Senegal - Technical Assistance Project

Document of The World Bank FOR OMCAL USE ONLY Report No. 5956 PROJECT COMPLETION REPORT SENEGAL FIRTST TECHNICAL ASSISTANCE PROJECT TO THE PARAPUBLIC SECTOR (CREDIT 764-SE) December 2, 1985 Western Africa Regional Office Tbis docaunt has a restied distibutom and may be used by recipients only in the performance of tlir offica dutes. Its contemts may ot othrwise be disdosed withot World lak authoizaton.  LIST OF ABBREVIATIONS USED ACC Agent Comptable Central des Etablissements Publics, MEF BOM Bureau Organisation et Methodes, SG/PR CCCE Caisse Centrale de Cooperation Economique (France) CEP Centre des Etablissements Publics (ACC+COF), MEF CFP Controle Financier de la Presidence, SG/PR CNCA Commission Nationale des Contrats de l'Administration, SG/PR COF Controle des Operations Financieres, MEF CPSP Caisse de Perequation et de Stabilisation des Prix CVCCEP Commission de Verification des Comptes et de Contr^ole des Etablissements Publics, Cour Supreme DCA Development Credit Agreement DTAI Direction du Traitement Automatique de l'Information, MEF EP Etablissement Public EPA Etablissement Public a caractere Administratif EPIC Etablissement Public a caractere Industriel ou Commercial EPP Etablissement Public a caractere Professionnel ESGE Ecole Superieure de Gestion des Entreprises, Dakar FY Fiscal Year (July to June: Senegal and IDA) GESP Groupe d'Etudes du Secteur Parapublic IUT Institut Universitaire de Technologie, Dakar MEF Ministere de 1'Economie et des Finances ONCAD Office National de Cooperation et d'Assistance au D6veloppement OPT Office des Postes et Te1lcommunications du SEn6gal PCR Project Completion Report PE Public Enterprise (generic term) PMU Project Management Unit RCFS Regie des Chemins de Fer du Senegal SAED Societe d'Amenagement et d'Exploitation des Terres du Delta SEM Societe d'Economie Mixte SENELEC Societe Nationale d'Electricite SG/PR Secretariat General de la Presidence de la Republique SICAP Socifte Immobiliare du Cap-Vert SN Societe Nationale SODESP Socifte de Developpement de l'Elevage de la Zone Sylvo-Pastorale SOMIVAC Socifte pour la Mise en Valeur de la Casanance SONACOS Socifte Nationale de Commercialisation des Oleagineux du Senegal SONAR Societe Nationale d'Approvisionnement du monde Rural RCFS Regie des Chemins de Fer du Senegal TA Technical Assistance TOR Terms of Reference USB Union Senegalaise de Banque  FOR OMCIAL USE ONLY PROJECT COMPLETION REPORT SENEGAL - FIRST TECHNICAL ASSISTANCE PROJECT TO THE PARAPUBLIC SECTOR (CREDIT 764-SE) TABLE OF CONTENTS Page Preface ......................................... i Basic Data Sheet ......................................... ii Highlights ......................................... v I. PROJECT BACKGROUND AND OBJECTIVES ........................................ 1 II. PROJECT IMPLEMENTATION ............. ............................ 2 A. Audit Program ...2................ 2 B. Strengthening of Government Oversight Agencies ...................... 4 a) Strengthening of CFP .................................... 4 b) Strengthening of CEP ................................... 5 C. Assistance to the Government's Data Processing Department ........... 6 D. Training ............................... 7 E. Measurable Results ................. ... 7 III. MAJOR ISSUES .............................................................. 8 A. Project Design ........... .. 8 a) Government Oversight vs. Enterprise Autonomy ..................... 8 b) Dispersion of Resources among Competing Agencies ................. 10 c) Choice of Key Institutions for Implementation of the Reform ...... 10 d) Assessing the Need for Resident Advisors ......................... 11 B. Government Commitment and Implementation Capacity ........ ........... 12 C. Project Management Unit - GESP Chairmanship ......... .. .............. 12 D. IDA Supervision ......... ................ ............................ 13 IV. CONCLUDING REMARKS ...................................................... 14 Attachment: Comments from the Borrower ................. .. ..................... 16 ANNEXES Annex 1 - Project Background and Preparation .................................. 18 Annex 2 - Project Management .................................................. 26 Annex 3 - Project Supervision ................................. 29 Annex 4 - Strengthening of the Parapublic Oversight System .................... 31 Annex 5 - Technical Assistance to the DTAI .................................... 43 Annex 6 - Audits of Public Enterprises ........................................ 45 Annex 7 - Training Component ............ ...................................... 50 Annex 8 - Organigram of Parapublic Oversight System ........................... 55 This document has a restricted distribution and may be used by recipients only in the performance of their ofricial duties. Its conlenst may not otherwise be disclosed without World Bank authorization.  PROJECT COMPLETION REPORT SENEGAL - FIRST TECHNICAL ASSISTANCE PROJECT TO THE PARAPUBLIC SECTOR (CR. 764-SE) PREFACE This report is an evaluation of the First Technical Assistance Project to the Parapublic Sector supported by Credit 764-SE. The credit was approved in January 1978 and closed in November 1982. The Project Completion Report was prepared by the Public Enterprise Unit of the Western Africa Regional Office of the Bank, and is based on a review of the Senegal Parapublic Sector Report (No. 1619a-SE), dated June 1, 1977, the President's Report (No. P-1244 SE), dated January 5, 1978, the Development Credit Agreement dated March 17, 1978, internal Bank memoranda and reports on project preparation and supervision, progress reports prepared by the project director and the resident advisors financed under the project, a completion report prepared by the Borrower dated June 25, 1984 and other documents from the Bank's files or provided by the Borrower. A mission visited Senegal in July 1984 to interview officials involved in the project; consultants and Bank staff involved in the project were also interviewed. This report is divided into two parts. The project completion report proper, and a series of annexes providing detailed information on facts and issues relating to project preparation and implementation. The draft report was sent to the borrower for comments; the comments of the Director of the Second Technical Assistance Project are shown in the attachment to the report. The credit has not been audited by the Operations Evaluation Department.  - ii - PROJECT COMPLETION REPORT SENEGAL - FIRST TECHNICAL ASSISTANCE PROJECT TO THE PARAPUBLIC SECTOR (CREDIT 764-SE) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Expectations Current Estimate Total Project Cost (US$ million) 7.9 7.9 Credit Amount (US$ million) 6.0 6.3 Disbursed and outstanding - 6.3 Date Physical Components Completed June 30, 1982 June 30, 1982 Proportion Completed by above 100% 100% Date Economic Rate of Return NA NA Financial Performance NA NA Institutional Performance weak NA = Not Applicable OTHER PROJECT DATA Original Plan Actual First mention in files or timetable - February 1977 Government's Application February 1, 1977 Identification February 1977 February 15, 1977 Appraisal June 1977 July 4-15, 1977 Decision Memorandum August 12, 1977 August 10, 1977 Negotiations January 3, 1978 Dec. 12 and 13, 1977 Board Approval February 21, 1978 January 24, 1978 Credit Agrepment Date March 1978 March 17, 1978 Effectiveness April 1978 June 6, 1978 Expected Comple-ion date June 30, 1982 June 30, 1982 Closing Date December 31, 1982 November 16, 1982 Borrower Republic of Senegal Executing Agency Office of the Prime Minister Fiscal year of Borrower July 1 - .'une 30 Follow-on Project Name Second Parapublic Technical Assistance Project Credit Number Cr. 1398-SE Credit Agreement Date August 10, 1983 Credit Amount SDR 10.2 million (US$ equivalent 11 million) PPF214 October 15, 1982  - iii - DISBURSEMENTS BY FISCAL YEAR (US$ million) FY79 FY80 FY81 FY82 FY83 ANNUAL Appraisal Estimates 1.60 2.10 2.10 0.50 - Actual 1.03 1.96 1.29 1.64 0.38 Actual as % Estimate 64 93 61 328 CUMULATIVE Appraisal Estimates 1.60 3.70 5.80 6.30 - Actual 1.03 2.99 4.28 5.92 6.3 Actual as Z Estimate 64 81 74 94 DISBURSEMENTS BY CATEGORY uS$ million; Appraisal Estimate (DCA) Actual (1) Salaries and allowances of 2.11 3.24 advisors (including recruitment costs) (2) Consulting services 0.80 0.70 (3) Auditing services 1.63 1.97 (4) Equipment, materials 0.31 0.20 and vehicles (5) Training and 0.65 0.19 scholarships (6) Unallocated 0.80 - TOTAL 6.30 6.30  - iv - MISSION DATA No. of No. of Total Status/ Item Month/Year days persons Staff weeks Trend Date of Report Identification February 1977 1 February 1977 Appraisal July 1977 12 6 14 July 1977 Supervision I July 1978 6 4 5 July 1978 Supervision II April 1979 3 3 2 2/1 April 1979 Supervision III June 1979 4 1 1 2/1 July 1979 Supervision TV December 1979 12 1 3 2/1 January 1980 Supervision V March 1980 12 1 3 2/1 April 1980 Supervision VI October 1980 12 1 3 2/1 October 1980 Supervision VII November 1981 10 1 2 1/1 December 1981 Supervision VIII June 1982 * 6 3 4 1/1 July 1982 Supervision I-VIII FY79-FY82 23 * excludes time spent on appraisal mission for follow up Parapublic TA Project, Cr. 1398-SE COUNTRY EXCHANGE RATES (Yearly averages) Name of Currency : CFA Franc (CFAF) 1977 (Appraisal year) Exchange Rate US$1 = 246 CFAF President's Report (Jan. 1978) US$1 = 245 CFAF 1978 US$1 = 226 CFAF 1979 US$1 = 213 CFAF 1980 US$l = 211 CFAF 1981 US$1 = 272 CFAF 1982 (Completion year) US$1 = 329 CFAF  -v PROJECT COMPLETION REPORT SENEGAL - FIRST TECHNICAL ASSISTANCE PROJECT TO THE PARAPUBLIC SECTOR (CR. 764-SE) HIGHLIGHTS The parapublic technical assistance project to Senegal was the first parapublic project to be financed by the World Bank Group. The project was carefully prepared, building on the conclusions of the Bank's 1977 report on the parapublic sector in Senegal. The project yielded some important benefits. It contributed to a significant improvement in the overall quality and timeliness of public enterprise accounts and other financial documents. It introduced the audit concept on a large scale and contributed to an increase in accountability of public enterprise managers (paras. 2-10). It gave the Government a better picture of the situation and problems of the parapublic sector, and more importantly established mechanisms to update this information (paras. 11, 17). This may be a contributing factor in the Government's gradual move (uuder the current President's impulse, who was Prime Minister during this project's implementation) towards more rational policies regarding the parapublic sector. By providing a market for audits, the project significantly strengthened the local accounting profession, in numbers as well as quantity. These project benefits should now allow the Government to grant more autonomy to those enterprises whose management has improved and to relax the stringent regulations that govern most aspects of parapublic companies' activities. Nevertheless, implementation and supervision faced many problems. Instead of promoting a coherent and coordinated Government oversight system, the project strenghtened the various agencies' powerbase (para. 12); agencies' responsibilities overlapped which resulted in excessive controls over some enterprises and in insufficient controls over others. Morever, an agency set up to be a formal financial control agency should not have been entrusted with the key responsibility of exerting operational and strategic supervision over the parapublic sector. The importance of a strong project management unit was underestimated (para. 36). The unit did not intervene to attempt to coordinate the various agencies' activities, nor did it try to solve the problems that arose in the execution of the project's components or to ensure adequate follow-up of the various recommendations made by auditors, advisors or consultants. Weak supervision only compounded this problem (para. 37), as did the uneven qual'ty of consultant's work and lack of Government leadership over the parapublic sector reform process.  - 1 - PROJECT COMPLETION REPORT SENEGAL - FIRST TECHNICAL ASSISTANCE PROJECT TO THE PARAPUBLIC SECTOR (CREDIT 764-SE) I. PROJECT BACKGROUND AND OBJECTIVES 1/ 1. The Technical Assistance Project to the Parapublic Sector, signed on March 17, 1978, was the first of its kind, not only in Senegal or the West Africa Region but also Bank-wide. Traditionally the Bank had approached the problems of the parapublic sector through assistance to specific companies in the context of its various projects. In 1976, the Bank as well as the Government of Senegal realized that a horizontal approach was necessary to address problems faced by all or most public enterprises and by the Government's management of its enterprise portfolio. A Bank mission was sent to Senegal to prepare a report on the parapublic sector as a whole, and the Government established an inter-agency task force on the parapublic sector (GESP). 2. The Parapublic Sector Report, of which a first draft was discussed with the Government in the Fall of 1976, emphasized the need to assign specific objectives to public enterprises (PEs), to find ways to improve their operations and to develop the Government's strategic and operational control over the parapublic sector, which had grown without restraint in the early and mid 1970s. The Government agreed with the analysis and the recommendations of the Report, and approved a major reform law of its parapublic oversight system integrating a number of the Bank's recommendations. It also requested IDA assistance to implement these reforms. 3. The reform law (law 77-89) became effective in August 1977. Following the Report's recommendation that public enterprises be given more autonomy, it created a new category of wholly Government-owned public enterprises, 2/ the Societes Nationales (SN), which benefit from a much larger degree of autonomy than the existing Etablissements Publics (EP). Law 77-89 also modified the Government's oversight system 1/ Further details on project background and preparation are provided in Annex 1. 2/ The term public enterprise (PE) refers in the rest of this report to all SNs, all EPICs, and all SEMs in which the Government has a majority stake or that are submitted to CFP controls by decree.  over the parapublic sector substantially. 3/ It located the main responsibility for overall financial oversight over the parapublic sector and for control over the SNs, the mixed companies - Societes d'Economie Mixte (SEM) -, and selected EPs to a special parapublic division to be established in the Controle Financier de la Presidence (CFP), the Presidency's financial control agency. 4. The IDA project was identified in February 1977 and appraised in July 1977, after approval by the Cabinet of the parapublic reform bill; the Credit became effective in June 1978. The project's objectives as summarized in the President's Report were to support the Government's reform program by i) improving the efficiency of Government agencies in charge of strategic analysis, operational supervision and formal financial control over the parapublic sector; and ii) upgrading the auditing, financial management and accounting procedures for the sector as a whole and for a number of the most important public enterprises in particular. 5. The Credit financed i) eight expatriate advisors to three Government oversight agencies for a three-year period while calling for a substantial increase in their local staff; ii) audits of seven public enterprises and technical assistance to two of these enterprises (compared to four envisaged at appraisal); and iii) a training program for the staff of the oversight agencies, for Government representatives on the boards of public enterprises and for public enterprise staff. During implementation various components were added to the initial project design. II. PROJECT IMPLEMENTATION 6. The implementation of the various project components is reviewed in detail in Annexes 4 to 7. Major highlights are reviewed in this section. A. Audit Program 7. The audit program was the project's most successful component. Beyond the direct benefit of a greater reliability of company accounts the audits provided a greater awareness i) of the weakness of the existing accounting systems; ii) of the precise situation and problems of the audited companies; and iii) of the usefulness of audits, which hitherto had not been widespread in Senegal. They allowed to build up the Senegalese accounting profession through: i) the establishment of additional local firms; ii) joint ventures between such firms and 3/ An organigram of the Parapublic Oversight System is provided under Annex 8.  - 3 - established international firms; iii) return of Senegalese auditors working abroad, lured back by an increased demand for top-level Senegalese auditors; and iv) general improvement in the quality of local auditing. The award of audit contracts to joint ventures of local and foreign firms was a very positive mechanism for the transfer of know-how, and was continued under the follow-up project; similar mechanisms should be explored for the transfer of know-how in other fields, such as management consulting. 8. The audit component under the project was designed as a pilot project to be replicated if successful. Managed by CVCCEP, it allowed this judiciary audit agency to expand its work program to a larger number of companies. The audits led other parapublic companies not included in the project to request the inclusion of their company in future audit programs and in some cases to finance their own limited audits in order to establish data on which management could assess performance and base its policy decisions. The experience gained under the first parapublic project led to improvements in the design of audit programs under the follow-up project. With one exception, it appeared that the accounts of the companies audited under the project were so poorly kept that full scale audits were not warranted; what these companies needed first and foremost was accounting assistance. The project provided such accounting assistance to two of the seven audited companies. An important lesson to be drawn from this component is the opportuneness of adapting auditors' TOR to the specific situation of each company, mainly by deciding on the optimal mix and phasing of classical auditing and technical assistance (for accounting, systems and procedures, etc.). When the available information on the company's situation is not sufficient to make that determination, the contracts should allow for enough flexibility to modify the TOR in view of the needs that will emerge from the first phase of the auditors' intervention. 9. The major shortcoming of the audit program was the inadequate implementation of the auditor's recommendations. One reason for this was weak monitoring by management, Government and IDA. Indeed, audits are not an end in themselves and become fully effective only if they are used as a key input in the management or oversight of the company. We would recommend that future audits be to the extent possible focussed either on PEs included in the Bank's lending program, or on PEs whose importance is such that a Bank division is willing to formally allocate the necessary staff resources to monitor the audit and the implementation of its recommendations. Audits of other companies should be financed by the Bank if sufficient conditions or guarantees are present; these could cover PEs identified as potential candidates for divestiture or liquidation, for example; they could also include PEs which are closely monitored by another donor or by a Government agency. Follow-up of the implementation of audit recommendations by the various oversight agencies has also left much to be desired. A better division of responsibilities and closer cooperation between these agencies is necessary. CFP should be made formally responsible for monitoring such implementation by the companies on an on-going basis and for assisting companies where needed. CVCCEP, which alreadv reviews the auditor's  - 4 - reports in plenary sessions and addresses official recommendations to PE management, would check on the implementation of these recommendations when it does its own audits (most PEs are only audited every four or five years by CVCCEP). 10. Now that the principle and practice of external audits of PEs has been established in Senegal, the first priority should be to optimize their use as an effective management tool. To fully justify the US$2.0 million spent on these audits, improvements in company management would need to exceed substantially the marginal levels achieved by project completion. B. Strengthening of Government Oversight Agencies 11. The main objective of the Project was to support the reform of the parapublic sector oversight system introduced by law 77-89. A major benefit of the project in this area was the substantial improvement in the quality and availability of data on the parapublic sector made possible by the closer monitoring of these companies. This has allowed the Government to have a better idea of the enterprises' performance, and is also starting to provide the basis on which sounder sectoral and macroeconomic strategies can be built. 12. The strengthening of the various oversight agencies also had its drawbacks. Indeed it reinforced each agency's powerbase, at the expense of overall coordination and resulted in a relatively inefficient and bureaucratic oversight system. The weak PMN1 did not make a serious attempt to integrate the various components i-1 coherent whole. Various coordination instances were established, but had no authority to effectively coordinate and could not prevent the agency heads from trying to build up their own agency. A lack of cooperation between ministers at the cabinet level amplified this problem. These issues are further discussed in Section III (Major Issues). The strengthening of the various oversight agencies makes it now even more difficult to restructure the government oversight system around clearly defined tasks. a) Strengthening of CFP 13. As is explained in par 32 below, the choice of CFP as the centerpiece of the parapublic oversight system was not a fortunate decision. CFP management was not committed to the project's objectives and it took a long time before an acceptable division chief was appointed to head the newly established parapublic division. Although the appointment of a competent CFP division chief was a condition of negotiations, it was later dropped when all other conditions were fulfilled; this was a mistake. The sudden influx of four technical assistants in an agency that did not have any befote and was used to a very independent way of operating also caused maijr problems. The lack of unity within the TA team and their insensitivity to Senegalese work methods and priorities made matters even more difficult. Conflicts  - 5 - between CFP staf' and the TA team were endemic, centering on opposite management styles and personality conflicts. 14. At the stage of project preparation, precautions had been taken to guarantee adequate and gradual recruitment procedures to build-up CFP. However, the trippling of staff envisioned over a period of two to three years, and which actually occurred over a one-year period (from eight in 1978 to about 25 in January 1979), was too much for CFP to absorb. Similarly, recruiting and retaining high level national staff with experience in administration and business management, which was called for by the DCA, may be seen as unrealistic due to relatively unattractive salaries for managers and lack of promising career prospects and incentives. Most recruits lacked adequate experience or a strong background in finance. Training was provided i) on-the-job by the TA team; ii) through internships in foreign accounting firms; and iii) through seminars organized by BOM. However, the absence of career prospects at CFP has led to a high turnover in staff whose career prospects after a few years of training and experience at CFP are much better in private or parapublic enterprises than at CFP. Although this high turnover had adverse effects on the development of CFP's capabilities, it had positive side-effects through the transfer of trained state controllers to management positions in public enterprises. In order to build up CFP's capabilities, it would have been necessary to establish attractive career prospects to retain the best staff. 15. The President's Report had identified the recruiting and retaining of suitable Senegalese at CFP as a major project risk, but not much was done to find a satisfactory solution to this problem. Finding suitable advisors and ensuring their cooperation with local staff was also identified as a project risk. IDA staff was heavily involved in the selection of consultants (with disappointing results however) but did not attempt to solve accute problems between technical assistants and local staff when such problems arose. 16. A major weakness in the implementation of this project component, however, was due to the opposite views held by CFP management on the one hand and the TA team and IDA on the other hand regarding the agency's mandate and priorities. Indeed, CFP management never agreed to take on the strategic supervision of the sector as a whole, preferring to limit its role to the less controversial function of formal control, whereas IDA considered strategic analysis of the parapublic sector as a key functi n of CFP. b) Strengthening of CEP 17. Tne assistance to CEP, the agency in the Ministry of Finance in charge of EPs, was more successful (especially at ACC). This success was to a large extent due to the quality of management and TA as well as a clear mutual understanding of the objectives of the assistance. The TA team worked in close cooperation with CEP management and staff.  Their contribution in training CEP staff and strengthening CEP and in improving EPs' budgeting, accounting and financial management was very valuable. 18. The progress achieved in EP management and financial systems should have made possible the phase-out of a priori controls and of the central treasury pooling scheme 4/ , as was expected at project appraisal. The last two directors of COF, the department within CEP in charge of a priori controls, agreed that the phase-out of these cumbersome controls would allow companies to recover the degree of autonomy that is necessary to successfully manage a public enterprise and would enable CEP to redeploy staff resources from unproductive controls, which take up more than 75% of their time, to management assistance. There was a built-in tension in this component between the short-term objective of strengthening EPs' accounting and financial procedures by strengthening CEP and the longer-term objective of increasing EP autonomy by phasing out a priori controls and the central treasury pooling system. Phasing out of a priori controls and introducing a more flexible treasury system (if not an autonomous enterprise treasury) for the industrial and commercial EPs are objectives on which most parties seem to agree and that should be further pursued under the follow-up project. In case such a program is implemented, the role of CEP will have to be reassessed. Should CEP be involved in the business of management assistance, or should another agency (we think of BOM in particular) take over such responsibility? As CFP and CVCCEP also claim a role in management assistance to PEs, the future role of CEP would have to be reassessed in the broader context of a general restructuring of the Senegalese parapublic oversight system. C. Assistance to the Government's Data Processing Department 19. The assistance provided under the project to the Direction du Traitement Automatique de l'Information (DTAI), the Ministry of Finance's data processing department, was the least successful of all project components. It provided 36 man/months of technical assistance mainly for training of DTAI staff and was added during project execution without proper appraisal, prior TOR or competitive bidding. The objectives of the component did not fit into the project's overall objectives and were not well specified. This lead to different interpretations of the contract by the consulting firm and DTAI management, to numerous conflicts and to a very meager output. The implementation of this component was not supervised at all. 4/ EPs do not manage their own treasury operations; they are "correspondants du Tresor" which means that all their receipts go to the Finance Ministry (Treasury) which also handles all their payments.  -7- D. Training 20. On-the-job training was provided by the various TA teams under the CFP, CEP and DTAI coimponents. The project's training component proper was entrusted to the Bureau Organisation et Methodes (BOM) at the Presidency, which received six man-years of technical assistance to help manage it. The TA team's effectiveness was seriously undermined by high turnover (six experts for two position, over a three year period). Short and longer term seminars were organized for staff of the oversight agencies, for PE general managers and accountants, as well as for Government representatives on PE boards. State controllers and CVCCEP staff benefited from overseas training in international accounting firms. The quality of these programs varied widely. The authority of BOM regarding training of oversight agencies' staff was challenged from the beginning by the other agencies; the component was not well integrated in the project. Various supervision reports characterized BOM management as weak and lacking initiative. The main weakness of the component, however, was that it did not establish a permanent system for the training of the various target groups. With the end of the project also came the end of most of the --aining activities that had been initiated under the project. The suggestion made by a 1979 Supervision Mission of the Training component to transfer training activities from BOM to ESGE was unfortunately not further examined. Entrusting training activities to the Dakar management school would most likely have improved the sustainability prospects of this component. E. Measurable Results 21. The President's Report indentified a number of measurable results to be achieved by project completion, namely a) Improve basic data on the operations of public enterprises: The final objective was achieved. By June 1982, detailed financial data were available on the financial performance of about 90% of the public enterprises, including aggregate data by sub-sector of the economy and consolidated data for the whole parapublic sector. b) Develop indicators at the sector and company levels to check on the achievement of development objectives: The final objective was not achieved. The link between the Government's overall development objectives and sectoral strategies on the one hand and company objectives and targets on the other remains very weak. The data base referred to above, however, includes key financial ratios for the enterprises, the sub-sectors and the parapublic sector as a whole.  - 8 - c) Definition and initial implementation of a program of improved financial management for the enterprises that benefited from in-depth audits, TA and training: Partially achieved. Implementation of such a program started in two of the seven enterprises audited, namely OPT where it was relatively successful, and SAED where it ran into major difficulties. III. MAJOR ISSUES 22. Four key factors contributed to the problems that arose during project implementation, namely problems of project design, waning Government commitment to parapublic sector reform, a weak project management unit (PMU), and poor supervision from Bank headquarters. A. Project Design 23. In general the project was well prepared, building to a large extent on the recommendations of the Bank's 1977 Parapublic Sector Report. The weaknesses mentioned below may be partly explained by political expediency and lack of prior Bank experience in parapublic sector management. a) Government Oversight vs. Enterprise Autonomy 24. The Bank's parapublic sector report, the Government's parapublic reform program and the IDA project emphasized primarily the problems related to the central management of the parapublic sector, and not the direct assistance to the individual companies. This approach followed a trend started with the parapublic sector law of 1966, which created the Centre des Etablissements Publics (CEP) at the Finance Ministry and established stringent controls on EPs' operations in exchange for the takeover by Government of their debt burden, and continued with -he parapublic sector law of 1972 which created CVCCEP, the parapublic sector auditing agency at the Supreme Court. Centralized Government control was deemed necessary as a check on PEs' poor management; simultaneously, the oversight agencies were seen not only as controllers but also as advisors to the enterprises. 25. The inherent tension between centralized control and increased autonomy became a serious problem in this approach. Strengthening enterprise performance by strengthening Government oversight agencies, and strengthening Government control by creating new control agencies turned out to be an internally contradictory and ineffective approach, dealing with long-term problems through short-term ad hoc responses. To increase enterprise performance it would have been necessary to grant management sufficient autonomy, to hold it accountable for the enterprise's performance and to link such performance with positive or negative sanctions. This would have required a change in the institutional and legal framework governing the parapublic sector, a  break in the tradition of civil-service-style management of public enterprises in favor of an entrepreneurial style. The strengthening of oversight agencies, staffed by career civil servants who often resent the higher pay PE managers get and who are not overly concerned with the promotion of an entrepreneurial spirit, resulted in closer government supervision (tutelle) and did not favor increased accountability of PE managers. 26. A second condition of successful strengthening of PE performance not well addressed by the project was the strengthening of BE management. Assistance to PEs was provided on an experimental basis to two companies under this project, namely OPT and SAED and was to some extent successful in the first but not in the latter case. The follow-up project now under execution includes the provision of such assistance to up to thirty public enterprises. 27. The inadequacy of this "State Control" approach led to the introduction of the contrats-plans system. The issue of a contractual form of Government-PE relations had already been raised incidentally in the Bank's 1977 report (par 6.05): "Supervision needs to be carried out on a contractual basis for enterprises to be able to manage themselves. Management by objectives could thus meaningfully be introduced under which each enterprise would negotiate with Government annually the targets to be set for it and would then be judged according to the results it achieved." This interesting idea was not reflected into the project design, however, but emerged again in 1980 with the introduction of the contrats-plans (program contracts between a public enterprise and government) and to a lesser extent with the issuance of "circulaire 008" introducing budgeting by objectives. When the Government decided in 1980 to modify its oversight mechaaisms over key public enterprises by introducing the contrat-plan concept, it entrusted the management of this new instrument to a newly established unit in the Primature (Cellule des Contrats-plans), and asked IDA to support this unit. 5/ This decision apparently resulted from the failure of CFP to accomplish the objectives of operational and strategic oversight it had been assigned under the project; it also reflected the Government's intent to shift oversight over PEs from control over their operations to monitoring of the implementation of mutually agreed objectives. Once more however, a new agency and new procedures were established, but the terms of reference of existing agencies remained unchanged and old procedures remained in effect. 28. ien if the initial project design can be attributed to the Bank's inexperience, it is more difficult to see why the approach was never reconsidered, especially as the performance of virtually all 5/ Cr. 1061 financed the salary of the unit's director. One may wonder, however, why the Parapublic Project was not reoriented to support this new strategy.  - 10 - enterprises in the sector began to show dramatic deterioration over the project period, with aggregate operating losses (even excluding ONCAD and CPSP) increasing according to Bank staff estimates from roughly 6 billion CFAF in 1978 to close to 30 billion CFAF by 1982. (In fact, the aggregate situation of the sector at the start of the project was a substantial surplus.) In the face of this, one cannot help but get the impression that the First Parapublic Project was not getting a handle on the real problems, and remained oblivious to them throughout its implementation. 29. The danger of reinforcing the Government's control apparatus without concomittant strengthening of PE management and eEtablishment of a regulatory framework favoring entrepreneurship may be one of the stronger lessons to be drawn from this project. b) Dispersion of Resources among Competing Agencies 30. The project supported the various oversight agencies on an individual basis, without clear delineation of each agency's terms of reference, with little coordination among them and with little integration of their interventions with the Technical Ministries, the Ministry of Finance or the Ministry of Planning. Specific project components were designed for each agency without adequaLe linkages. 31. Furthermore, the reform introduced by law 77-89 and supported by the project resulted in practice in a hybrid government oversight system over public enterprises--modulated on a legal classification of Etablissements Publics (EP) supervised primarily by CEP on the one hand and Societes d'Economie Mixte (SEM) and Societes Nationales (SN) supervised by CFP on the other--that does not rest on solid economic or managerial grounds but rather on historical ones; e.g., some commercial and industrial enterprises come under the tight control of CEP and to a lesser extent the regular supervision of CFP, whereas others are only under CFP's supervision. This system disperses scarce resources over different agencies when they could be better put to use in one agency. c) Choice of Key Institutions for the Implementation of the Reform 32. The second important issue in project design was the choice of CFP, the Presidency's financial control agency, as the main locus for Government oversight of the parapublic sector. This choice, which was dictated by decisions made by the Government, was not very fortunate. First, CFP was not sufficiently associated in the project preparation although it was to become the centerpiece of the Government's parapublic oversight system. This resulted in misunderstandings and differing views on the agency's mandate and priorities as well as on the technical assistance team's terms of reference. Second, the objectives assigned to CFP by the IDA project were too ambitious for an agency perceiving its function as one of financial control over about seventy public enterprises and not of macroeconomic and strategic analysis or even of operational supervision; moreover, the mandate given to CFP by the reform law clearly emphasized its formal financial control function,  - ii - which was not reflected in the project designi. Thilrd, CFP was not under the authority of the Prime Ministor, wh. w.; the main driving force behind the parapublic reform progrrnm; iL became difficult for the Prime Minister to influence CFP management wlho repolrted directly to the President. Fourth, no adequate linkages were established between the Ministry of Economy and Finance, whichi controls the key instruments in the Government's policy towards thc par;apiublic sector (subsidies, taxes, debt management, portfolio mangcent, cenLral treastry pooling system for EPs, controls over EPs, data processing, etc.), and CFP, which was supposed to be responsible for m -roeconomic and strategic analysis over the parapublic sector. The in.-ragencv t;ask force on the parapublic sector (GESP), on which CFP' ard the Ministrv of Finance were represented and which was to be "responsible for the coordination and the execution of the project" (DCA, section 3.08(a)(A)(ii)), did not have the necessary authority or influenret t' coordinate their activities; this was compounded by the absence of cabinet-level coordination of the Government's approach to the parapublic sector. With hindsight, an attempt should have been made to involve the Ministrv of Finance more closely in the oversight of the parapublic sector even though a major aspect of the reform program supported by IDA was to remove such responsibility from this Ministrv in response to its poor past performance and related political problems; these issues should have been addressed directlv instead of reshuffling responsibilities among agencies. 33. The parapublic project's experience shows that i) the choice of the proper institution(s) for the m.anagement Lind oversight of the parapublic sector, ii) a clear delineation (f responsibilities at all levels, and iii) the establishment of efficient linkages between the institutions (actors) involved are critical factors for the success of a parapublic reform program. d) Assessing the Need for Resident Advisors 34. Another pervasive issute during project implementation pertained to the proper role and .rerogatives of resident advisors. The main lesson is that the preparation of these projects should address the following questions. Is there a real need to involve expatriates in the process, or are there qualified nationals available? Is there a real need for resident advisors or could part-time advisors do a more effective job? Should resident advisors be in line or in staff positions, should they have direct authority over local staff, should they have the right to communicate directly with Cabinet members or high officials outside their agency? Should their primary function be to run an agency, to design systems and procedures to have it run efficiently by national staff, or to train counterparts, or should it be a blend of the three functions or a gradual shift over the assignment from the first to the latter? Also, did the agencies that received such assistance request it in the first place, and did they get involved in the preparation and approval of the advisc,rs' TOR? How should personal conflicts between resident advisors and agency staff or management be addressed by the PMU or the IDA supervisors'? These issuies were not  - 12 - satisfactorily addressed by the Project. The TOR of the resident advisors were not sufficiently defined; their role (advisory, operational, training) was not clearly identified, which led to differing interpretations and tensions between the agencies and the technical assistants. Local staff resented the authority claimed by some resident advisors. Some agency heads used the advisors as their personal assistants, entrusting them with specific missions at the expense of helping build up the agencies' own capabilities. Also, in the case of CFP, the influx of four technical assistants was too large for the agency to absorb. B. Government Commitment and Implementation Capacity 35. Government commitment at project appraisal appeared strong, as illustrated by the enactment in August 1977 of the parapublic sector reform law. The Prime Minister was the driving force behind this reform program, whereas the agencies to be strengthened under the project reported to the Presidency or the Ministry of Finance. The only assistance to the Prime Minister's office came at mid-project in the form of the part-time secondment of a CFP consultant for the establishment of a Contrats-plans unit at the Prime Minister's office, dealing with the negotiation of agreements between key public enterprises and the Government. The PMU although located at the Prime Minister's Office did not contribute to the strengthening of that office. Once the law and the ensuing decrees had been enacted, the Government apparently did not follow-up on their proper execution or on the broader objectives of the reform. Each agency- was allowed to develop its own approach without any real attempt to coordinate their activities. Fading Goverrment commitment to a more rational and efficient parapublic sector was identified by the appraisal mission and the President's Report as a project risk. It seems that the Prime Minister did not manage to instill this commitment into the key oversight agencies and ministries. Clearly, the commitment of one person, be he Chief of Government or Chief of State, was not sufficient to secure the support of ministries and agencies in charge of the implementation of project objectives. Moreover the Prime Minister apparently became aware of the need to consolidate the control over the parapublic sector and the management of the Government's portfolio into one ministry. Indeed, in January 1980 he suggested that GESP report to the Minister of Finance rather than to the Prime Minister. IDA opposed this move and GESP remained at the Primature. C. Project Management Unit - GESP Chairmanship 36. The need to coordinate project management with the management of the parapublic reform process had been perceived at appraisal. The DCA (section 3.08(a)) stipulated that GESP, the interagency taskforce on the parapublic sector, would be responsible for the coordination and the execution of the project and for the supervision of the PMU. The GESP chairman was also PMU director; a deputy-director was to be in charge of  - 13 - the project's day-to-day management. The performance of the successive project directors (GESP chairmen) and deputy-directors left a lot to be desired. The importance of a strong PMU management and GESP chairmanship was underestimated; it was felt that it was more important to place c:ompetent managers in public enterprises or in the oversight agencies cather than in the PMU (GESP chair), which was perceived primarily as a unit in charge of project administration and not as the driving force behind the reform of the parapublic sector. This left both the reform and the project without real leadership, and contributed significantly to weak performance on both fronts. 6/ Fu-thermore, even the administrative function was not dealt with properly by the PMU. Further details on project management are provided in Annex 2. D. IDA Supervision 37. The high turnover of project supervisors, the absence of formal supervision during a full year of project execution, the lack of feedback and follow-up from headquarters on audit reports and other project-related documents contributed to the project's weak performance, despite the Resident Mission's efforts to compensate for headquarters' shortcomings. The amount of supervision needed for such a complex institution-building process was clearly underestimated, as was the difficulty of coordinating the various Bank divisions' inputs; the Bank's organizational structure was not well adapted to the needs of the supervision of a cross-sectoral project of this type. The project should have benefited from above average supervision coefficients and strong cooperation between concerned Bank divisions. The weak IDA supervision effort was a problem, especially since this was IDA's first Parapublic TA project, the project generated an enormous number of reports, a large number of resident advisors and short-term consultants had to be supervised, and major problems plagued the project throughout its execution; some project components should have been reassessed or readjusted during project implementation, when problems arose. The project agreement included $100,000 for three interim evaluations and ore final evaluation of the project by consultants, but this proviso was unfortunately waived in 1980. In view of the novelty of this type of project, it would have been useful to include at least one thorough interim evaluation after 12 or 18 months involving all concerned parties. Such evaluation could have led to the reassessment of the project and to changes in its design; they would have been extremely valuable contributions to project supervision and to the preparation of the follow-up project. Further details on project supervision are provided in Annex 3. 61 The follow-up project tried to address this problem by locating the PMU in the Cellule des Contrats-Plans at the Presidency.  - 14 - IV. CONCLUDING REMARKS 38. Institution building is a sustained long-term process. Designing an alternative administrative structure, as was done during project preparation, sanctioned by law is not sufficient to ensure the success of the planned reform. Although we believe that the basic design of the parapublic oversight system and of the internal organization of CFP, the system's centerpiece, was far from optimal, we also think it could have been made to work more efficiently during project execution. In view of the problems that arose, some project components should have been reassessed. More importance should have been given initially to making sure we had a real, profound agreement on project objectives with Government in general and implementing agencies in particular and to securing adequate leadership and commitment to implement the parapublic sector reform. Also the lack of mutual understanding, the absence of real leadership and waning commitment we noted should have been dealt with more forcefully during project execution. i9. The success of a parapublic operation of this type critically depends on effective coordination at various levels: i) coordination between the various agencies and ministries involved in the management of the parapublic sector; ii) coordination of project components and TA efforts; iii) coordination at the Cabinet level; iv) coordination of IDA supervision, and finally v) coordination between these various levels. The number of actors involved and the cross-sectoral nature of parapublic issues make the management of such parapublic operations extremely complex. 40. The assistance to CVCCEP, CEP, CFP, as well as the various audits financed under the project have revealed where the main problems or bottlenecks of public enterprises and of management of the parapublic sector lie. These diagnoses and recommendations have not always been followed-up adequately and tend to be repeated in consecutive reports. More attention should thus be paid to the actual implementation of the recommendations, which may require specific and punctual assistance. Identifying what the problems are is one task, for which the controllers of the oversight agencies are now better equipped; finding ways to solve them and implementing these solutions is another task which these agencies have not been able to deal with. The challenge facing the Government today is to use this vastly improved knowledge of the parapublic sector to develop innovative and courageous reform programs, to integrate them in a broader context, to make some tough but necessary decisions. 41. The project focussed too much on the control mechanisms and not enough on the management of the public enterprises. The problems with this centralizing approach were compounded by the moderate to poor quality and bureaucratic background of staff in these oversight agencies. The Structural Adjustment Loan (Cr. 1084-SE/Ln. 1931-SE), building on the experience of the First Parapublic TA Project disbursed much of its local funding to parastatal companies which had concluded a  - 15 - contrat-plan with Government approved by the Bank. This effort, which also had its problems, was guided by the Prime Minister's contrats-plans advisor financed under the TA Project for Economic and Financial Planning (Cr. 1061-SE), which began to turn the Bank's attention toward helping individual public enterprises. The follow-up parapublic project changed this emphasis by focussing on the strengthening of the management of the country's key enterprises. 42. The need for a follow-up operation had been anticipated in the President's Report (par 47) "since it will take more than three years to establish an effective institutional structure for exercising strategic analysis, operational monitoring, and in some cases, financial control over companies in the parapublic sector". A project completion report, or at least a preliminary evaluation report, should have been prepared prior to the appraisal of IDA's follow-up project, however.  - 16 - COMMENTS FROM THE BORROWER (Translation) Attachment REPUBLIC OF SENEGAL OFFICE OF THE PRESIDENT OF THE REPUBLIC SECRETARIAT-GENERAL Project Unit: Technical Assistance Project to the Parapublic Sector Office of the Director Dakar, October 21, 1985 Our ref: No. 1,224/PR/SG/CP The Director Operations Evaluations Department Dear Sir: Subject: Comments on Project Completion Report on First Technical Assistance Project to the Parapublic Sector (Credit 764-SE). Thank you for the Project Completion Report on this matter. Having read it, I should like to make the following remarks: 1. Audit Program It is suggested on page 3 of the Report that the audit program should focus on those public enterprises marked for divestiture or liquidation. In our opinion, enterprises to be liquidated should be excluded from the program, since they are not the subject of auditors' recommendations. Any audit should, in principle, produce a series of recommendations on how the enterprise could effect improvements; there is no point, however, in improving an enterprise in liquidation, since its dissolution is inevitable. */ *1 IDA comment: The PCR states that the audit program could also cover public enterprises identified as potential candidates for divestiture or liquidation. In such cases audits would be used to help decide whether the enterprise should be privatized or liquidated.  - 17 - 2. Annex 4: Strengthening of the Parapublic Oversight System Page 18, B (b): This position title in French should be Chef de la Division du Secteur parapublic and not Directeur de la Direction du Secteur parapublic. The discussion here concerns a Division of CFP headed by a Chief and not a Director. Page 19, paragraph 18: CFP should be said in French to consist of three Divisions (and not three Directions), each of which is headed by a Chief (and not a Director). 3. Annex 7: Training Component A number of CEP staff members also received overseas training, although Section D on page 35 notes only that CFP and CVCCEP staff did so. ** From my personal viewpoint, the Report provides a full inventory of the various components of the first technical assistance project to the parapublic sector. However, since I did not follow execution of all those components closely, being at the time a member of one particular institution, I was fully aware only of the part of the project which involved that institution. Lastly, on page iv of the Report a list of project disbursements by category is provided. In our view, the Report would be still more useful were it to show, in respect of each institution affected, the different sums spent under such headings as training, accounting assistance or audit program, for instance. This would give a clearer idea of the overall cost picture to be taken into account in the futire. ***/ Many thanks for having forwarded me a copy of the Report. Sincerely yours, /s/ Serigne Ahmadou CAMARA Director, Second Techrical Assistance Project to the Parapublic Sector **/ Comment was incorporated in Annex 7, par. 10. ***/ We agree, but the project's poor accounting practices and the lack of project audit would make such a breakdown difficult.  - 18 - ANNEX I Page 1 of 8 PROJECT BACKGROUND AND PREPARATION I. Bank Parapublic Sector Report A. Background 1. In early 1976 a World Bank mission visited Senegal to study the parapublic sector. This decision was the result of several factors. The performance of tbe parapublic sector had become a concern not only to the Bank but also to Government. The sector accounted for a large part of economic activity and the accumulating deficits of a growing number of PEs were a cerain on public finances. Most Bank Group projects were channelec through parapublic sector companies whose generally poor performance financially and operationally was affecting project implementation. In some cases, the companies themselves were judged at fault. In others, however, poor performance had been caused largely by restrictive Goverrment control, disorderly financial and management relations between Government and the parapublic sector, and by lack of clear Government policy towards and effective coordination of the parapublic sector; the traditional project-by-project approach had often failed in addressing such problems which were beyond the control of the projects' management. In suamarizing its preliminary conclusions the Bank emphasized the need for Government to assign specific objectives to pnr=public companies, improve its strategic and operational control over the sector, which had grown without restraint in the early and mid 1970's, and find ways of improving company operations. B. Findings 2. The Parapublic Sector Report, discussed with the Government in November 1976, identified many problems. The Government could not monitor performance since objectives were rarely clear. In many cases, Government interfered heavily in company operations. Overstaffing and high salaries prevented many companies from being efficient and producing at competitive costs. They were rarely able to finance their own investments. A few companies accounted for the bulk of the sector's deficit. There were few trained managers and technicians at the middle and upper levels, and those selected for top jobs were often civil  - 19 - ANNEX I Page 2 of 8 servants whose training in law and public administration was usually inappropriate for solving problems in parapublic companies. 1/ 3. The report found that Government control of parapublic companies was diffuse and ineffective, and described supervision by CEP, (although lacking formal authority) as more effective in influencing company policy decisions than that done by the Ministry of Finance. The report referred to the CFP as the agency with the most complete body of information on public enterprise and stated that "the success of the Financial Controller proves that a full-time team of highly qualified persons close to the center of Government strategy-making is necessary to achieve successful monitoring" (par. 6.12). C. Proposed Approach 4. The Report emphasized the need for Government to define national and sectoral development priorities including the role of parapublic companies. Supervisory ministries would be responsible for implementing national strategy in sectors under their control but should not interfere in day-to-day operations of companies. Government should establish contractual relationships (ma agement by objectives) with public enterprises on an annual basis with precise goals and provision of adequate financial means and operational autonomy to achieve them. Monitoring should be designed to collect data and measure public enterprise performance against a set of operational indicators. With emphasis on the strategy and monitoring, the control function would focus on a posteriori financial oversight and audits rather than a priori financial control which hampered company operations. The report recommended that Government place less emphasis on "control" - a curative approach which was not working - and more on prevention through strategy and monitoring. D. Government Reaction to the Report 5. The Government, particularly the then Prime Minister (now President), reacted favorably to the report but expressed fear that without outside help the Government would be unable to implement the report's recommendations. The Bank's position was that the Goverrment 1/ The report also argued that (i) Government often placed parapublic companies in impossible financial positions by keeping tariffs artificially low without providing offsetting subsidies; (ii) large accounts receivables mainly by the Government were a major cause of the financial problems; (iii) many parapublic companies were inadequately capitalized at their establishment and mnable to realize their objectives; and made sub-sector specific recommendations for agriculture (assuring long term financial viability) and tourism (creating a holding company for Government's scattered investments).  - 20 - ANNEX 1 Page 3 of 8 needed to make some political decisions about how to reform parapublic sector control and which institutions should handle the reform, before the Bank could consider a technical assistance project. Within a day of his decision on how to reform control of the sector, the Prime Minister sent an official request to IDA for such assistance (February 1977). II. Parapublic Sector Reform Law (Law 77-89) 6. The National Assembly approved a Parapublic Sector Reform Law (Law 77-89) in August 1977 which changed the organization and basic legal structure of public enterprises and transferred the primary responsibility for oversight of the paraptiblic sector to a new parapublic division at CFP in line with the recommendations contained in the Bank's report; this shift followed a trend by which each major parapublic reform law created a new institution (CEP in 1966, CVCCEP in 1972 and parapublic division of CFP in 1977). IDA reviewed and agreed to the text of the new reform law. 7. Law 77-89 classified public enterprises into three categories. The Societes Nationales (SN), incorporated companies owned by the Government (or other public entities) to which all corporate laws apply unless otherwise stated by law or decree; SNs enjoy a large degree of autonomy, and are controlled by CFP. A second category, the Etablissements Publics (EP), includes administrative agencies (EPA), professional organizations (EPP) and wholly Government owned commercial or industrial companies (EPIC). Most EPs are submitted to the controls of CEP and to a lesser extent CFP, with the exception of a few major EPICs, which are controlled by Government Commissioners appointed and supervised by CFP. The remaining parapublic enterprises are mixed companies, Societes d'Economie Mixte (SEM), which are controlled by CFP if the Government has a majority stake in the company or if a special decree submits a company with minority Government ownership of at least 10% to CFP controls. 2/ 8. Law 77-89 gave the CFP the primary responsibility for the supervision of SEMs, SNs and selected EPs. The CFP powers, defined following the recommendations of the Bank Report, were broad. The law gave it the right to suspend any decision of the Board of a public enterprise for two weeks pending a decision by the President of thie Republic, and the right to conduct any investigation deemed necessary in a public ent rprise. 9. The President's Reportts interpretation of CFP's new responsibilities under law 77-89 was partly accurate and partly a 2/ The term public enterprise (PE) refers in this report to all SNs, all EPICs, and all SEMs in which the Government has a majority stake or that are submitted to CFP controls by decree.  - 21 - ANNEX 1 Page 4 of 8 reflection of IDA staff views. The table below compares the description of the law made in the Report with the law and decree 78-085 of February 1978 (taken in execution of law 77-89), which defines CFP's expanded responsibilities in the following terms: a) the control over the financial operations of the Government, the local governments and the public enterprises; and b) the financial information of the President and the Prime Minister on the management of public finances and of the entities placed under the Government's control. President's Report Law (L) / Decree (D) A. Strategic Analysis of PP Sector 1/ A. Financial Information of the President and Prime Minister on Management of PEs (D) (no mention of strategic analysis) including advising Government on: 1 - sub-sector policies 1 - no mention 2 - creation of new companies 2 - creation of new SEMs and capital increases (L art. 20) 3 - large capital expenditures 3 - profitability of investment programs prepared by SEMs and SNs (L art. 20) 4 - annual company budgets 4 - same, but control function B. Operational Supervision 1 - monitor implementation of objectives 1 - no reference given by Government to PEs 2 - anticipate problem area through 2 - no reference supervision of company management (permanent control over SNs and SEMs; L. art. 20) D. Formal Control ex post facto formal financial control formal financial and over SEMs and over two EPICs management control (ex post facto and in real time) of SEMs, SNs and selected EPICs (L. and D.)  - 22 - ANNEX 1 Page 5 of 8 1/ The DCA (Schedule 2, Part A (i)) describes this strategic role of CFP as the primary objective of this project component. "Development within CFP of an appropriate follow-up and control system to meet medium and long-term targets of the Parapublic Sector." The above table indicates that the Law or the ensuing Decree did not establish CFP as an agency responsible for overseeing or advising on economic aspects of the parapublic sector, but primarily on financial aspects. The decree lists the various responsibilities entrusted to the Financial Controller, which are all related to the central function of CFP, namely ensuring that public funds are spent properly (appropriately and legally), with the emphasis on the respect of the Senegalese legislation (including the National Plan). Furthermore, the Law (art. 19) refers to the Ministry of Finance as the agency responsible for the overall management of the Government's portfolio. III. Project Preparation and Appraisal 10. The project was identified in February 1977, but appraisal was delayed until law 77-89 was approved by the Cabinet and President Senghor. Appraisal took place in July 1977 and the parapublic reform law was voted by the National Assembly in August 1977. A. Initial Project Design 11. The project identification report (February 15, 3977) contains all the major elements of the future project, although the design of the CFP component was substantially different than the one proposed by the appraisal mission and included in law 77-89. In the identification report, CFP was to be divided into three functions: a) a study bureau, to be staffed by two Senegalese project analysts and a resident advisor-economist, to analyze the creation of new companies, review capital investments, do sector reviews and an annual report on the parapublic sector; b) an accounting department, to be staffed by five Senegalese and a resident advisor-accountant, to supervise the Government Commissioners in the EPs; c) a department of State Controllers, to be staffed by ten Senegalese and a resident advisor-management expert/financial analyst, who would represent CFP on a full-time basis on the boards of mixed companies . The structure the appraisal mission agreed on with CFP was different, however, and less adapted for the implementation of project objectives. It was decided to divide CFP's parapublic division in seven sections: a) two functional sections: documentation and data collection, and legal and accounting; and b) five operational sections divided by sector of activity (agriculture, agro-industries and fisheries; industry and mining; infrastructure and transport; tourism, communications and  - 23 - ANNEX 1 Page 6 of 8 services; and, financial institutions). No reason was given for the elimination of the study bureau. Not surprisingly, CFP did not analyze the rationale for the creation of new companies, nor did it review PE capital investments; sector reviews as well as annual parapublic reports have lacked serious economic or strategic analysis, and CFP's economic analysis capability at the end of project implementation was still very weak. Moreover, the absence of a study bureau reflects CFP management's low interest for strategic supervision. B. Project Preparation 12. Project preparation thus consisted mainly in i) defining the assistance required to allow the oversight agencies to take on the role they were assigned by the reform law, ii) developing a pilot program of audits and technical assistance to key enterprises to start dealing with the dismal situation of accounts and financial management, and iii) designing a training component for staff of the oversigl-t agencies, for public enterprise managers and accountants as well as tor government representatives on company boards. The project was processed ahead of the initial schedule. C. Major Issues 13. The Decision Meeting (August 1977): (i) recognized that Government commitment to the project was strong and warranted proceeding with the project 3/ ; (ii) requested that a) the executive decree organizing CFP in application of law 77-89 be discussed with the Bank before its final adoption, b) a high-level Senegalese acceptable to the Bank be appointed as parapublic division chief (this condition was later waived), and c) that Government propose at negotiations a program to recruit qualified Senegalese to staff the expanded CFP (no reference seen in minutes of negotiations);(iii) stressed the danger of weakening parapublic performance through repressive controls and excessive constraints;(iv) emphasized the need to shift the activities of the oversight agencies from control to monitoring and assistance and to define clearly what the agencies respective tasks are; (v) focussed on the importance of the project's training component; (vi) recognized the difficulty of measuring project success, which justified the inclusion of funds for annual evaluations by consultants (which were dropped later on); (vii) noted that sufficient guarantees were given for the further implementation of project objectives by the Government after project completion; and (viii) organized the split of responsibilities for project supervision. 3/ The July 1977 Issues Paper noted: "However, control of public enterprises and mixed companies is a delicate area, and despite the mission's belief that the project has been carefully conceived, the Government's commitment to a more (Footnote Continued)  - 24 - ANNEX 1 Page 7 of 8 14. This Credit was signed on March 17, 1978 and became the Bank's first Technical Assistance Project to the Parapublic Sector. This factor should be stressed; indeed, some of the project's shortcomings are due to its innovative character; there was no previous experience in the Bank with this sort of project to learn from. IV. Other Bank Projects in the Parapublic Sector 15. Bank involvement in the parapublic sector was further developed under other projects 4/ : a) The Technical Assistance Project for Economic and Financial Planning (Cr. 1061 of October 27, 1980) which included financing of a) an advisor in the Prime Minister's Office (November 1980-June 1983) to organize and participate in negotiations leading to the signature of contrats-plans between the Government and about ten public enterprises, b) overseas training for senior civil servants respons4ble for the implementation of these contrats-plans, and c) resident advisors at CFP, CEP and BOM (January-July 1983) previously financed by the first parapublic project; b) The Structural Adjustment Program (Ln. 1931/Cr. 1084 of January 26, 1981), under which the CFAF counterpart funds emanating from the currency payments for the SAL's import scheme were used for financing enterprises which had signed Bank-approved contrats-plans, mainly in the agriculture sector where the prospect of such financing certainly accelerated the agreement on contrats-plans; the use of these counterpart funds was poorly monitored, however. Experience to date shows that SALs are not good instruments to pursue gradual process-type reforms. c) The follow-up Second Parapublic Technical Assistance Project (Cr. 1398 of August 10, 1983) which continues the financing of the contrats-plans advisor (located at the Presidency after the abolition of the Premiership in May 1983), and of resident advisors at the various oversight agencies, previously financed by the First Parapublic Technical Assistance Project and the Technical Assistance Project for Economic and Financial Planning. The project's main objectives are to improve the efficiency of public enterprises and help reduce the public sector deficit by: (i) providing technical assistance to up to twelve key enterprises for the development and implementation of rehabilitation (Footnote Continued) rational and efficient parapublic sector may conceivably weaken in the future." It did. 4/ The Third Education Project (Cr. 908 of May 23, 1979) financed construction, equipment and technical assistance for ESGE, a graduate management training institute training managers for both private and public enterprises.  - 25 - ANNEX 1 Page 8 of 8 programs and contrats-plans; (ii) helping selected enterprises to finance audits and accounting assistance; (iii) developing and implementing training programs at local educational institutions geared to the needs of public enterprise personnel; (iv) strengthening the ability of Government oversight agencies to monitor parapublic sector performance efficiently and to guide Government policy; and (v) financing studies relating to the sc:pe, organization and management of the parapublic sector.  - 26 - ANNEX 2 Page 1 of 3 PROJECT MANAGEMENT I. Project Management Unit 1. The July 1977 Issues Paper recognized that since a large number of organizations were involved in the project, coordination should be a serious concern. The Government agreed to set up a project management unit (PMU) in the Prime Minister's office to coordinate project administration. The nomination of its director, who would also be chairman of the interagency task force on the parapublic sector (GESP), and of the deputy director in charge of day-to-day management was a condition of negotiations; both candidates were acceptable to IDA. The first director, who was acceptable to IDA, participated in project preparation but was named as head of a parapublic financial institution during negotiations in Washington. The second project director, also acceptable to IDA, was a lawyer who moved to the Supreme Court before effectiveness. The third director, unknown to IDA, was in place at effectiveness by which time his lack of familiarity with the project's objectives had become a serious problem. 2. A weak PMU and ineffective management remained one of the key problems of project implementation. The first supervision mission (July 1978) unsuccessfully requested that the deputy director be replaced. The next supervision mission (April 1979) found an undynamic project director as a major reason for the lack of coordination of the various project components and of the parapublic sector as a whole. The third supervision mission (December 1979) found large management problems. The PMU not only failed to provide policy guidance, coordinate project activities, submit progress reports, and keep any accounts, but it also had incurred a number of unauthorized, lavish and possibly fraudulent expenditures. The project director was found to serve no useful function and the deputy-director to be uncooperative. The mission recommended that if no improvements were made by March 1980, consideration should be given to suspending disbursements until the management had been replaced. Both the project director and the deputy director were replaced in February 1980. The new project director was characterized by the April 1980 mission as being more cooperative but showing little energy or initiative; the new deputy director "seems to be making no contribution at all to the project." The project director's performance later improved, but the deputy-director's remained unsatisfactory; the latter nonetheless became project director in November 1981. Hence, the various project components were managed indeapendently, without consideration for the overall objectives of the project and the reform law. 3. Many of the problems encountered by the project could have been overcome by a strong PMU director (Chairman of GESP). These problems included i) conflicts between the outside advisors and the  - 27 - ANNEX 2 Page 2 of 3 staff of CFP and DTAI, which paralyzed the execution of these project components; ii) the fragmentation of the government oversight system over the parapublic sector into ill-coordinated agencies whose powerbase the project consolidated; iii) conflicts between oversight agencies and BOM regarding the management of the training component; iv) a series of administrative irregularities and weaknesses, going from improper use of funds to excessive delays in processing of disbursements and poor accounting practices; and v) poor supervision, lack of regular evaluations of progress made toward the objectives of the project and the reform law. Poor performance under the first parapublic project led IDA to handpick the PMU director (Chairman of GESP) for the second project, a solution that also had its drawbacks. II. Selection of Consultants 5/ 4. In view of the critical importance of recruiting qualified experts for the eight resident advisor positions financed by the project, a special selection procedure was set up. Out of 25 firms contacted, eleven were invited to submit CVs. Out of a total of 140 CVs, five candidates were pre-selected for each position, most of whom were then interviewed in Europe or North America by a mixed Senegalese-IDA committee. The final selection was made following these interviews. An American consulting firm was awarded the CFP contract (four advisors), a Canadian accounting firm the CEP contract (two advisors) and the same Canadian consulting firm the BOM contract (two advisors). The firms were not asked to submit a proposal or to comment on the TOR. The outcome of this special selection procedure was not satisfactory. III. Counterpart Funding 5. Goverrment failed to provide counterpart funds for the project, causing important delays in project execution and accumulation of outstanding bills. By October 1980, the FY80 budget of CFAF 90 million, and the FY81 budget of CFAF 130 million had not been made available to the project. The problem was solved in early 1981, when CCCE allowed CFAF 130 million of its special loan to be allocated to the project's counterpart funds. IV. Project Evaluation 6. The April 1980 supervision mission recommended dropping the $100,000 allocated for project evaluation (three interim and one final evaluation) on the basis that is was difficult to find suitable consultants, that each beneficiary organization already prepared progress reports, that outside inquiry was costly and not we'lcome, and 5/ See Annex 5, par I for a description of the selection of two resident advisors at DTAI.  - 28 - ANNEX 2 Page 3 of 3 finally that IDA could do the job. Consequently section 3.08 b) of the DCA was waived. As a result, no systematic evaluation of project execution or of consultants' performance was made, project accounts were not audited, and the project completion report, which should have been prepared by the PNU with the help of consultants and submitted to the Bank by May 16, 1983, was not completed before the end of June 1984, after considerable pressure had been exerted on the last PMU director. V. Audit of Project Accounts 7. The inadequate PMU and its lack of concern for administration resulted in no project accounts being kept during the first years of the project; irregularities were noted in the use of the project's bank account; there was no follow-up of the disbursements on the IDA Credit; financing of some eligible expenditures was not requested; and requests were often allocated to the wrong category. The consultant recruited in November 1979 to do the project's supervision left a detailed note on these shortcomings with the Secretary General of the Government in 1980. 8. One of the experts of the CFP team was detached to the PMU from January to June 1982, with the purpose of writing up project accounts, both for IDA disbursements and counterpart funds. Accounting records were virtually non-existent, requiring that these be re-established for the three years since project inception. The expert left these records with the Project Director, but we were not able to review them. CVCCEP audited the counterpart funds of the project. Although no major legal irregularities were found, CVCCEP did note that the frequent turnover in PMU management and insufficient attention to financial and accounting matters resulted in deficient control over project execution and in the impossibility to properly assess project achievements. The Commission concluded in the necessity to establish for the follow-up project a PMU that would effectively "drive" the project. The main shortcomings identified by CVCCEP were: i) lack of procedures for financial management and accounting, which did not allow to keep track of commitments and disbursements, and partly accounted for the overcommitment on IDA funds, which left the government with more than $100,000 in unpaid bills that were still outstanding in October 1984; ii) inadequacy of supporting documents for expenditures; iii) no control over proper use of the nine vehicles purchased under the project; iv) violation of local legislation in recruitment and payment of PMU personnel; v) excessive cost of training seminars for PE managers; and vi) delays in availability of counterpart funds and in payment of bills, resulting in delays in deliveries of equipment and impairing proper project execution.  - 29 - ANNEX 3 Page 1 of 2 PROJECT SUPERVISION 1. The Decision Memorandum entrusted the primary responsibility for coordinating supervision of the project to the Programs Division, because of the multi-sector nature of the project and because of that division's large role in project preparation. Programs staff was expected to spend about six staff-weeks annually and the five projects divisions six staff-weeks. The time formally allocated to supervision of the project 6/ was around 12 staff-weeks a year. During the first months of project execution, the Programs Division became concerned about its technical and budgetary capacity to supervise the project adequately, especially as the appraisal mission leader was scheduled to become the Bank's first Resident Representative in Senegal and the decision was taken to de-emphasize his role in project supervision lest it interfere with his task of establishing a Resident Mission. Consequently in April 1979 the Transport Division, which had shown interest in the project and had sent a member on the appraisal mission, was asked to coordinate the supervision of the project. The Transport Division, however, did not have competent staff available on a regular basis to supervise a parapublic TA project, and a former agricultural division chief who was leaving the Bank at that time and who had participated in the appraisal mission, was hired as consultant to supervise the project twice a year starting in November 1979. This arrangement continued for about one year and was unfortunately terminated due to the unavailability of the consultant. His supervision missions and reports were thorough, but the follow-up by IDA on his recommendations was not always adequate and his contacts with headquarters were quasi-nonexistent. No supervision mission visited Dakar between October 1980 and November 1981. 2. The crucial importance of adequate supervision of this project was not recognized. Responsibility for supervision remained ambiguous 6/ Supervision time for the project was recorded by the TRS as follows (fiscal years, in staff-weeks) TOTAL 78 79 80 81 82 83 84 FY78-83 4.8 15.3 16.8 6.4 7.7 1.3 2.8 52.3 (PCR) Of the 52.3 staff-weeks recorded for FY78 through FY83, 5.0 were spent by the Programs Division, 13.8 by the Resident Mission, 28.7 by the Transport Division (including the consultant hired for supervision) and 4.8 by other Project Divisions.  - 30 - ANNEX 3 Page 2 of 2 throughout the project. The Resident Representative carried out an informal, de facto supervision, often through personal contacts with the Prime Minister or other high-level officials. Communications between the Resident Mission and Headquarters remained weak throughout the project. Clearly the Bank's internal organization was not adapted to handle projects crossing traditional divisional lines and no adjustments were made during project implementation to address this major problem. 3. The Senegalese complained about the high turnover of project supervisors and abGut the lack of IDA comments or feedback on audits and other project-related documents, which had a demoralizing effect. 7/ Projects divisions had agreed to budget staff resources for commenting on audit reports but failed to a large extent to do so. 4. Despite the serious problems that plagued the project throughout implementation, it was rated "status:2 / trend:1" by the five supervision missions held between April 1979 and October 1980, and "status:1 / trend:1" by the two last missions (December 1981 and June 1982). The project thus escaped the scrutiny of Regional Management during the full period of its implementation. 5. The uneven Bank supervision effort was a serious problem, especially since this was IDA's first Parapublic TA project, the project generated an enormous number of reports, a large number of resident advisors and short-term consultants had to be supervised, and major problems plagued the project throughout its execution. 7/ Mr. Alain Pichon, (expatriate) President of CVCCEP, addressed some harsh criticism in that regard to the Bank in a January 10, 1980 letter, of which an excerpt follows: "Je saisis cette occasion pour vous signifier mon mecontentement a legard des initiatives et des pratiques des experts de la Banque Mondiale qui excellent a provoquer des etudes de tout genre, mais negligent apres d'en suivre le contenu, le deroulement, les consequences, la coherence et l'application.... Je vous prierai de bien vouloir faire connaltre ma position a votre siege."  - 31 - ANNEX 4 Page 1 of 12 STRENGTHENING OF THE PARAPUBLIC OVERSIGHT SYSTEM I. Coordination of Oversight Agencies 1. The main objective of the Project was to support the reform of the parapublic sector oversight system introduced by law 77-89. A major benefit of the project in this area was the substantial improvement in the quality and availability of data on the parapublic sector made possible by the closer monitoring of these companies. This has allowed the Government to have a better idea of the enterprises' performance, and is also starting to provide the basis on which to build sounder sectoral and macroeconomic strategies. 2. Three shortcomings resulted, however. First, the project supported the oversight agencies on an individual basis, without clear delineation of each agency's terms of reference, with little coordination among them and with little integration of their interventions with the Technical Ministries, the Ministry of Finance or the Ministry of Planning. Specific project components were designed for each one of them, with the exception of training which was centralized at BOM. The weak PMU did not make a serious attempt to integrate the various components in a coherent whole. Various coordination instances were established: i) GESP, the interagency task force on the parapublic sector, ii) regular meetings of the agencies linked to the Prime Minister or Presidency, including CVCCEP; and iii) occasional (rare) meetings of the heads of the project components. However, this did not prevent the agency heads from trying to build up their own powerbase, often at the expense of overall efficiency. A lack of cooperation between ministers at the cabinet level amplified this problem. In January 1980 the Prime Minister suggested to move GESP from his own responsibility to that of the Finance Minister. This would have consolidated the responsibility for coordinating the Government's parapublic policy in the Ministry of Finance. IDA rejected this proposal, however, on the basis that it would increase friction between MEF and CFP and would lower the authority of GESP. The strengthening of the various oversight agencies achieved under the first parapublic project makes it now even more difficult to restructure the government oversight system around clearly defined tasks. 3. Second, the reform introduced by law 77-89 and supported by the project established a hybrid government oversight system over public enterprises-modulated on a legal classification of Etablissements Publics (EP) supervised by CEP on the one hand, and Societes d'Economie Mixte (SEM) and Societes Nationales (SN) supervised by CFP on the other--that does not rest on solid economic or managerial grounds but rather on historical ones. This system disperses scarce resources over different agencies when they could be better put to use in one agency.  - 32 - ANNEX 4 Page 2 of 12 4. Third, law 77-89 and the project focussed the major responsibility for control over the parapublic sector in CFP whose management was not committed to the reform and did not have to report to the Prime Minister, who was the key actor behind the reform. In fact, the Prime Minister's office did not ask for any support from the Bank until the end of 1980, when a Cellule des Contrats-Plans was established at the Primature with a half-time advisor detached from the CFP team and a director financed by IDA's Technical Assistance Project for Economic and Financial Planning. II. Assistance to the Controle Financier de la Presidence (CFP) 5. The main thrust of law 77-89 was to shift the locus of parapublic sector reform from MEF to CFP. The first parapublic project supported this shift by providing about 30% of disbursements for extensive technical assistance, training and equipment to CFP. This shift was motivated by MEF's unsatisfactory past performance and by the objective of building up at CFP capacity for strategic analysis of the entire parapublic sector, as well as for operational supervision and formal control of SNs, SEMs and selected EPs. However, no thorough assessment was made of CFP's willingness and ability to take this function over or of CFP's performance in controlling the central government and local government expenditures, which was its main function prior to 1978. During project implementation it became clear that CFP's main concern was with its formal control function, and to a lesser extent with operational control; strategic oversight of the parapublic sector was never a priority however. Consequently, at project completion (June 1982), CFP's strategic analysis capability was almost nil and has not much improved since. Operational supervision to be entrusted to CFP was defined by the President's Report as monitoring "to see if the objectives given by Government to parapublic companies are being achieved" and trying "to anticipate problem areas through supervision of company management". However, this is also the iask of the Ministries supervising the companies, although in practice their involvement in the companies has been closer to interference in the daily operations rather than ensuring Government policies were implemented or anticipating problems or bottlenecks. The proper role of Technical Ministries in the oversight of companies was not addre-sed by the project. Operational supervision as defined by the President's Report was not carried out by CFP, because of the weakness of th- state controllers. However, the capacity of CFP at formal control, both concomitant (in real time) and a posteriori, checking mainly if the companies respect the legislation and regulations in effect, did improve considerably during project execution. A. Technical Assistance Team 6. The technical assistance team was to include (President's Report, Annex IV, Appendix 1): i) a tean leader, management specialist/advisor to the parapublic sector division chief; ii) an agricultural economist; iii) an industrial sector specialist; and iv) a senior financial and accounting specialist. The mix/composition of the  ANNEX 4 Page 3 of 12 team and their TOR wer, well prepared and coherent, but did not relate to the i2aa!didre !IccUds ,I cF'P's recently created parapublic division. The first t.i';k to ht carried out by CFP was the apprehension/collection of key data on the 50 SE?is or SNs and 20 EPs they were to control. This was by itself a large task in view of the dismal state of most companies' accounting and reporting procedures, and of the lack of experience and poor background of most newly hired con.rollers. 7. The President's Report listed strategic control, operational supervision and formal control over PEs as the CFP's primary tasks. However, exerting formal control over the PEs was perceived by CFP as its first priority, to be carried out prier to the other functions. This explains why the presence of sectoral specialists was not deemed necessary and why a second financial analyst was finally hired instead of the industrial sector specialist. In retrospect, it would have been preferable to limit the TA to one or two experts in accounting, financial analysis and basic management information systems. The same situation still prevails today, i.e., two years after the closing of the first parapublic project and one year into implementation of the second parapublic project. Although the idea of having top-level sectoral specialist as ch. cks on the policies implemented by the sectoral ministries may be appealing, one still has to choose in light of scarce resources at CFP between reinforcing its financial and management potential or its sectoral and macroeconomic capability. Furthermore, locating such sectoral specialists in a central economic ministry (Plan, Finance) or in an executive unit at the Presidency might be more effective than locating them in a semi-autonomous control agency. 8. The consulting firm was selected among nine firms to pro-ide 12 man-years of assistance to CFP. The first three advisors arrived in October 1978. The first teamleader left in July 1979 and was not replaced until December 1979 and the last advisor (the second financial analyst) arrived in April 1980. The lack of unity and coherence among the TA team was a problem. Although the second teamleader tried to coordinate their efforts, the fact that two experts had already been in place for more than a year before he arrived and had thus established their working patterns and methods, the fact that none of the experts had even seen one another before arriving in Senegal, and that all four possessed widely differing backgrounds and qualifications, were bound to produce problems in this respect. 9. C nmunication problems between the TA team and CFP staff arose quickly. The team tried to implement western management techniques at CFP, with little attention paid to CFP's existing procedures. The introduction of a Management Information System, which the advisors referred to as MIS, the English acronym, was one example. Although CFP staff now recognize that the system was an improvement, they resented the sudden imposition of such a foreign management tool. The main conflict was one of management style: the technocratic, efficiency oriented approach of the TA team vs. the bureaucratic and legalistic approach which most controllers had been accustomed to. The team did not pay sufficient attention to the adaptation of their system to CFP's  - 34 - ANNEX 4 Page 4 of 12 needs and did not put sufficient emphasis on explaining the benefits to be expected from such system. Interviewed CFP staff complained about this lack of cultural awareness and said that after a few months the relations between staff and advisors had reached a level of near rebellion. 8/ The hostility towards the TA team was not general however; among the 26 state controllers, there were about 6 to 8 who appreciated the assistance of the team and benefited from the transfer of technology by working in close collaboration with them. 10. The climate at CFP improved however after the first year with the appointment of a new parapublic division chief and of a new TA teamleader, but tensions remained throughout the project. When IDA suggested to extend the team's mandate past December 1981, the Financial Controller declined and chose to keep only one financial analyst, whose relations with the state controlleLs was judged positive, less intrusive, more cooperative. This experience with resident advisors explains in part the current reluctance of the Senegalese to hire resident advisors under the Second Parapublic TA Project. B. CFP Management 11. Many of the problems encountered during project execution can be traced back to poor management at CFP, combined with management's 8/ In April 1979, 16 controllers addressed a note to the Financial Controller to complain about the TA team, stating that: - the experts were excluding the col'trollers from the decision-making process, and were signing notes t., the President themselves; - the experts refused to adopt the control mechanisms and procedures they had designed to the "realities of the parapublic sector" and to the Senegalese legislation; - this control system was ill-suited for financial companies and utilities; - PE managers could not understand that the Financial Controller was represented by experts from a consulting firm and made complaints about the experts' attitude; and requesting that: - the TA team be limited to an advisory function; - the follow-up system be modified; - meetings between the Financial Controller and the controllers be resumed; - the experts be prohibited from communicating directly with the President or Prime Minister; - the division chief be associated with all decisions concerning the division.  - 35 - ANNEX 4 Page 5 of 12 lack of dynamism and of commitment to parapublic sector reform. Some IDA staff and some members of the TA team locate the main problem with this project component at the level of the Financial Controller himself, and of the initial parapublic sector division chief at CFP. a) The Financial Controller 12. The Financial Controller was sick during the appraisal mission and discussions were carried on with the Prime Minister and his parapublic sector advisor. Shortly after effectiveness, the Financial Controller returned and it became clear to IDA that he did not have the motivation to make the CFP into the instrument envisaged by the then Prime Minister, and had no real interest in the project or in the reforms to be pursued under the project. Referring directly to the President, he was quite independent from pressures by the Prime Minister, who was the main driving force behind the reform. Indeed, CFP management was not able or willing to take on the strategic supervision of the parapublic companies and of the sector as a whole, preferring to limit its role to the less controversial function of formal control. b) The Parapublic Division Chief 13. The appraisal mission recognized that the CFP division chief for the parapublic sector needed to be replaced and made this a condition of negotiations. However, as all other conditions of negotiations and Board presentation had been fulfilled, this condition was waived with the Prime Minister's promise that the replacement would be made. IDA overestimated the Prime Minister's ability to influence CFP. When the division chief was finally fired late in 1979 (i.e., close to two years after negotiations), the Financial Controller agreed orally that he would observe the three top internal candidates for six months before choosing a division chief and this would be done in consultation with IDA. Instead, the FinancJ'al Controller promoted in November 1979 a state controller with no practical or managerial experience, but who was much more cooperative; this division chief left CFP and was replaced by another state controller in early 1981. This high turnover contributed to the slow progress made under this component. c) Personnel Management - State Controllers Recruitment of Controllers 14. The DCA stipulated that the Government had to recruit new staff for CFP acceptable to IDA and in accordance with a timetable acceptable to IDA. (Section 3.05(b)). The President's Report further specified that "the eight Senegalese staff in CFP dealing with the parapublic cector would be progressively increased to 26 by 1981 to carry out the CFP's new duties and as the technical assistance personnel are replaced by Senegalese... The recruitment of high-level Senegalese with experience in both administration and business management is  - 36 - ANNEX 4 Page 6 of 12 essential for the CFP to perform its strategic analysis, operational supervision, and formal control over the parapublic sector. Government discussed with the Association and will carry out a program to recruit and train qualified Senegalese to staff the expanded CFP...." (par. 34, President's Report). Finding and retaining suitable Senegalese for CFP was identified as a project risk. 15. Although the important responsibilities and tasks entrusted to CFP by the recent law 77-89 and reinforced by the project's covenants required to hire additional controllers, they did not justify the recruitment by January 1979 of fifteen additional state controllers, most of whom lacked the proper training and experience, and were hired more for their connections than for their qualifications. This sudden and large influx of new staff that was not directly operational was too large for CFP to absorb especially in the absence of a strong training program and given the strained relations between CFP staff and expatriate advisors. Absence of Career Incentives 16. Most controllers belong to a corps (French administrative system): corps des admiz1istrateurs civils, des inspecteurs des imp8ts/du tresor/des douanes, etc. While at CFP, they remain part of their corps d'origine and get promotions and salary increases within that corps. The Financial Controller evaluates their performance, but does not have any input in the decision to promote them or not, which is made by the corps. i7. On top of their regular salary (about 100,000 CFAF/month) they receive flat allowances of 25,000 CFAF/month for working at CFP and of 50,000 CFAF/msnth for their duties as state controllers in PEs. Despite a decree forbidding the concurrent drawing of these two allowances, all professional staff do receive them. 18. CFP has three divisions headed by division chiefs; the parapublic division is composed of five sectoral units, headed by unit chiefs. The functions of unit or division chief are not officially recognized; they do not come with a higher salary or allowance (the division chief actually makes less than quite a few of the state controllers), nor ao tI'ey give authority (hierarchical power, power to sanction, to recruit, f_re or promote) over the unit or division staff, or over the use of its budget. Only the Financial Controller, who is appointed by decree, has the hierarchical authority, the power to sanction and to evaluate. There is thus no real career track at CFP, and few individual incentives. Turnover 19. Sixteen state controllers (out of 22 working full-time on the parapublic sector) have left CFP between 1979 and 1984; of those, eleven have remained in the public or parapublic sector. Most staff joined CFP for training and experience that would allow them to get a more  ANNEX 4 Page 7 of 12 lucrative ioh a few- years later. Only half of the full-time state controllers have been with CFP for four years or more. Over the last fiscal year, ending June 1984, nine new controllers have been hired (and hence need intensive training). Such a high turnover is only partly explained by the low salaries and the bleak career prospects at CFP, however. Indeed, it is not uncommon in the auditing profession, which is often a springboard to management positions in enterprises. 20. Not only was staff turnover high, but state controllers' assignments to individual PEs also changed regularly. Most companies have been supervised by three, four or more different controllers over the time of the project. Assignments were made arbitrarily by the Financial Controller himself and were often the result of intense politickinlg as some companies offer greater opportunities to establish credentials and contacts for a future career outside CFP. Attitude of State Controllers 21. Managers of public enterprises complained, especially in the first years of the project, about the controller's inexperience and arrogance and felt they were often more an obstacle than a help. The work environment within CFP rapidly deteriorated. Besides conflicts b'etween part of the staff and the TA team ( cf. par. 9), important tensions also existed within the staff, due to a large extent to i) differences of pay status between civil service controllers and contractual controllers; not only were the latter often better paid, they were also not subject to the obligation to remain for 10 years in the administration; ii) resentment regarding the allocation of enterprises between controllers; and iii) resentment regarding the choice of section chiefs. d) Weak Logistics 22. Support Staff. Secretarial staff was inadequate throughout the project. Until 1980, only four secretaries were working at CFP, of which only two were available to the parapublic division. This represented one secretary for 12 controllers, not counting the technical assistants, which resulted in excessive delays in the typing and in further demotivation of controllers. IDA agreed to finance two secretaries and a messenger under the project for the use of the TA team. A -hough this may have been a necessary step to allow the team to operate, the Government should have been made responsible for the proper implementation of the credit agreement (section 3.05(a)) requiring adequate staffing of CFP. The parapublic division chief had no control over the support staff, who reported directly to the Financial Controller. 23. Equipment. Lack of budgets was never a problem for CFP which benefited from project funds and from a special treasury account (funded by compulsory contributions from the PEs to the operation of the oversight system, and under the direct control of the Financial Controller) to purchase equipment, cars, and to finance other operating  ANNEX 4 - 38 - Page 8 of 12 expenditures. Equipment and cars financed by the project were not specifically allocated to the parapublic division, but served the whole CFP. Controllers who needed cars for missions complained about the difficulty of obtaining them. 24. Buildings. CFP buildings were scattered over three or four different locations, increasing communication problems and proper allocation of support staff and equipment. All CFP services have recently moved to a new building (1984). C. Achievements under the Project Component a) Organization of CFP 25. The experts contributed significantly to the strengthening of CFP's parapublic division. They organized the state controller's work, produced a procedures manual which is still in use, although the state controllers initially refused to use it, and provided the state controllers with valuable guidance and training. Their efficiency, however, was limited due to causes which have already been highlighted. CFP is now better equipped to supervise parapublic companies, but has not been able to reach the objectives it had been assigned by law 77-89 and the project. The cargets set for CFP and the pace at which it expanded were too ambitious. The role of CFP will have to be reassessed within the overall framework of the parapublic sector oversight system, and its management will have to be dramatically improved if CFP is to establish a significant role in the parapublic sector. b) Development of a Comprehensive Deta Base 26. The development of a comprehensive data base covering the EPs, the SNs and the SEMs controlled by CFP was identified as one of the objectives and one of the measurable results to be expected from the project. By project completion such data base had been established and computerized in cooperation with DTAI, but included only financial data (balance sheets and operating accounts), which are now more reliable and more timely. Production of these data remains too slow, however, limiting their operational usefulness; consolidated financial statements of the companies monitored by CFP for 1983, for example, were not available before December 1984. Non-financial data on the other hand are not kept in a systematic way; it is difficult to get data on employment or productivity for example. This results from the priority given by CFP to formal financial control and reflects its limitations in exerting operational supervision or strategic analysis/planning. IDA has now asked CFP to produce regular financial statements on the parapublic sector for the use of -he Government and to expand its database to cover production, employment, productivity and other key data needed to monitor the sector.  -39 - ANNEX 4 Page 9 of 12 c) Company Supervision. 27. The quality of company supervision is hard to assess, as it mostly involves routine operations like participating in board meetings and making observations on the documents (financial statements, investment programs, etc.) submitted to the board, visiting company headquarters to gather information or investigate specific issues. From the documents we were able to review, this supervision remains too formal, focussing on the respect of the accounting legislation or of the numerous decrees regulating various aspects of company management. The trivial, such as the size of the car a company manager is allowed to have, often dominates more fundamental issues, such as the appropriateness of a given subsidy level, the review of the investment programs and their proposed financing, critical assessment of company management, etc. 28. Reports tend to be descriptive; operational recommendations are too rare, but the quality of the descriptive part and the analysis of company data have improved considerably. Also, all the companies with majority government ownership are now being monitored on a regular basis, which was not at all the case in 1978, when a number of companies didn't even prepare annual budgets. It is difficult to assess to what extent this improved data gathering and control system has been used as an input in the political decision-making process. d) Sectoral Reports 29. Annual reports have been addressed to the President starting in 1977 covering the various economic sectors in which the State is involved. The draft and publication of these reports take an enormous effort, immobilizing the entire staff for months. They are extremely long, of varying quality and read in their entirety by almost no one. They do not lead to real decisions by the Government. These observations which were already made by IDA in January 1979 on the 1977 and 1978 reports remain true today; we had agreed with the Financial Controller that future reports would focus on specific and immediate company or sectoral problems, be published throughout the year and be complemented by an annual macroeconomic and parapublic sector-wide report. Similar observations and suggestions had to be made in June 1984 concerning the 1983 sectoral reports; the draft of the 1983 reports was completed in July 1983, but they were not issued before May 1984; they consist of 9 volumes (over 900 pages total) with sections on the 70 companies supervised by CFP (and CEP); they do not include sector reports nor an overall macroeconomic analysis. 30. A sector-wide parapublic report was prepared and published in 1982; it commented on the aggregate financial statements of the parapublic sector for 1978, 1979 and 1980, but did not provide insights on the ways to solve the sector's acute financial and managerial problems, which tend to be highlighted in every report on the parapublic sector in Senegal. The strategic dimension of this document is almost inexistent; the proper role of Government in the various sub-sectors is  - 40 - ANNEX 4 Page 10 of 12 not addressed. Furthermore, its publication in 19F2 made comments on financial statements of 1978 to 1980 more of historical rather than operational interest. 31. To get the CFP to draft timely, operational, focussed documents specific instructions to that effect should have been given by the Presidency and the Ministry of Finance (the primary potential users of such information). As for the sector-wide macroeconomic reports, these should be drafted by senior officials from Plan, Finance and the Presidency, using company data provided by CFP. e) New Management Techniques 32. Two major management tools were introduced during this project with the assistance of the TA team, namely a new budgeting format based on cost/profit centers, and an enterprise plan negotiated with the government (contrat-plan). The new budgeting procedures have now been adopted by a majority of public enterprises and have helped management to focus on its objectives for the fiscal year and on the means to achieve these. A more detailed audit would be needed to assess progress made under the project in company budgeting and accounting. ContrEts-plans were signed with five companies in 1981 and 1982. The contrats-plans procedure has been developed and promoted under four IDA projects (supra, Annex 1). The first parapublic ptoject financed two missions of Frenclh contrats-plans experts to develop the methodology and detached one member of the Berger team on a half-time basis to the contrats-plans unit at the Prime Minister's office 9/ between October 1980 and July 1981. D. Conclusion 33. Although part of the problems encountered at CFP could have been anticipated and taken into account at the appraisal stage, the major weakness on IDA's part was at the supervision stage. In light of CFP's lack of commitment to take on the responsibility it had been assigned by the project agreement and of the tensions existing between advisors and CFP, as well as between CFP staff and between the TA team, the assistance to CFP should have been reassessed. However, the project did contribute to the improvement in financial reporting by PEs and to the establishment of a financial data base on the parapublic sector. 9/ The unit Chief's salary was financed under the Second Technical Assistance Project for Economic and Financial Planning (Cr. 1061-SE).  - 41 - ANNEX 4 .age 11 of 12 III. Assistance to tll a.entre des Ecablissements Publics 34. The Centre des Etablissements Publics (CEP), established in 1966, has two main divisions: the Agence Comptable Centrale (ACC), which centralizes company accounts, supervises accountants and manages cash flow through a central treasury pooling system, and the Controle des Operations Financieres (COF), which authorizes expenditures (a priori control), reviews budgets and provides occasional management assistance, for about twenty Etablissements Publics (EP). 35. Four EPs have been emancipated from CEP controls, including SAED which became a SN in 1981. The others are RCIS (since 1976), SODESP and SONAR (since their creation), where a Government Commissioner appointed and supervised by the CFP is entrusted with the responsibility CEP has over the other EPs. The not; defunct ONCAD was submitted to the same control mechanism; this larger autonomy, however, did not result in improved performance, as all these companies continued to be plagued by increasing indebtedness and deficits and were in need of major restructuring. The failure of this decentralization of controls is to a large extent due to the lack of accompanying measures and to the severe structural and financial problems faced by these companies, which were not properly addressed. 36. A primary objective of the assistance provided to CEP was to gradually shift the function of this agency from control over EPs to assistance to EPs. The project financed i) a management expert for three years (initially two) to set up a management assistance unit, review EP management systems and procedures, scale down and improve COF controls; ii) an accounting expert for three years (initially two, and later extended for another three years under the follow-up parapublic project) to improve accounting systems and procedures in EPs; and iii) training for CEP staff, EP management and personnel (cf. Annex 7). 37. Existing a priori controls and the treasury pooling mechanism were seen as important constraints on the operations of most EPs. However, removing these constraints at once was seen as not feasible, in light of the weak management and accounting systems in place in most EPs. The centralized controls and treasury pooling were imposed by law in 1966 in exchange for the take-over by the Government of the EP's consolidated debt, and were intended to put an end to the mismanagement and abuses +hat characterized EPs up to 1966. It was felt, however, that by str ngthening EP management through the experts' assistance, EPs could be made gradually more autonomous. One of the tasks assigned to the CEP experts was to look at ways to scale these controls down and to recommend ways to grant EPs the appropriate level of autonomy. 38. The advisors' contributions were considered as very valuable. They were instrumental in upgrading the quality of EP management and accounting procedures. With help of the experts, CEP achieved the following results during the project:  - 42 - ANNEX 4 Page 12 of 12 - the establishment of internal audit units in EPs, considered as a preliminary for the phase-out of a priori controls (eight EPs had established such units by the end of the project); - implementation in EPs of the January 1980 directive of the Prime Minister to set up a decentralized budgeting system based on cost centers (management by objectives); by project completion, 80X of the EPs had complied; - submission of more timely and more realistic budgets; - introduction of quarterly reports on status of budget execution (80% compliance by project completion); - centralized computerization of EP accounts in cooperation with DTAI; - reduction by six months in the submission of the annual financial statements, combined with an increased reliability of such statements, which has lead to better and more timely information on the EPs; establishment of new accounting systems and procedures, ana of limited management information systems; improvement in the accountability of the Agents Comptables Particuliers (enterprise accountants appointed by the Minister of Finance and subordinated to the Agent Comptable Central), through training and assistance; speciic programs, such as the organization of teams of interns to track down and collect client receivables at OHLM (housing agency) and PAD (port) or the preparation of reorganization proposals for MSAD (tapestry manufacture) and ITA (food research institute). 39. The progress achieved in EP management and financial systems should have made the phase-out of a priori controls and (to a lesser extent) of central treasury pooling possible, as was expected at project appraisal. The former director of COF had concluded to the necessity of such phase-out but could not organize it due to his departure at the end of 1980. His successor has reached the same conclusion, but has not yet made specific proposals. Both feel, and we concur, that by phasing out these cumbersome controls, companies will recover the degree of autonomy that is necessary to successfully manage a public enterprise, and CEP will be able to redeploy staff resources from unproductive controls, which currently take up more than 75% of their time, to management assistance. However, some EPs, in particular the non-commercial ones, would still remain submitted to stricter controls.  ANNEX 5 Page 1 of 2 TECHNICAL ASSISTANCE TO THE DTAI 1. The technical assistance component to DTAI, the Finance Ministry's data processing department, was added to the project during project execution. The December 1979 supervision mission recommended to use project funds to finance a study of the EP computer systems and programming: "The main objective should be to eliminate the hand-written cash journals and have the journals printed on the spot from cards punched from the basic documents, to prepare auxiliary ledgers (for fixed assets, stocks receivables, payables, ctc.) and to analyze expenditures by cost center. It will probably be more economical and practical to instal mini-computers in each of the large EPs and a central installation for the smaller ones, but this should be studied." The mission asked the project director and the consulting firm that was awarded the BOM contract to prepare TOR. No other firms were contacted. The 'irm drafted a contract for the provision of two computer experts to train DTAI personnel in the use of a new methodology to develop computer systems; the assistance would cover training of DTAI staff in this methodology and applications of the new techniques in one or more pilot projects. The consultant's mission was not further described in the contract, and no TOR were prepared. The firm's priposal, however, did not address the problem that had been identified, but dealt mainly with the training of DTAI staff in general. The firm submitted four candidates for the position, of whom one did not meet the basic recruitment conditions. Ar. IDA - Government selection committee ranked one of the three remaining candidates as very good and the two others as "interesting". 2. Although the competence of the experts was not contested, the assistance provided by the firm to DTAI was not satisfactory. It took five months (10 man-months) for the experts to agree with DTAI on the methodology and techniques to be used, and to launch the preparation of the first pilot project. Substantive delays in start-up were due to the lack of preparation of the mission by the firm, and to the firm's failure to deliver the required training manuals in a timely fashion. When the manuals finally arrived from Montreal they turned out to be ill-suited for DTAI's needs. The back-up from the firm's headquarters was totally insufficient. 3. Three seminars were organized on the first three steps (out of six) of the computer system development methodology proposed by the firm (Yourdon method). Most seminars were judged by DTAI staff to be inadequate and ill-prepared. No manuals were provided at the seminars; draft manuals were submitted, but only one out of three was finalized and approved by DTAI. The proposed method has not been successfully introduced at DTAI, which is now considering to ask a French institute to help them introduce another method. It was envisaged to conduct up to six pilot-projects. Only two were started and neither one was completed. One of the two pilot projects initiated by the experts, but  - 44 - ANNEX 5 Page 2 of 2 never completed, dealt with computerization of OPT's billing and outstanding debts; the second one was a car registration project for the Transport Ministry, which had no connection at all with the parapublic sector. OPT, however, plans to use the work done on telephone billing and receivables as an input in its computerization program, for which IDA assistance has been requested under the follow-up parapublic project. The firms's assistance was most appreciated by the OPT computer specialist who collaborated closely with the experts in the development of the pilot project. The DTAI director, however, expressed his extreme dissatisfaction with the firm's performance to the appraisal mission for the second parapublic project. 4. The absence of TOR, the different views held by DTAI and the firm regarding the objectives and the execution of the mission led to repeated conflicts. The project component suffered from a dual leadership structure and lack of communication: the director of DTAI, on the one hand, who wanted to keep full control over the implementation of the component, including the pilot projects; the consultants, on the other, who felt they needed direct authority over their Senegalese counterparts in order to be able to execute their mission. Effective leadership at the PMU or supervision by the Bank would most likely have contributed to the solution of these problems. 5. The main problems with this component can be traced to: i) inadequate preparation and appraisal of this component; poor identification of DTAI needs in parapublic sector work, poor design of project component; ii) unclear relation to overall project objectives; iii) absence of TOR for the consultants; no evaluation of the firm's proposed work program; no other firm contacted; iv) inability of the consulting firm to provide required back-up from headquarters to consultants in the field; the firm did not have the proper manuals nor the appropriate staff for the training side of the mandate; v) limited availability of DTAI counterpart staff; vi) excessive centralization of project management (including of pilot project) at level of DTAI director; vii) failure of DTAI and of PMU to flag problems in a timely manner; viii) tensions between the TA team and experts from another firm, which had been assisting DTAI on a long-standing basis; and ix) lack of IDA supervision of this component (the only supervision mission that came to Dakar during the 18 months of the experts' presence, apparently did not cover this component; an IDA mission evaluated the component in June 1982). 6. The major problem identified prior to the inception of this component remains as important in 1984 as it was in 1980: what are the precise data processing needs of the EPs and is DTAI the proper locus/institution to handle these computer needs? The question was not addressed under the project.  - 45 - ANNEX 6 Page 1 of 5 AUDITS OF PUBLIC ENTERPRISES 1. Two audit programs were included in the project, covering in each case three different aspects: i) audits of financial statements, with possible certification; ii) review and strengthening of accounting procedures; and iii) analysis of economic performance. They were financed for 902 from credit funds and for 10% by the Government; the companies did not contribute to the costs of the audits. The objective was to improve the financial and accounting systems and procedures of these PEs in order to provide management and Government with appropriate company level data to base their policy decisions on. The audit component was managed by CVCCEP 10/ , the Senegalese public enterprise auditing agency under the Supreme Court, and fully integrated in the agency's work program. I. Auditing, Financial and Accounting Assistance to OPT, SAED, SOMIVAC, and CPSP 2. Audits plus technical assistance packages were designed as a pilot program to be replicated in other PEs if successful. The program consisted of annual audits, review and strengthening of accounting procedures and training of accountants and consulting services over a three-year period. The dismal state of the ONCAD accounts, as revealed by a Arthur Andersen report, did not allow to audit them and ONCAD was thus dropped from the audit program. In 1980, the Government was forced to liquidate ONCAD and absorb accumulated losses of close to CFAF 100 bil'lion (roughly 15Z of Senegal's 1980 GDP). It was later decided to audit CPSP accounts instead. The accounts of SAED, SOMIVAC, and OPT could not be certified, but auditors made important recommendations. No meaningful analysis of economic performance could be made for these companies, as the necessary data were either not available or not reliable. A. SAED Audit and Follow-up TA 3. SAED accounts were thoroughly audited. The second year audit revealed ti-at no progress had been made in the quality of the accounting 10/ CVCCEP was the only oversight agency involved in the project that received no technical assistance or equipment from the project. Its staff, however, benefited from in-depth training both in Senegal and overseas, which contributed to further Senegalisation of CVCCEP staff. Moreover, about half of its staff (including its President) were expatriates financed through bilateral aid.  - 46 - ANNEX 6 Page 2 of 5 since the previous year, which was not surprising as little assistance had been given to SAED's accounting department over that period. It was then decided to limit the third year audit to a few key functions and to postpone the third full audit to year four. Relations between auditors and SAED were strained. Little effort was made by management to correct the major irregularities that had been identified by the auditors. 4. Meanwhile, following recommendations made by the auditors, the Government requested and obtained financing for a management support unit consisting of three senior experts attached to the General Manager to be financed jointly by CCCE/FAC and the Parapublic Project (up to $200,000). The assistance yielded little benefit. The TA team was not well integrated into SAED's operations. It would be worthwhile to assess the quality and effects of the assistance provided by IDA and other donors to this company over the last five to ten years. 5. In September 1981 SAED was transformed into a SN against the recommendation of the oversight agencies, which argued that the poor management of this company, as revealed by the auditors, did not warrant the large autonomy given to SNs. The company remains financially and managerially weak; its deficits have soared in recent years. B. SOMIVAC Audit and Follow-up TA 6. The first two phases of the audit of SOMIVJAC, a major rural development agency, were completed by January 1980. They revealed the need to provide TA to strengthen the company's accounts and procedures and to help implement the auditors' recommendations. The second audit report noted that the accountant had successfully started to implement the recommendations made in the first report, but that in other areas company performance had not been improved or had even worsened. In parallel, IDA approved the financing under the Parapublic Project of experts (four months) and of an accountant (two years) to help implement the procedures manual prepared by the auditors. SOMIVAC management, however, showed no interest in either program and refused the assistance. The audit and TA program was thus stopped after the second audit phase. Since then, SOMIVAC has continued to grow and its situation to worsen; most donors, including the Bank, have phased out their assistance to SOMIVAC, which is now heavily dependent upon the national budget. C. OPT Audit and Follow-up TA 7. The association of an international and a Senegalese firm that was awarded this contract was the only one among the seven contracts awarded under the project not to work out, mainly due to conflicts between the two firms resulting in poor audit work. After one year, CVCCEP cancelled the contract and a new association was hired to audit FY79 and FY80 accounts. 8. The audits/assistance resulted in some improvement of the financial documents, a greater attention to the way money was spent, the  - 47 - ANNEX 6 Page 3 of 5 implementation of somewhat more performing financial and management systems, and in general better informa-ion about the cimpany's performance. 9. The audits identified the need to grant OPT larger operating autonomy and laid the groundwork for the change in legal status which is being pursued under the second project. It provided better information based on which tariff increases could be requested, leading in turn to a better balanced budget. It stressed the danger of growing personnel expenditures and overstaffing and led to a freeze in recruitment. It revealed some irregularities in OPT procurement procedures and highlighted the catastrophic international mandate situation; action was subsequently taken to start correcting these situations. 10. In parallel a $500,000 TA program, of which $200,000 was financed by the Credit, was set up to strengthen OPT's accounting and financial management and train OPT staff. Two experts from a Canadian consulting firm were hired for two years (July 1979 - June 1981). OPT retained their services until June 1984 paying them directly out of its operating budget. They were highly valued by OPT management. The accounting expert helped improve OPT accounts, established a system of cost accounting, drafted a procedures manual, and trained the company's accountants. However, the splitting of the post and telecommunications accounts took longer than expected and was onlv introduced in the FY83 accounts; also, additors have so far not been able to certify the accounts. The management expert prepared procedures manuals for inventory management anu procurement and helped set up an internal management audit unit. Both experts contributed to the implementation of a system of management by objectives. 11. Financial statements are now issued within six months (compared to delays of up to three years previously), and OPT accounts might now be "certifiable". A new audit cycle was initiated under the follo'-up project in 1984. The combination of audits and technical assistance has contributed to the improvement of OPT's management, which should now allow the company to obtain more autonomy through its transformation into a SN, which was recommended by a recent (1984) diagnosis report p-Rpared by an IDA-financed consultant. D. CPSP Audit 12. The financing of an audit of CPSP, the price stabilization agency for major commodities, was included in the project in November 1981 at the Bank's and the IMF's request. The project financed CFAF 100 million (about US$370,000) of the cost of this audit. Almost no reference to this project component was found in the files. The company and Government failed to a large extent to implement the auditors' recommendations, despite President Diouf's instructions to that effect.  - 48 - ANNEX 6 Page 4 of 5 II. Audits of three major SEMs 13. The project financed one major audit and one follow-up audit of the accounts of three major SEMs: SICAP (housing), SONACOS (oil mill) and USB (banking). The SONACOS audit resulted in the first certification by auditors of the accounts of a public enterprise in Senegal. The USB and SICAP audits identified major weaknesses in company management and finances and made recommendations to correct these, most of which were not implemented; both companies are still managerially and financially weak. These were followed by more limited annual audits carried out by commissaires aux comptes financed by the companies themselves. In the past CVCCEP had had major diffic .lties in convincing such companies to hire competent auditors on a regular basis; the follow-up project includes a component to help draft legislation on mandatory auditing of PE annual accounts. III. Evaluation of the Audit Programs A. Adjustments in Auditors' TOR 14. These various audits illustrated, to an extent that had not been expected by IDA or the Government, the poor quality of PE accounts. During execution of the project it appeared necessary to amend auditors' TOR, when it was still possible, in order to: (i) help companies build up their accounting systems prior to carrying out full audits; (ii) space the pace of full audits in order to allow the company to implement auditors' recommendations prior to the next full audit; (iii) adapt the nature and level of recommendations to the company's implementation capacity; and (iv) provide TA to implement the recommendations made by the auditors. B. Benefits of the Project Component 15. Beyond the direct benefit of a greater reliability of company accounts they provided a greater awareness i) of the weakness of the existing accounting systems, ii) of the precise situation and problems of the audited companies, and iii) of the usefulness of audits, which hitherto had not been widespread in Senegal. They allowed to build up the Senegalese accounting profession through: i) the establishment of additional local firms, ii) joint ventures between such firms and established international firms, iii) return of Senegalese auditors working abroad, lured back by an increased demand of top-level Senegalese auditors, and iv) general improvement in quality of local auditing. The award of au.dit contracts to joint ventures of national and international firms has been a very positive mechanism for the transfer of know-how, and has been continued under the follow-up project. C. A Pilot Program 16. The audit component under the project was designed as a pilot projact to be replicated if successful. The audits led other parapublic  - 49 - ANNEX 6 Page 5 of 5 companies not included in the project to request the inclusion of their company in iuture J p;ograms and in some cases to finance their own limited audits in order to establish data on which management could assess performance and base its policy decisions. The experience gained under the first parapublic project has led to improvements in the design of audit programs under the follow-up project. With one exception, it appeared that none of the audited companies had accounts worth auditing and that what they needed first and foremost was accounting assistance. The type of 'chnical assistance to be included in the auditor's TOR covers: 1) the review and strengthening of accounting systems and procedures; ii) the establishment of an internal audIt unit within the company and related training; and iii) the monitoring of and assistance in the implementation of the recommendations made by the auditors. D. Shortcomings in Follow-up of Audits 17. The follow-up of the implementation of audit recommendations by the Bank, the companies and the Government was weak. The audit of SOMIVAC and SICAP for example should have led to a phase-out or downscaling of these companies. USB, as well as the other companies, should have been submitted to performance criteria to monitor improvement in its management. Only OPT benefited from a relatively successful follow-up program, which it financed itself to a large extent; the OPT audit and follow-up TA contributed to the currently ongoing sector restructuring (1984). Such comprehensive and longer term perspective is necessary tc optimize the use of these costly audits (IDA financing of 902 of cost reached US$2.0 million total vs. US$1.6 million originally allocated, ranging from US$170,000 to over US$500,000 per company). 18. The implementation of audit recommendations requires in many cases some external support, as the changes recommended Ere often managerially or politically difficult. SICAP or USB were not included in Bank operations and failed to a large extent to implement the recommended changes. The Bank has been repeatedly and rightly criticized by the Sexiegalese, espec..lly CVCCEP, for failing to provide comments on audit reports and to follow-up on the implementation of the recommendations. This was the case for companies not included in the Bank's project portfolio, but also, though to a lesser extent, for companies of direct interest to the Bank's lending program. Follow-up of the implementation of audit recommendations by the various oversight agencies ha; left much to be desired; only CVCCEP took the matter seriously by organizing plenary sessions in which the audit reports were discussed and formal recommendations addressed to company management. CVCCEP, however, did not have sufficient staff to follow-up on the implementation of these recommendations.  - 50 - ANNEX 7 Page 1 of 5 TRAINING COMPONENT I. Objectives of Training 1. The success of the project depended on improving management and accounting skills i) within the Government's control agencies concerned by the project, ii) at the level of company directors, and iii) in the finance/accounting departments of parapublic sector companies. The target groups were well defined as staff of CFP, CEP and CVCCEP, key members of boards of directors of selected enterprises, and finance and accounting officers of public enterprises, with particular attention to the staff of the EPs to be audited under the project. II. Training Design 2. Because technical assistants who are asked to do both operational work and training of counterparts often neglect the latter, it was decided to separate training from operational technical assistance to the project agencies. The means to achieve the project's training objectives were: (a) 6 man/years of specialist services in modern management techniques and decision-making; (b) funds to expand the instructional resources of the Bureau Organisation et Methodes (BOM) to meet the tasks imposed by the project; and (c) funds for short-term overseas scholarships for training in specialized fields ($650,000 of IDA credit for (a) and (b), of which only 30% were actually spent). The specialists were to work with the staff of the BOM to: (a) develop courses and case studies reflecting the types and level of management problems faced LJ managers of public enterprises in Senegal; (b) collaborate closely with the agencies and companies concerned to develop course materials which should comprise reusable packages of printed and audiovisual materials, together with tests to allow trainees' acquisition of new knowledge and skills to be measured. III. Training Component Management 3. Management of the training component was entrusted to the BOM at the Presidency, the central agency assigned the tasks of improving administrative structures, systems and procedures and development of staff skills in the public and parapublic sector. The BOM created a Training Division to implement the training activities under the project. IV. Implementation 4. A Canadian consulting firm was selected by the Government to provide two training specialists. The content and scope of their tasks were defined in the President's Report (Annex IV; Appendix 3), but were  - 51 - ANNEX 7 Page 2 of 5 not reflected in the contract. The firm was also asked to coordinate and administer the training of Senegalese officials selected under the project for scholarships in Canada. Two experts arrived 3 months after the date of project effectiveness and concentrated on planning short seminars on varied subjects. They also agreed on the need for intensive two-week seminars, particularly for financial controllers, and on a program in accounting training for enterprise accountants to be organized by the Institut Universitaire de Technologie (IUT) at the University of Dakar. For the intensive seminars, expatriate trainers (from France) had to be invited to run the different sessions. There was a high turn-over of experts at BOM (6 persons in 3 years), leading to different perceptions of the experts' role. During project execution the emphasis moved away from training; the experts did very little training, did not develop adequate training materials and concentrated instead on organizing seminars. A. Short-Term Seminars 5. The BOM organized 30 seminars each of 2-5 days duration and covering 21 subjects. Some had to be repeated three times to allow for a larger audience and to compensate for participants' frequent absenteeism. In view of the limited time available to senior staff, particularly public enterprise board members, the seminars were orgarized on the basis of 2 afternoons per week for 3-4 weeks. Approximately 300 people attended, of whom many attended more than one seminar. B. Intensive Seminars 6. Six intensive seminars of 10-15 days were organized during the project. One of these seminars organized by BOM was for training the Directors General of public enterprises. The Audit Commission (CVCCEP) organized its own intensive seminar during two weeks for CVCCEP auditors and CFP controllers with the assistance of an international auditing firm which was also selected to organize a practical training program for 7 auditors at its Paris office. CFP also organized its own seminar; they sought bids from three firms and one university and chose the university; the seminar was organized in two parts, one in 1980, the other in 1981. 7. Another type of intensive training was offered only during the last year of the project by the IUT. It took a very long time for the BOM, IUT and the Public Contracts Commission (ChNA) to agree on two programs, their content and cost. During the first program of 9 months, 25 public enterprise accounting clerks were trained (academic training) and received a basic degree in accounting. The second program, also for 9 months of academic training in accounting, led to a higher degree in accounting.  - 52 - ANNEX 7 Page 3 of 5 C. Training Materials 8. No significant training materials were developed. Very few and brief case studies of little relevance were developed. These were western cases using Senegalese names. No reusable printed or audiovisual materials were prepared as had been planned in the training objectives. D. Overseas Training 9. CFP sent 12 persons (50% of its staff) for an 8 week training program either in France or Canada. The consulting firms in France and Canada received the CFP trainees to train them in theory and practice of auditing. A few attended short courses at a school of public administration i!: Canada and went through a study tour to familiarize themselves with government public finance operations. Of the 12 trainees, only 5 have remained at CFP, while the others transferred to other jobs. 10. Seven auditors from CVCCEP received training at the Paris office of an international accounting firm. While the other firms charged the project for their training services, this firm did not; however, it was adequarcl, cczipensated by the cost of the one-week seminar it organized in Dakar. Of the seven auditors given training in auditing techniques, only 3 have remained at CVCCEP; two of those who left became directors of CEP. Some CEP staff members also benefited from overseas training. V. The Training Component in Perspective 11. It is difficult to assess the success of the training component. Among the achievements one should mention that six of the trainees now occupy leading positions in Government agencies and public enterprises; undoubtedly their training contributed to their promotion. However, many problems arose: a) There was an excessive turnover in the TA team of 2 experts, which resulted in a lack of continuity in the implementation of the project component. b) The progress reports prepared by the training experts did not reveal the implementation problems. There was little depth in the training; the background of seminar attendants was too heterogeneous; many trainees were unmotivated and absenteeism was significant. Letters were sent to the employers of absentees, but were ignored. c) The intensive seminars fared better because the duration allowed some in-depth treatment of the subjects. The expatriates from two organizations came well prepared with their training materials, while other experts came ill-prepared.  - 53 - ANNEX 7 Page 4 of 5 d) Management of the component was conflict-laden because: (i) the authority of the BOM vis-a-vis the heads of the other components had not been made clear; (ii) the expatriates assisting the different project agencies decided to organize the local training of their agency staff in their own way, collaboration between them and the BOM experts was difficult to achieve since each claimed primary responsibility for training implementation. e) Overseas training was beneficial as means of broadening the horizons of trainees but was not not always very relevant. Trainees complained that the coordinating firm and a training firm did not care much and neglected their needs. They reported gains in general knowledge, but not in skills needed to improve their performance. There was no analysis of their tasks at home, nor appropriate planning of the content of their training overseas to meet the needs for improved performance. f) Experts at the BOM were also asked to do consulting tasks, deviating their attention from their primary tasks. Although the tasks sometimes benefited training, these benefits were not demonstrated through the preparation of case materials. g) Supervision Reports repeatedly stressed BOM management's lack of initiative and drive. h) There is no evidence of guidance from IDA supervision missions to contro' and improve the quality of training. i) The accounts of the training components were not kept at the BOM and it has been very difficult to obtain complete data showing the costs of local training. VI. Conclusion 12. The training component was well prepared, but too ambitious. Its implementation left much to be desired. At the outset of project implementation, agencies financed by the project challenged the authority of the BOM for controlling their training activities. This impaired the effective implementation of the component. Agreement on the allocation of responsibility and commensurate authority should have been reached among all parties concerned before the start of implementation. IDA staff should have played the role of catalist or facilitator in this matter. Care should have been taken in approving the contract with the consulting firm for spelling out the objectives of the component and the specific tasks needed to achieve them. The component was not adequately supervised by IDA. This led to the deviation of the consultants from their training role to an advisory and consulting role. Had supervision been adequate, the design and content  - 54 - ANNEX 7 Page 5 of 5 of the courses would have been improved and corrective measures for better management of the component would have been possible. 13. Corrective measures were not taken to address the problems raised in the reports that trainees submitted upon their return. Valuable information pointing to the real value and issues of the training is contained in those reports, but nobody seems to have read them. 14. No permanent training system for public enterprise personnel and oversight agencies' staff was set up. With the end of the project came also the end of the training program administered by BOX. The follow-up parapublic project finances an advisor at BOM whose TOR include setting up such a permanent training program.  SENEGAL - PARAPUBLIC OVERSIGHT SYSTEM FINANCIAL AND ADMINISTRATIVE CONTROLS LEGAL/ADMINISTRATIVE CONTROLS INSTITUTIONS MINISTRIES PRESIDENCY - PREMIERSHIP Technical Finance President Prime Minister Supreme Court of IIERARCIIICAL Ministries Ministry (abolished in 1983) Court Budgetary Discipline AUTHORITIES S GOVERNMENT CEP I CFP j C C C_E_P (CFP-DP OVERSIGHT C ACENCIES PUBLIC EP S E M (general) S N ENTERPRISES CPEP (exceptional) SGPR - Secrdtariat Gdndral de la Prdsidence de la R6publique CCP = Cellule des Contrats-Plans (est. 1980) CEP - Centre des Etablissements Publices (est. 1966) CVCCEP - Commission de Virification des Comptes et de Contrale COF = Contr5le des Opdrations Flnancires (division of CEP) des Etablissements Publics (est. 1972) ACC = Agence Comptable Centrale (division of CEP) EP o Etablissement Public CFP - Controle Financier de la Prdsidence (est. 1959) SEM a Socidtd d'Economie Mixte CFP-DP - Division Parapublique (crdde en 19'7) SN - Soci4t4 Nationale s_ X IGE - Inspection Grndrale d'Etat (est. 1906) ACP - Agent Comptable Particulier 0 co BOM = Bureau Organisation et Methodes (est. 1966) Mi  - 56 - ANNEX 8 Page 2 of 2 Comments on the Organigram of the Parapublic Oversight System The organigram outlines the complex Government oversight system over public enterprises in general prevailing during the period of project execution. One institution was added in November 1980 to this oversight system, namely the Cellule des Contrats-Plans, reporting to the Prime Minister until May 1983, when the Premiership was abolished, and to the Secretary General of the Presidency since then. All the oversight agencies mentioned on this diagram with the exception of the Inspection Generale d'Etat (IGE) were included in the Parapublic TA project and are further described in the body of this report. IGE is a special elite corps of inspectors working directly for the Presidency on specific assignments, including special reports on public enterprises. (In 1980 IGE prepared at the request of the Presidency a report on the management of CFP; the report which was quite critical was never released.) With the exception of CEP (ACC and COF), whose responsibilities only include control over EPs, all oversight agencies have some degree of intervention in EPs, SNs and SEMs. IGE, BOM and CCP involvement in a specific company normally occurs only on specific request from the Government or the company.