESOURCES

86. This report has emphasised in several instances the unusual and particular nature of the South African aid scheme; it combines high political relevance, administration complexity, volume of funds and diversity of actions.

87. All these factors called for high quality organisation and reinforced management resources, which were not, however, provided. Firstly, between 1986 and 1991, the Commission could not set up a structure on the spot for the administration of the scheme. In 1991 the Programme Coordination Office was opened, but only with limited technical responsibility. It was not until 1994 that the Commission established in Pretoria a full Delegation.

88. Secondly, the Commission failed to give enough priority to the management of the South African aid programme: insufficient management resources were allocated, there were inadequate written procedural instructions and the organisation set up was inadequate. At Commission headquarters in Brussels, the South African Unit has been insufficiently staffed, with frequent staff movements and gaps in posts.

89. In South Africa, the tasks and responsibilities of the PCO, and the procedures that it was to follow, were not clearly set out. This contributed to management shortcomings and deficient monitoring.

90. Even when the Delegation was established in Pretoria in early 1994, the Commission did not provide sufficient staff for the duties involved. For more than one year the Delegation consisted of the Head, his assistant and a secretary; the Development Counsellor was not appointed until 1995 and the posts in the staff organigramme were not all filled until the end of 1996.

91. In addition, the lack of written instructions and guidelines has led to shortcomings in the management routines of the Delegation; inter alia:

(a) there is no clear division of responsibilities between the staff of the Delegation, so that, for example, it is not clear who is responsible for monitoring which project;

(b) information systems are weak: project documentation is not put together in a single file; in several instances information has not been processed and various reports and other documentation have been lost;

(c) the monitoring system does not provide sufficient information for assessing the current status of the projects: actual disbursements; reporting deadlines; progress made towards objectives, constraints encountered, etc; under these circumstances it is not possible to ensure adequate project monitoring, supervision and continuity in the service.

92. A particular point that affects the organisation of the Delegation in South Africa is the need to maintain a presence in both Pretoria and Cape-Town, where the National Parliament alternates meetings. The Delegation has opened an office in Cape-Town where the Head is frequently obliged to remain in order to deal with political affairs. That has led to delays in the resolution of important matters, especially affecting the development cooperation. There is a need to nominate a Deputy Head of Delegation in Pretoria responsible, in particular, for development cooperation and project monitoring.

CO-ORDINATION IN DEVELOPMENT CO-OPERATION BETWEEN THE COMMUNITY AND MEMBER STATES

93. Article 130u of the Treaty on European Community (TEC) stipulates that Community development co-operation policy shall be complementary to the policies pursued by the Member States. With this aim, and seeking for coherence and effectiveness, Article 130.x of the same treaty, requests the co-ordination and consultation between Community and Member States. Since 1993, the Council and the Commission have taken significant steps in this direction: a pilot experience was launched in six countries (25) and a series of resolutions and conclusions were issued (26).

94. In South Africa, meaningful coordination should be regarded as a major issue. Firstly, 14 of the 15 Member States are represented in the country and face similar problems; secondly, strong links and co-ordination, which were developed during the earlier more difficult political periods, are already in place.

95. Council Regulation (EC) No 2259/96 of 22 November 1996 on development cooperation with South Africa (27) regards coordination as a central instrument to guarantee optimal effectiveness of the aid. The Regulation invites the Commission to take all necessary measures; these would involve the organization of regular meetings, the establishment of a system for the exchange of information and the transmission to the Member States of a cooperation strategy paper and project outlines. Finally, the Regulation extends the need for coordination to other donors.

96. Current coordination mechanisms in South Africa consist of regular meetings and of exchange of information on the programmes implemented. These mechanisms have permitted a significant involvement of the Member States in the definition and inception of the Community programmes. Documentation has been regularly sent by the Commission to Member State representatives and Community operations have been discussed in detail. However, the exchange of information regarding Member States' actions has not been formalised and little progress has been made on the establishment of a common data base. Insufficient information has hindered the achievement of a more effective coordination.

97. To strengthen coordination, the Commission and Member States should take practical steps to implement the provisions laid down in the above-mentioned Regulation:

(a) setting out detailed procedures for consultation and exchange of information and for the adoption of common approaches; these procedures should ensure that information flows in both directions between the Community and the Member States, including meeting schedules, compilation of a data base, transmission of reports, and common appraisals;

(b) defining and enhancing the role to be played by the Commission Delegation; adequate resources must be provided so that the Delegation can fulfil this role;

(c) developing common sectoral strategies; efforts should be directed to improving the administrative capacity of the South African government to deal with and coordinate international donors;

(d) extending coordination to other donors and adopting a consistent approach, which should entail common accounting and reporting requirements, agreed disbursement schedules and joint auditing procedures.

CONCLUSION

98. In September 1986 the European Community decided to support actively organisations that were providing assistance to the victims of apartheid, and to contribute significantly to the efforts to secure a peaceful transition to democracy. As such, it was a unique programme, opposed by the South African Government, and controversial in Europe. The Commission showed considerable imagination and foresight in promoting the programme. It was responsive to the needs of the target population. The Special Programme is regarded by those in South Africa who were involved with it as having made a positive contribution during the period of transition. As a result, the European Community built up considerable political goodwill with the new democratic leadership of South Africa (paragraphs 1 - 5 and 22).

99. The Community, therefore, was in a key position in 1994 to develop relations between South Africa and Europe in the economic and cooperation spheres. It took immediate steps to do so. Like other donors, however, the Commission, in order to show without delay its solidarity with and support for the new government, approved through the EPRD more programmes spread over a wider number of sectors than could be effectively managed in the short term. The constraints on the capacity of the Commission to handle such a portfolio of actions and of the new government to absorb the funds within the timescales foreseen were not sufficiently taken into account. Nor was sufficient preliminary ground-work done to identify what institutional strengthening was necessary before projects could be implemented effectively, and what role the Community should play in assisting the South African authorities in this regard (paragraphs 56 - 57). As a result, substantial implementation delays were encountered in the implementation of the EPRD. Similar problems have already been identified by the Court in other areas of development cooperation, as PHARE (28) and the Mediterranean (29).

100. Efficient and effective implementation of the Community's programmes of assistance to South Africa has throughout the whole period from 1986 to 1997 been adversely affected by the inadequate procedures, systems and resources applied by the Commission:

(a) initially, the Commission sought to apply 'light and flexible` procedures, but without defining what these consisted of and setting out clearly to the partners what would be acceptable (paragraph 4);

(b) as the situation improved and it was possible to operate in a more open way, the opportunity to establish appropriate procedures and systems was not taken. Although the administration of the projects was entrusted to the Channels, the Commission, in an attempt to ensure strict control, retained in the central services the authority for the approval of each individual project, contract and contract amendment, report, and payment. Since the Commission failed to provide sufficient resources to carry out and check properly and promptly all these operations, the approach followed was unworkable and failed to provide adequate management and control;

(c) the specific exercise to close the 750 open projects with NGOs and CBOs, going back to 1986, was poorly designed and implemented, so that the limited resources available were used inefficiently (paragraphs 45 - 52);

(d) as far as the EPRD is concerned, the chronic problem of insufficient resources to manage a programme of its size and diversity has been made worse by the lack of operating instructions for staff in the delegation and in Brussels, and staff movements (paragraphs 87 - 92);

(e) implementation efficiency has been adversely affected by the slowness of Commission payment procedures, shortcomings in financing agreements and the failure to agree on the tendering procedures to be applied.

101. Many of the shortcomings highlighted in this report are similar to those set out in Court reports concerning other external actions, for example the Special Reports on the PHARE programme (30), on the MED-Programmes (31), and observations concerning NGO cofinancing (32). The Commission should, without delay, take steps to overcome these weaknesses.

102. The Commission should seek, therefore, in the area of development cooperation generally, to establish a more effective decentralised management approach that not only places decision making authority at the most appropriate level of the system, but also strengthens ownership on the part of the beneficiary organisations and encourages a greater degree of partnership. Within its own organisation, a degree of operational and financial responsibility should be decentralised away from central services to the field, where decisions can be taken closer to the actions concerned. Such an approach should also aim to utilise the limited resources of the Commission for managing and controlling the use of Community funds in an optimal manner. Such an approach would require:

(a) establishing a clear framework and terms of reference for the operation and financial implementation of the programmes, including precise and straightforward reporting requirements;

(b) determining what decisions are best taken at the level of implementing organisations, the Delegation, and the central services in Brussels, and what are the pre-conditions in terms of capacity and resources for such a decision making structure;

(c) as one of the above pre-conditions, being able to rely on the management and control systems of the implementing organisations: to achieve this, the Commission needs to assess existing systems and provide support for their improvement;

(d) focusing monitoring on the overall execution and results of the programmes, complemented by selected checks of individual actions.

103. The EPRD forms one part only of the relations between the EU and South Africa. The negotiation of a Trade and Cooperation Agreement which will establish, after a transitional period, a free trade area (33) with the Community is a matter of great importance for South Africa. The volume of development aid from the Commission to South Africa is considerable - South Africa benefits from one of the largest single country financial and technical cooperation programmes with the Community. However, the economic and social progress of the country depends to a greater extent on the successful development of efficient producers able to compete in international markets in the many sectors (natural resources, agriculture) in which it has a comparative advantage. The trade negotiations are long and complex, and were not completed at the end of 1997. In this context, it is important to ensure that European commercial policy is coherent with development strategy.

104. Member States have very keen interest in and have been closely involved in the trade negotiations. It is disappointing to note, however, that, in the context of a new financial and technical cooperation programme with a new government, coordination between the Commission and the Member States is limited. They have had a significant input into the definition of the Community programmes, but there has been a lack of information provided by Member States on their actions. This limits the possibilities to define a coherent European Union programme of assistance. Further, greater efforts are needed, not only between the European Union donors but also with other donors, to agree common approaches on procedural and accountability requirements.

This report was adopted by the Court of Auditors in Luxembourg at its meeting of 27 and 28 May 1998.

For the Court of Auditors

Bernhard FRIEDMANN

President

(1) Ministerial level political cooperation meeting on 10 September 1985.

(2) EC development aid scheme regarding South Africa has been funded between 1986 and 1990 from budget line B9530; between 1991 and 1995 from B7-5070 and, since 1996, from B7-320.

(3) The term 'Channels` is used to describe these associations because the Commission channelled through them the funds of the Programme to the implementing organisations.

(4) The Agreement was signed in Pretoria in October 1994 and approved by the Council on 19 December 1994 (OJ L 341, 30.12.1994).

(5) The TNDT is the first step towards the creation of a National Development Agency which will ensure NGOs funding in the future.

(6) Council Regulation (EC) 2259/96 of 22 November 1996 on development cooperation with South Africa (OJ L 306, 28.11.1996).

(7) The EC aid programme to South Africa will, however, continue to be financed from the EC General Budget as South Africa's accession to the Convention is limited.

(8) See the Evaluation of the European Special Programme on South Africa. SPM Consultants; October 1996.

(9) Of the original amounts committed, 10,2 Mio ECU were decommitted.

(10) National Indicative Programme signed by the Commission and the South African government on 14 May 1997.

(11) 'Evaluation of the European Special Programme on South Africa`, SPM Consultants, October 1996.

(12) See paragraph 10 in the Annex on the Bursary Programme.

(13) (Observations concerning the Bursary Programme are included in the Annex).

(14) Between 1993-95 the Commission approved 189 projects with the Channels, most of them for amounts of less than 0,2 Mio ECU. To this number should be added the running projects approved in previous years. Around 700 projects remained open during the considered period.

(15) The logical framework is a management tool aimed at improving project designing, planning and monitoring. It is used by major development aid donors.

(16) The South Africa Project Management (SAPMAN), a spreadsheet application used by the Commission, did not contain reliable information, since it was not regularly up-dated. The application was abandoned in December 1995.

(17) The Council Regulation (EC) No 2259/96 of 22 November 1996 provides in Article 10 for a financial reference amount of 500 Mio ECU for the period 1.1.1996-31.12.1999.

(18) The task force responsible for the negotiations was not set up until March 1995.

(19) There being no real demand for the ECU in South Africa, the banks are able to set their own exchange rate, which may differ considerably from the rates quoted on the international money markets.

(20) Payment by the Commission in ZAR would make it possible to minimise these losses and the possibility is not excluded by the Financial Regulation.

(21) For example, the Parliamentary Support Programme (AFS/1996/9) stipulates in the body of the financing agreement that the Ministry of Finance represents the Republic of South Africa, even though the agreement was signed by the Speaker of the Parliament.

(22) Projects Urban Sector Network (94/75070/060) and Micro-project Programme Trust (95/75070/005).

(23) See, for example, the projects Participatory Primary Health Care Programme, the Urban Sector Network, the Microprojects Trust and the Land and Agricultural Policy Center.

(24) See paragraphs 15.27 - 15.31 of the Court of Auditors' Annual report concerning the financial year 1991 (OJ C 330, 15.12.1992) and Special Report No 1/98 in respect of bilateral financial and technical cooperation with non-Member Mediterranean Countries (OJ C 98, 31.3.1998).

(25) Bangladesh, Ivory Coast, Costa Rica, Ethiopia, Mozambique and Peru.

(26) See notably the ones issued on 2.12.1993; 6.5.1994 and 15.7.1996.

(27) OJ L 306, 28.11.1996.

(28) See paragraphs 2.4 - 2.12 of the Special Report No 3/97 concerning the decentralised system for the implementation of the PHARE programme (period 1990-1995) OJ C175, 9.6.1997.

(29) See paragraphs 150-151 of the Special Report No 1/98 in respect of bilateral financial and technical cooperation with non Member Mediterranean Countries.

(30) Special Report No 3/97 concerning the decentralised system for the implementation of the PHARE programme (period 1990-95) (OJ C 175, 9.6.1997).

(31) Special Report 1/96 (OJ C 240, 19.8.1996).

(32) Annual Report concerning the financial year 1990 (OJ C 324, 13.12.1991).

(33) Subject to a number of protocols protecting certain key sectors from tariff-free imports of products from the Community.

ANNEX

BURSARY PROGRAMME

Programme objectives and description

1. The objective of the Bursary Programme was to assist higher education students from disadvantaged communities and to contribute to the establishment of a national bursary scheme. It provided bursaries consisting of a grant (60 %) and of a loan (40 %) to be repaid by the students to the Tertiary Education Fund of South Africa (TEFSA). Other components of the Programme targeted the improvement of administration and student support services of historical black Tertiary Institutions (TI) (1). Between 1993-1995 the EC allocated to the Programme 59,9 Mio ECU, of which 45,5 Mio ECU has been disbursed. Around 34 000 bursaries were granted.

2. The Programme was managed by the Kagiso Trust (KT), which established a Bursary Office for its administration. The bursaries were disbursed through the TI to which the Office advanced funds according to the amounts awarded to each student. At the end of the exercise, the TI forwarded to the Office an audit report stating the amounts of the grants and the loans provided by beneficiary, the actual figure expended and the remaining amount to be refunded.

3. A Bursary Programme Committee was set up to establish policy criteria and to oversee the administration of the programme. The Committee consisted of representatives of the TI, the Ministry of Education, TEFSA, KT and the EC Delegation.

Control system

4. The financial control of the programme chiefly relied on the submission by the TI of audit reports. However, this mechanism did not function satisfactorily:

(a) deadlines for the submission of the audit reports were not established and a penalty clause in case of non-submission was not foreseen;

(b) until 1995 the Bursary Office did not request from the TI the reports due for 1992, 1993 and 1994;

(c) comprehensive instructions were not given to the auditors; the reports were not consistent and in most cases did not meet the control objectives. In several cases, the auditors themselves did not consider the exercise as an audit in the strict sense, in some cases the scope of the work was restricted, in some only qualified opinions were formulated or none at all;

(d) there was a lack of follow-up of the reports by the Bursary Office. Unacceptable reports were not rejected (for example those issued by the TI's own internal auditors) and those containing a qualified opinion or containing observations were not cleared up.

5. No other control mechanisms were operated. The Bursary Office did not carry out annual reconciliations between Commission, KT and TI accounts to establish the balance of funds available. As a result, around 11 Mio ECU (18 % of the funds available) remained unused (2).

6. One particular weak element in the control system was the absence of an audit review of the Bursary Programme as a whole, which was not provided for in the contract concluded between the Commission and KT.

Programme supervision

7. The poor functioning of the control system was made worse by the lack of effective supervision. Firstly, the Bursary Office did not review the procedures and systems of each TI.

8. Secondly, the Bursary Committee did not function properly and did not follow up matters of a financial nature. The Committee held fewer meetings than foreseen, there was inadequate segregation of duties (the secretary of the Committee, was, at the same time, the director and financial manager of the Bursary Office) and data examined mixed up projections and actual figures.

9. Thirdly, neither KT main office nor the Commission supervised the work performed by the Bursary Office. Until the end of 1996 they were not aware of the absence of reliable figures on expenditure and of the non-repayment of the loans by the students.

Evaluation

10. In 1994 a mid-term external evaluation of the Bursary Programme was conducted (3). The evaluation did not assess financial management and did not tackle the loan repayment problem. The recommendation that the financial overview role of the Bursary Committee should be strengthened was not followed up.

11. No final external evaluation of the Programme has been carried out nor is foreseen.

Achievement of objectives

12. The 1994 evaluation underlined the success in reaching the targeted group and concluded that the Programme provided financial assistance to disadvantaged students and was contributing to the redress of racial and gender imbalances. This conclusion was drawn in 1994 and needs to be further confirmed for the period 1995-96.

13. In a number of respects, however, effectiveness, sustainability and visibility of the Programme have not been satisfactorily achieved.

Effectiveness

14. Effectiveness was reduced as a consequence of poor financial management. As explained in paragraph 5, 11 Mio ECU remained unspent; this could have financed more than 8 700 additional bursaries.

15. Further, funds arrived very late in the school year to final beneficiaries. As a result, the TI and the students were obliged to advance the expenditure, sometimes taking out private loans.

Sustainability

16. Sustainability of the Programme has not been ensured in that no priority was given to the repayment of the loan component of the bursaries.

17. At the end of 1996, only 1,13 % of loans had been repaid; this represents a loss of more than 72 Mio ZAR (approximately 12,5 Mio ECU at December 1996 exchange rates) for the national bursary scheme. The feasibility of recovering the loans is dramatically decreasing with time; this undermines the sustainability of the Programme and puts at risk the financial capacity of the national scheme.

Programme ownership

18. The beneficiary students did not receive an annual statement of account (loan granted, interest accrued, etc.) to record and confirm their financial position. They were poorly informed on the mechanisms of the scheme and were not involved at any stage in its administration. As far as the final beneficiaries are concerned, there was a lack of programme ownership, which is a chief element for ensuring successful development programmes.

Visibility

19. In most of the TIs visited by the Court, neither the TI's staff nor the beneficiary students were aware that the European Union was the source of the funds. KT did not make public the source of funding until the end of 1996, when problems arose because the Programme came to an end. The Commission's Financial Controller pointed out the lack of visibility in 1995, but no steps were taken to redress it.

20. The Commission did not develop an information strategy (via brochures, press releases or other actions) for the dissemination of the EC origin of the funds. Visibility is necessary to reinforce partnership and the cooperation relationships both in South Africa and in Europe.

>TABLE>

(1) Tertiary Institutions are universities; technikons; and teacher training, technical and vocational colleges.

(2) This amount corresponds to 10,67 Mio ECU of Community commitments not disbursed to KT and approximately 0,5 Mio ECU in funds unexpended in the TI's accounts.

(3) The European Union/Kagiso Trust Bursary Programme, an evaluation. British Council. August 1994.

REPLIES OF THE COMMISSION

INTRODUCTION

The Court's audit policy towards EC's South Africa Aid Programme

14. The Commission takes note of the global positive assessment of the Court on its programmes in South Africa in spite of the difficult conditions prevailing there during the first two phases, and agrees that the basic principles underlined by the Court are those which must - and in fact do - form the guidelines for its activities. In order to ensure financial control and accountability a certain amount of procedural groundwork is required which can create delays in implementation, given that our development partners are not familiar with donor's requirements, or sometimes have a different administrative approach.

IMPLEMENTATION OF THE BUDGET

15 - 19. As regards the disbursement rate of payment appropriations for the projects relating to the period 1986-1990 - a percentage of 85 % was achieved. During the project closure exercise, unspent amounts have been identified (up to 4.2 MECU) and are being decommitted.

Concerning the commitments for projects launched during 1991-1994 a disbursement rate of 70 % was achieved. During this period large exchange gains ECU/Rand occurred. This prevented liquidation of total contract amounts in ECU. Due to lack of personnel it was not always possible to close the projects and to decommit rapidly.

As the Court rightly points out, and with the introduction of the new European Programme for Reconstruction and Development (EPRD) in 1995, it is true that payment levels declined sharply. Between 1994 and 1996 the major unspent commitments were in government projects for the years 1995-1996, when the disbursement percentage commitments/payments reached only 29 %. This is to a large extent due to the initial expectation by South Africa that the EC funds could be used to reimburse expenditure already incurred by the implementing Departments or public bodies. The South Africa administration realised only in a later stage that their legislation did not allow a pre-financing of Community projects by national funds, it subsequently became apparent that the Departments were unable to make initial payments to the projects. This held up the execution of these projects for 12 to 18 months at least. This problem was addressed in 1996, when it was decided that the corresponding Financing Agreements should be amended by riders to allow the payment of advances on the basis of approved annual work plans. It took five months for the South African Government to sign the riders. In 1997, after adapting the Financial Agreements it was possible to double the amount of payments compared with the previous year.

Another reason for the low execution rate was the change in government administration following democratic elections bringing about major reorganisation of the staff responsible for our programmes.

These delays could not be anticipated by the Commission since their origin was within the competence of the South Africa government.

There were also slow start-up times for large programmes, which could not be fully anticipated at that time for which partial solutions have now been found (e. g. preparing tender dossiers and/or work plans before the financing agreements are signed in order to be ready to launch the programme immediately).

SYSTEM FOR IMPLEMENTING PROJECTS THROUGH THE CHANNELS

Introduction

22. The changes in the way the Commission was operating in South Africa from 1991 onwards led to difficulties which were and are addressed. Shortage of staff was often an obstacle to quick reactions. It has always been a problem for the South Africa programme. Staffing levels are gradually being increased, but this takes time. The streamlining of administrative processes (see responses to points 39 and 66) and aid delivery mechanisms is also expected to have a positive effect on the reported weaknesses. It is also considered that a more focused programme with clear priorities in our co-operation with the South African Government would lead to a higher level of efficiency in development achievements.

Observations concerning the Channels system

Design weaknesses

25 - 26. After the period of 'light and flexible` management due to the particular circumstances in South Africa - 1986-1991 - the Commission had to tighten the management procedures for the next round of development projects in South Africa. Unfortunately, as mentioned above, the necessary human resources were not rapidly available, causing delays in the day to day execution of the programme.

27. The special difficulties of South Africa, and especially the problems resulting from the Apartheid era, encountered by the Commission and the learning process in which it was engaged were quite different from other programmes. In addition, it should be mentioned that the Channels themselves are under constant change. The role NGOs play in the development sphere is still a sensitive matter in the South African political context. The Commission has consistently sought to adapt itself to this dynamic situation, the approach being to act as much as possible through well equipped and well organised intermediaries (such as Transitional National Development Trust (TNDT) or INTERFUND) which were identified via an appraisal study launched in 1997, and which will have to be reinforced (capacity-building) when necessary (see also point 66).

Formulation of projects

28. There was no delegation in South Africa at the time. The Programme Co-ordination Office (PCO) did not have enough staff with experience in project preparation. In 1993/94, efforts in South Africa were concentrated on the forthcoming elections, and it was difficult to train partners at that time in project management. Project proposals were thus by necessity often accepted in the format in which they had been prepared, and were indeed sometimes poorly formulated.

29 - 30. Project time frames were invariably too short, failing to allow sufficient start-up time and even sufficient time for implementation and reporting. Additional information was often requested before agreeing to extend time frames, but communication between the Commission headquarters and individual projects, often in remote areas, was not easy, and replies or justifications, not always understood by the project managers, often took months to be submitted. Extension of contract deadlines, which places a considerable administrative workload on all concerned, is a requirement of budgetary discipline. The particular exercise referred to in this paragraph was due to the lack of human resources, which made it impossible to observe all the deadlines during the budget year. And it must be remembered that time frames were extended in order to give the required budgetary cover, i.e. to keep the commitment open, and to allow the project to report and receive funds if these were due.

Financial and treasury management

31. The rather weak follow-up of the partners' accounts was, to a large extent the result of the extremely difficult circumstances under which the programme started. As the Court rightly points out, the programme started in the face of severe active opposition from the South African Government. In order not to put its partners in danger, the Commission had to adapt management to the reality of the situation. The fact that the Commission did not immediately apply the strictest reporting and follow-up requirements as from 1991 on is due to various reasons. These range from a lack of resources to the positive experience of partners' financial management during the previous years. This was recognised by the evaluation team in their report on the European Special Programme (ESP) (Oct. 96).

The Commission recognises that the monitoring of the funds could be improved, certainly in relation to the Southern African Catholic Bishops' Conference (SACBC) and South African Council of Churches (SACC). The closure exercise in South Africa by the Technical assistant (period: 1986 to 1991) has already helped the Commission to identify unused funds which are being recovered. The ongoing closure review by the audit firms (period: 1991-1994) will allow the Commission a detailed insight of the actual situation on outstanding funds. To facilitate the recovery of these funds, the Commission is considering opening an account in South Africa to receive them.

32. As the Court rightly observes, the strict observance of the conversion system was impractical. The Court's report relates to not amending contracts. It is true that the Commission did not modify the contracts to fit the practice at that time: this was not an ideal situation, but given the priorities at that time and the pressures of managing so many contracts with so few resources, this issue was not addressed.

Payment and reporting

33-38. As far as reporting is concerned; a timeframe for the project was always indicated following the budget table. Normally the reporting should have been done during this timeframe (unless an extension had been granted). However, the signature of the contract often occurred much later than the indicated starting date of the project. As the payment arrived only after that signature the whole project, and consequently the reporting, was delayed. Most projects only started after the first payment had been received.

The reporting required by the Commission was not too difficult for the beneficiaries. It required only a narrative report (2 or 3 pages were acceptable) and a financial report indicating - according to the budget in the contract - what had been spent of the advances paid by the Commission. Especially for the earlier projects the Channels due to their requirements often rejected the reports. As the projects were often located in remote areas and the reports transferred via local offices to the headquarters of the Channels, considerable time was lost in this process.

The fact that decision-making was rather slow in the Commission obliged the Channels to introduce requests for continuation of a project well in time. This could give the impression that the funding for the initial project and the second or extension project were executed together. Wherever possible the Commission took into account a certain closing date for the first project - based on the audit reports of the project -and only from that date onwards the next project could start. With the large number of ongoing projects it was not always possible to track the different contracts for the same project in due time.

34. During the first two phases the projects were for the most part quite small. The system of linking reporting to payments was appropriate for that type of project, especially given the shortage of Commission staff dedicated to South Africa. Of course, some projects did not submit their reports, but they would have been even less likely to do so without any financial incentive.

35. As reporting was linked to payments and to the time frame of the project, as mentioned above, this did not always correspond with the financial year. The projects had not asked for this linkage. On a few occasions, new funding agreements were signed before ongoing ones with the same organisation, had been closed, but care was always taken that there was no overlap of funding.

36. The Commission was always as flexible as possible, within the constraints of the Financial Regulation, in the knowledge that many of the projects had limited capacity for reporting. Thus, the minimum required was that a financial statement be prepared showing expenditure against the budget lines in the contract, and that a short narrative report should give information on the activities carried out.

37. The Court observes that the last payment is often outstanding. The current closure exercise will bring some clarifications on this. Preliminary results indicate that any overfunding was at the level of the Channels, which could often be explained by exchange rate differences at that time. The overfunding - or exchange gain - stayed then with the Channel, until recovered by the Commission.

However with the arrival of the democratic government in 1994 there was a change in the approach to NGO's and also in the funding habits of donors who turned more towards government led projects. The NGO's received less funding and often the projects closed before all objectives were achieved thus not allowing them to present the final report, a condition for the disbursement of the last instalment. Hence the abandonment of the last instalment.

The reporting on grants provided by other donors was not explicitly indicated in the earlier projects and the Commission services often had the impression that the projects just tried to keep things going with the grants that came in.

The Court rightly points out that the closure exercise should give some insight as to why a certain number of projects have not submitted a final report, or even in some cases interim reports. In the Commission's opinion its reporting requirements can not considered to be particularly cumbersome, even though the approval circuit may seem for the projects to be rather lengthy.

38. Audited financial statements, while sometimes of limited use regarding the precise use made of EU funds, nevertheless give a good indication of the general financial situation of the organisation and reveal incidences of misuse of funds. In the particular case of the Bursary Programme, because of the unsatisfactory situation, the Commission took action in 1996 and it was decided to stop funding the bursary programme through Kagiso Trust (KT) and to fund ongoing students through TEFSA (the Tertiary Education Fund of South Africa (see Annex).

Project monitoring

39. The Commission is currently improving procedures and has constituted a task force within the Delegation in order research solutions to improving the information system.

40. Thirteen of the fourteen projects referred to have all expired and are marked for closure, except for one project of SACC. Project time frames have been extended in order to retain the possibility of further action if required.

42. Delays were incurred, as stated by the Court; however, they were not always caused in the way described. It should in this context be noted that neither with the channels, nor with the Programme Co-ordination Office (P.C.O) was any expertise available in financial matters relating to the Commission.

As far as the Bursary Programme is concerned, the situation raised by the Court has now changed, since the visit of Commission representatives to the project in October 1996, when remedies were proposed and accepted. In 1997 TEFSA has taken over the management of the projects, remedied the situation and is beginning to collect the back payments - see comments to point 38.

43. The initial management information system was abandoned at the end of 1995, but a new tool was introduced at the beginning of 1996, namely GELIBU (Gestion des lignes budgétaires) for South Africa. This supplement to the information system used by the Accounting Service SINCOM allows for tracking of the project status, as well as providing all financial information. It is correct that the Commission did not have a copy of the contracts between the Channels and the project holders, preferring for practical reasons at that time to limit contractual relations to the Channels. This system was modified in 1995.

44. The second wave of the project closure exercise is now well under way and will address the problem of outstanding balances. It should be noted that 4 MECU of the 10 MECU outstanding can be attributed to one project, Uthombo Pride (92-75070-397), which it has not been possible to decommit because of on-going protracted legal proceedings.

Closure of Contracts

47. Meetings were held in April 1996 in Pretoria and July 1996 in Brussels where planning and responsibilities were defined. A definitive format of the required information to close the projects was decided in 1997.

48. In addition, the recruitment of the private sector audit firms necessitated a restricted tender launched in July 1996 with a deadline for submission of offers of 25 September 1996. Because the offers did not meet the requirements, the tender was cancelled and the two contracts were awarded by direct agreement.

45-52. The procedural difficulties the Commission encountered during 1996 have been solved during the second half of 1997 and the situation on the closure of ESP projects as outlined in the Court's report has definitively improved. Inter service agreements have led to a clear definition of responsibilities and tasks. Computerised tools to overcome the shortcomings of existing systems have been developed. This has allowed the finalisation of closure dossiers of all projects of the 1986-1991 period, 415 in total. Of these, 125 projects have already been closed, the remainder being in the final stages of closure.

The second wave of closures (period: 1991-1994) is ongoing with the assistance of two audit firms in South Africa. Although there were delays in producing the closure review reports by these firms, the Commission still intends to respect its deadlines to close 180 projects by the end of June and to close all non-active projects by the end of 1998. In order to overcome the restricted logistical means of the Delegation, measures are being applied to reinforce its equipment and resources. To date, 32 projects have been closed, and 23 are in the final stages. These projects have followed a 'natural` life cycle and were proposed for closure after receipt of closure reports, mainly from KT. These projects will still be reviewed by the audit firms but will not need further action assuming the auditor's findings correspond with the closure report of the partner.

THE EUROPEAN PROGRAMME FOR RECONSTRUCTION AND DEVELOPMENT (EPRD) 1994-1996

The EPRD system

57. As mentioned in points 18 and 27, it was politically and technically difficult to prejudge the implementing capacities of the new Mandela administration. The Commission also had to operate within the framework of the 1994 co-operation agreement and the 1996 Council Regulation. The latter determines the level of funds to be allocated to South Africa and specified large sectors of intervention. This opened the door to many requests from various South Africa Departments. This issue has been examined with the SA government and as a result a Multiannual Indicative Programme (MIP) was agreed upon with a concentration on basic social services (Education and Health), laying the foundation for direct budget-support mechanisms. Unfortunately most of the departments involved in the sectors prioritised by the MIP are facing major difficulties in the implementation of the programme. Therefore, the Commission is now seeking, in agreement with the NAO, to redirect its activities towards large financial and managerial capacity-building programmes in key departments and towards sectors where the governmental partner has well recognised and established implementing capacities, be it at national or provincial level. In parallel, operations with non-governmental partners will be reinforced.

60. b) In the case of the health programmes, the combination of several NGOs was unsuccessful and therefore that method of grouping was not repeated.

Experiences gained on earlier projects in the context of organisational structures for co-ordinating activities have been applied to new projects. Much attention is now paid to the capacity of such organisations. New programmes are always based upon either an appraisal study or, in the case of an organisation with which the Commission has already worked, an evaluation of the previous project.

61. The TNDT project started slowly, partly because one major partner left, but is now a model for successful implementation and monitoring. The Commission is satisfied with the work carried out by this organisation.

Government projects

62. The three health programmes also started slowly. These were in fact the first programmes funded through the government. There was a natural learning curve. Some problems occurred with tender procedures, but much effort was put into making these programmes work. At the present moment, the 3 projects have reached a level of execution of over 90 % and are in a final stage of implementation.

63. The project started very slowly. Now it is well under way and a mid-term review is currently being undertaken.

64. The aims were kept vague because the actual situation under the new government was in itself not clear (points 15 - 19, 27 and 57). The project therefore needed to be designed for initial flexibility. A mid-term review is currently being prepared for the Human Rights Programme.

65. The initial cost of the programme was estimated at 36 MECU. The Commission has accepted to contribute up to 16 MECU, the Swedish Co-operation up to 0.5 MECU on women and development, and the remaining was to be provided by the South African Parliament budget. The South African Parliament has decided that their contribution would be done in a flexible and gradual way. This explains why the first work plan was based on the Commission's contribution only. The successful implementation of the part financed by the Swedish government should also be noted, and co-ordination meetings with the Commission are being held. The staff have been appointed and the project is being executed. External expertise was recruited recently to help the project management to establish in detail the objectives of the project and how to achieve them in the most cost-effective way. The intention of the joint funding will appear in the next work plan.

NGO projects

66. The Court rightly mentions the difficulties encountered by NGOs in providing the Commission services with information needed to justify the release of funds and the delays, which resulted from this situation. It was recognised that very often the NGO's did not realise that advice was available from the Delegation. The Commission has taken a series of measures to improve this situation:

- 'Guidelines for the Implementation of EU funded Projects in South Africa` have been drawn up and distributed in early 1998 to all partners (Government services, NGOs, etc . . .). These provide clear and operational guidance on inter alia how to elaborate a financing proposal, a work plan, a tender dossier, on payments and fund flows, reporting, etc. . . Seminars are being organised to promote the use of the Manual;

- A capacity-building programme targeted at NGOs operating in the framework of the MIP, to be implemented by a specialised NGO, is under approval. It will provide immediate short-term technical assistance and support to these NGOs when needed; the setting-up of a specialised NGO Programme Management Unit is also under appraisal (project INTERFUND);

- The Commission is currently examining means to bring about greater decentralisation of decision making to the Delegation.

This set of measures should significantly improve the implementation of programmes particularly with non State partners.

68. The housing loan initiative was undertaken on a province-by province basis. Secondly, the provision of housing was the responsibility of Government, which, in the aftermath of the election, was overly optimistic as regards the speed of implementation. Nevertheless, it has to be stressed that, although in the first year only 879 houses could be delivered, there were over 1 800 in year 2, notwithstanding the absence of bridging funds.

Despite difficulties faced by certain NGOs, the overall programme remained on schedule. Another component of this programme was the servicing of sites (one of Urban Sector Network's principal responsibilities) with the objective of a total of 7185 sites in the first year. The project actually achieved a total of 7529.

69. Mention is made of the excessive time required for the signature of the financing agreement and disbursement of the first instalment to the Micro-projects Programme Trust (MPT). This type of problem will be minimised in future, as we are now encouraging the projects where possible to prepare their tenders and/or work plans prior to signature of the financing agreement, so that everything is ready to start almost immediately after signature.

Part of the problem is also to be related to MPT, which did not provide the Commission's services with the required information. This was partly due to inadequate communication between the Commission and the project, which since then has been substantially improved.

Weaknesses in the tools and procedures for implementation of the EPRD

70. As was explained in point 66, the Commission has now a manual providing guidelines on every stage of EPRD implementation.

Payment delays

71. In the two cases quoted, the Urban Sector Network and the Micro-projects Programme, there were particular problems with the work plans presented, which had to be referred back to the projects. This added considerably to the approval time. As regard the internal Commission circuits concrete steps are currently in hand to decentralise and simplify procedures in order to reduce delays consistent with the principles of budgetary discipline.

Exchange losses

72. Exchange rate fluctuations were a problem. In the earlier years appreciation of the Rand meant that the South Africans had substantial exchange gains. That is why at a certain moment the Commission decided to calculate the different payments in South African Rand (ZAR) and to convert into ECU at the rate of the month in which the payment was executed. With the depreciation or appreciation of the ZAR the beneficiaries receive more or less ZAR than intended. That is the reason why final payments are now again made in ZAR, despite the fact that the Commission would prefer to use the ECU. The problem of rates applied by private banks is a source of concern and is being discussed with the government.

Financing agreement

73. The financing agreements have now been modified to include very detailed and clear clauses relating to payments. It is true that in one or two cases (one of which was the Urban Sector Network) a pragmatic solution was put forward as a basis for payment of the 20 % balance, as the text of the agreement was rather unclear.

74. The system of bank guarantees for NGOs or the 'certificates` has been abandoned.

75. See reply to point 32.

77. For the last two years at least, riders have been required for all significant amendments to financing agreements.

78. This weakness was also identified by the Commission and a clearer picture has been presented to all parties involved in the recently released guidelines (see also point.66). The amendment of financing agreements is described in the financing chapter and indicates the respective responsibilities and the 'route` that such amendments should follow.

Tendering procedures

79. As early as in 1995, the European Commission and its South African negotiating counterpart agreed to use a slightly revised version of the standard EDF tendering rules for the first financial packages of the EPRD, known as the 'Draft General Regulations`. The major difference between the standard EDF tendering rules and the Draft General Regulations and Draft General Conditions of Contract is precisely that in the latter no price-preference is given to South African companies. Procurement for community financed development projects in South Africa has also been regulated by Council Regulation No 2259/96 of 22 November 1996, on development co-operation with South Africa. Article 7 of this Regulation requires that participation in invitations to tender and contracts shall be open on equal terms to all natural and legal persons in the Member States, South Africa and the ACP-States and in duly substantiated cases, other developing countries.

This also applies to the origin of the suppliers. Without any changes to the existing agreement, the ACP-EC Council of Ministers decided on 24 April 1997 the accession of the Republic of South Africa to the Lomé Convention. Some issues and Articles regulated in the Convention are not applicable to South Africa. Concerning tendering rules, the decision confirms that no price-preference shall be given to South African entities. In its efforts to harmonise tendering rules, the European Commission instructed its services on February 24, 1998 to apply uniform procurement thresholds for all budget line activities. This decision implies that projects governed by Financing Agreements signed after this decision, must be implemented in accordance with the revised thresholds, as agreed with Financial Control.

The issue of conflicting tender regulations arose on a case by case basis, with some Departments being less flexible than others in the choice of the procedure. The Commission initiated discussions with the Government and it was agreed (and stipulated in the Multiannual Indicative programme signed in May 1997) that EC tender regulations would apply for each new programme. For older programmes, the same line is followed in principle, but in some cases, South Africa is still facing difficulties. This will certainly be solved with the amendment of the Reconstruction and Development Programme Fund Act which governs external public donors funds. This Amendment will allow each donor to use its own procedures, and should soon enter into force.

Technical assistance

81. It may be true that skilled expertise is available in South Africa, but often the projects need high level Technical Assistants with experience in EU funded projects. These can be recruited more cost-effectively from Europe in a variety of cases. Nevertheless, the Commission always very carefully examines the possibility of recruiting local expertise prior to taking a decision.

82. At the time of the visit of the Court, the Commission had already organised an in-depth consultation with the TA's in order to develop precise instructions on responsibilities and definition of tasks.

83. The lack of experience, both from the side of the Department and of the expert, have resulted in the situation described. The situation has since been corrected.

85. The location of technical assistance (central, provincial, local) and the decision taking process are decided on the basis of appraisal studies and discussions between the Government, beneficiaries and the Commission.

COMMISSION ORGANISATION AND RESOURCES

86 - 88. See conclusion

89 - 90. The observations of the Court are fully shared by the Commission, which is currently trying to find solutions (see conclusion).

91. a) The Delegation is currently awaiting the replacement of 2 senior staff members, and a provisional distribution of tasks and responsibilities has been put in place. In the light of the experience, grade and background, of these new staff members, the new organization chart will be finalised.

b and c) A distinction should be drawn between the Special Programme (1986-1994) where close monitoring was impossible, and the European programme for Reconstruction and Development (EPRD 1995 onwards), where monitoring is ensured through close contacts between the delegation and the programmes. Certain elements of the monitoring process could be further developed. To this end, a task group has been created within the Delegation (see point 39) and will propose solutions in the near future. To the extent that funds can be mobilised to implement the proposed solutions, this should lead to a much improved situation.

92. The possibility of nominating a Deputy Head of Delegation is currently under examination in the Commission and will be taken into account in the appointments of the staff referred to in the above paragraph.

CO-ORDINATION IN DEVELOPMENT CO-OPERATION BETWEEN THE COMMUNITY AND THE MEMBER STATES

95. The Commission is currently implementing the Council Regulation in accordance with the views expressed by the Court.

96. The observations address the Member States activities. The Commission will handle this in the South Africa Committee of the Council.

97. The Commission has and is taking proactive measures to reinforce the co-ordination in Pretoria along the lines mentioned by the Court.

CONCLUSION

98-104. As the Court has acknowledged, it was for political reasons that the Commission was formally represented in South Africa through its Delegation only after 1994. It then entered into a process of continuous support and technical adaptation to the complex and changing circumstances of South Africa. This took place in a context, both for the Commission and, in some cases for the South Africa authorities, of scarce human resources and budgetary constraints. Furthermore, the conduct of the EU/SA negotiations for a long-term contractual framework from June 1995 has imposed an additional important burden on Commission services, which could not foresee at that time that those negotiations would be so complex and lengthy and which in fact are still ongoing. Management resources were and remain at a critical level and the Commission agrees with the Court that this situation had an impact on implementation efficiency. The Commission has not however remained inactive and lessons have been learnt from the difficulties experienced. On many occasions, Headquarters and the Delegation have had a common assessment of the problems they were facing. This has led the Commission to simplify, rationalise and codify its modus operandi (models exist for most standard documents, a manual of procedures has been elaborated, implementation problems are regularly reviewed and discussed with the SA Government, new instruments such as direct budget aid are being tested, etc.). At the same time measures have been and are being taken to ensure that the South Africa Unit and the Delegation are better staffed.

The Commission is currently undertaking an exercise with a view to bringing together the executive part of all the external actions in one common service in order to harmonise the execution and follow up of the external aid programmes. The Commission is also engaged in a process of decentralisation and simplification of its co-operation. The case of South Africa is being duly taken into account. At the same time, the Commission is negotiating a Bilateral Agreement on Trade, Development and Co-operation with South Africa, which will contain provisions on development co-operation. It will also prepare with the Member States a new Council Regulation for the post-1999 era. In both cases, the Commission intends to take advantage of these opportunities to improve its relationship with South Africa in terms of aid delivery, partnership, reliability and decentralisation. The Court's conclusions and recommendations form an important contribution to that process.

ANNEX

BURSARY PROGRAMME

It is true that, given insufficient human resources, the Commission relied too much on the KT Bursary Office for the management of this programme and that this management proved to be inadequate. This was realised in Brussels upon receipt of inadequate reports from KT. A mission from Brussels concentrated on these problems, and decided towards the end of 1996 to cease funding students through KT and decided to complete the funding foreseen through TEFSA. TEFSA has since verified and reconciled all KT data from 1992 to 1996. This reconciliation has enabled them to work more efficiently on the loan recovery component of the agreement, although it is too early to produce statistics at the moment.

TEFSA's loan recovery capacity improves annually and as they are confident of continued support from the government, the sustainability of the programme will be ensured through them as it was envisaged in the financing agreement.