Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16091 PROJECT COMPLETION REPORT ARGENTINA TAX ADMINISTRATION TECHNICAL ASSISTANCE LOAN (3015-AR) November 1, 1996 Public Sector Management and Private Sector Development Division Country Department I Latin America and the Caribbean Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EOUIVALENTS December 1, 1989: Currency Unit= Argentine Austral (A) Austral 1.00= 7.8 cents at the official commercial rate January 1, 1992: Currency Unit = Argentine Peso (P) = 10,000 A Peso 1 00 = 0.991 US$ June 1, 1994: Peso 1.00 = 1.00 US$ GLOSSARY OF ACRONYMS AUD[NT Performance Monitoring Component CAPPER Training Component DGI General Tax Directorate ('Direcci6n General Impositiva") DNPF Directorate for Taxpayer Audit Programming ("Direcci6n de Normas y Planes de Fiscalizaci6n") DOSMIL Large Taxpayer Collection System Component EQUIPA Equipment Component GOA Government of Argentina GDP Gross Domestic Product IT Information Technology METFIS Taxpayer Auditing Methodology Component ("Metodologias de Fiscalizaci6n") MOP Memorandum of the President of IBRD to the Executive Directors of the Bank PCR Project Completion Report PIU Project Implementation Unit PPF Project Preparation Facility PSMTAL Public Sector Management Teclnical Assistance Loan PSRL Public Sector Refonm l.oan SAR Staff Appraisal Report SELPER Personnel Recruitment Component SICOFI Sistema de Control de Fiscalizacion SITER Information System on Economically Relevant Transactions SSFP Undersecretariat of Public Fianaces ("Subsecretanra de Finanzas Publicas") STAP Second Tax Adminstration Project TATAL Tax Administration Technical Assistance Loan UNDP UJnited Nations Development Program VAT Value Added Tax (" Impuesto al Valor Agregado") FISCAL, YEAR January I December 31 Vice President: Mr. Shahid Javed Burki Director: Mr. Gobind T. Nankani Division Chief: Mr. Paul Meo Staff Member: Mr. Luis Jose Mejia FOR OFFICIAL USE ONLY The World Bank Washington, D.C. 20433 U.S A. ROBERT PICCIOTTO Office of the Director-General November 1, 1996 Operations Evaluation MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT Subject: Project Completion Report on Argentina Tax Administration Technical Assistance Loan (Loan 3015-AR) Attached is the Project Completion Report (PCR) for the Argentina Tax Administration Technical Assistance project (Loan 3015-AR, US$6.5 million, approved in FY89 and closed 2.5 years behind schedule in March 1994) prepared by the Latin America and the Caribbean Regional Office, with Part II of the report contributed by the Borrower. The objective of the loan was to increase the efficiency of the General Tax Directorate (DGI) by improving control of tax compliance of large taxpayers, recruiting additional tax inspectors, improving training, tax controls and tax inspection techniques and providing additional computer equipment. These efforts at improving tax administration were considered key to the Government's objective of reducing the fiscal deficit. The loan became effective 11 months behind schedule, (after a new government had taken office and initiated a new economic reform program) and was amended twice, in September 1991 and September 1993. Other loans with similar or complementary objectives include a Public Sector Reform Technical Assistance Loan, (US$23 million, FY9 1), a Public Sector Reform Loan (US$325 million, FY92) and a second Tax Administration Technical Assistance Loan (US$20 million, FY92). The PCR concludes that lack of Government commitment during a period of political transition delayed the start of most project components until late 1991. At this time, the Government and the Bank reformulated the project in a way that produced significant improvements in tax collections. The most important project component, which developed a system to monitor tax compliance of large taxpayers, was highly satisfactory. The objectives of other components of the project were partially achieved. These components aimed at developing internal auditing and tax inspection techniques; and to provide training for new recruits. A component which sought to recruit tax inspectors and technical personnel was achieved only at a considerable cost. The PCR also finds that the Government's commitment to improving tax administration continues to be demonstrated by a nationwide expansion of such activities and by a perceived improvement in DGI's image as a credible enforcement agency. Based on the above, OED rates the project outcome as satisfactory, sustainability as likely and institutional development impact as substantial. OED also rates Bank performance as satisfactory because project restructuring and intensified supervision corrected initial weaknesses in design. Lessons from this project include that: (a) political stability is particularly important for implementation of technical assistance operations; and (b) improved civil service incentives, including adequate pay and career development are a must to achieve improved tax administration. The PCR is of satisfactory quality. It provides a good description of the project, the changes in its design and the achievement of objectives. No audit is planned. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. I FOR OFFICIAL USE ONLY PROJECT COMPLETION REPORT ARGENTINA TAX ADMINISTRATION TECHNICAL ASSISTANCE LOAN (LOAN 3015-AR) TABLE OF CONTENTS Page No. Preface ............................................................i Evaluation Summary ........................................................... iii PART I. PROJECT REVIEW FROM THE BANK'S PERSPECTIVE .........1... Project Identity ...........................................................1l Background ............................................................1 Project Design ............................................................2 Project Objectives and Description .......................... .................................. 2 Project Implementation ............................................................3 Second Phase of Project Implementation ............................................................5 Results ............................................................6 Project Impact ............................................................8 Sustainability ............................................................8 Bank Performance ........................................................... 10 Borrower Performance ........................................................... 10 Project Relationships ........................................................... 10 Consulting Services ........................................................... 10 Project Documentation and Data ..................... ...................................... 10 PART II. PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE ............... ................. 11 PART IIH. STATISTICAL INFORMATION ....................................................... 16 Table 1: Related Bank Loans ........................................................... 16 Table 2: Project Timetable ........................................................... 17 Table 3: Loan Disbursements ........................................................... 18 Table 4: Projects Cost and Financing ........................................................... 20 Table 5: Project Results ........................................................... 22 Table 6: Use of Bank Resources ........................................................... 23 Table 7: Status of Covenants ........................................................... 24 This document has a restricted distribution and may be used by recipients only in the performance of their l official duties. Its contents may not otherwise be disclosed wiLhout World Bank authorization. I PROJECT COMPLETION REPORT ARGENTINA TAX ADMINISTRATION TECHNICAL ASISTANCE LOAN (LOAN 3015-AR) PREFACE (i) This is the Project Completion Report (PCR) for the Tax Administration Technical Assistance Project in Argentina, for which Loan 3015-AR in the amount of US$6.5 million was approved on January 24, 1989. The Loan was closed on March 31, 1994 (the original loan closing date was September 30, 1991). The Loan was fully disbursed on November 19, 1994. (ii) Parts I and III of this PCR were prepared by the Public Sector Management and Private Sector Development Division of Country Department I, Latin America and the Caribbean Region. This report is based upon examination of the Memorandum of the President, Loan Agreement, Supervision Reports, Project Correspondence, Final Statements and Audit Reports, as well as interviews and materials collected during two supervision missions and one completion mission. Part II of the PCR was prepared by the Borrower, the Government of Argentina, which also commented on Parts I and III. - iii- PROJECT COMPLETION REPORT ARGENTINA TAX ADMINISTRATION TECHNICAL ASSISTANCE LOAN (LOAN 3015-AR) EVALUATION SUMMARY Objectives 1. The US$6.5 million TATAL Loan to the Government of Argentina, approved on January 24, 1989, was designed as an important component of the effort to stabilize the Argentine economy. TATAL financed improvements in tax administration to increase revenues as a part of the strategy to eliminate the persistent fiscal deficit without recourse to inflationary domestic credit creation. This increase would be obtained through the following components: (i) control of large taxpayer compliance (DOSMIL); (ii) improvement of internal control (AUDINT); (iii) improvement of tax inspection techniques (METFIS); (iv) training of staff (CAPPER); (v) additional personnel to carry out tax inspections (SELPER); and (vi) provision of additional computer equipment (EQUIPA). Implementation 2. Continuing changes in senior management delayed project execution, and resulted in two clearly demarcated implementation phases. Very little progress was achieved during the first phase, which covered the period from November 1988 to November 1990. During the second phase, which lasted from November 1990 until the end of the project in March 1994, implementation proceeded swiftly under the direction of the new and committed DGI management team and project implementation unit. The DOSMIL and AUDINT components were revised at this time, and DOSMIL converted from a monitoring system to a comprehensive collection system. In January 1991 DGI introduced their first Large Taxpayer Collection System, controlling the 920 largest taxpayers, nationwide. The system was very successful, and was immediately replicated in 36 more agencies throughout the country. Eventually it was installed in all 130 DGI agencies nationwide (financed by the follow-on loan Second Tax Administration Project, STAP 3460-AR), controlling the approximately 300,000 largest taxpayers. 3. The emphasis placed on DOSMIL required a reallocation of funds to this component and consequently curtailed the scope and execution of AUDINT and METFIS. Out of the six areas identified for the operation of AUDINT only four were begun, and two finished. The remaining four were finished under STAP. Studies were carried out under METFFIS to evaluate evasion trends in specific economic sectors. CAPPER, the training component, was completed, and is being expanded to non-Metropolitan Area staff through STAP. New personnel was recruited as planned through SELPER, although this component was delayed due to a hiring freeze of government employees. Equipment needed to operate the DOSMIL system was acquired under the EQUIPA component. - iv - Results 4. The new collection system (DOSMIL), that began operating in January 1991, increased VAT revenues from US$6.8 billion in 1991 to US$13.2 billion in 1992 (94%). The use of this system substantially reduced collection and return verification, errors in returns (43% in 1990 to 4% by January 1992), late submissions (51% to 3%), and late payments (40% to 10%). During 1993 errors and delays were negligible. In addition, with the new automated system, a typical DGI branch in Buenos Aires, with 12 staff, can service four times as many taxpayers as before. 5. The AUDINT component was partially achieved, in that the analysis and development of control systems for four areas were begun. Results cannot be formally assessed at the present time, as the 1 systems (including systems finished under the follow-on loan) are just becoming operational. Nevertheless, the new systems are expected to make operations more efficient and provide information for decision-making in the effective allocation of internal resources. In one case these systems have already proved their worth, detecting a considerable fraud carried out by certain DGI employees. 6. The METFIS component was partially achieved. The sectoral studies financed by this component were not used (although taken up later under the follow-on loan). The most important contribution of METFIS were studies to reduce industnral promotion exceptions and loopholes. 7. The objectives of SELPER were achieved but in a relatively costly manner for the agency. The targeted number of 600 tax inspectors and 50 technical personnel (35 programmers and 15 analysts) was fulfilled. An important number of recruited personnel still remain employed. 8. The objectives for CAPPER were achieved; all recruits were trained. A similar course of action is being followed to train staff outside the Metropolitan Area through the follow-on loan; moreover, closer ties between DGI and universities have resulted in more efficiently designed courses. Through the SELPER and CAPPER components, the total personnel working on taxpayer auditing tasks increased by almost 300% between 1989 and 1992 and the audits per month increased from 360 in 1989 to 1,800 in 1992, 9. In conclusion, in spite of delays and modifications, we may consider this project to be a resounding success. In addition to the quantifiable results mentioned above, it is important to mention that during this time there was a marked improvement in DGI credibility and taxpayer mentality due to perception of strict enforcement which DGI created during this period. Sustainability 10. Control of the large taxpayers has continued under the follow-on loan; the DOSMIL system now controls 300,000 large taxpayers. Further increases in collections will now depend on the success of the taxpayer auditing component under the follow-on loan. However, the long-term sustainability of the project will depend on giving more emphasis to human resources, particularly Systems: General Accounting (Budget Progranmning and Execution, Financial Accounting, Accounts Payable and Receivable); Personnel Administration (Recruitment, Benefits, Transfers, Control); Target Selection and Control; Taxpayer External Auditing; Taxpayer Internal Auditing, Judicial; Revision and Appeals; Acquisitions and supplies; Treasury; Tax Accounting; and Assets Management. The last two were installed under Loan 3460-AR. V to the introduction of an adequate incentive system tied to performance and supported by training. This is being at present considered by DGI. Lessons Learned 11. Lessons drawn from Bank experience applicable to this project fall into three categories: (i) those conceming difficulties in the implementation of tax administration projects; (ii) those conceming critical issues for sustainability; and (iii) best practices in tax administration. (i) Implementation Difficulties 12. In the first category, the two most important lessons are: (a) the need for political commitment by the Government to the reforms and its active support for their implementation and (b) the presence of a politically stable environment to ensure continuity and consistency of tax policies. The Minister of Finance needs to support the Tax Administration in withstanding powerful vested interests accompanied by a vigorous enforcement of tax laws. The Secretary of Public Revenue needs to support the organizational and personnel reforms which are an intrinsic part of tax reform. Support from the Congress is crucial for enacting necessary changes in the legal framework and in tax structure. 13 The Project Implementation Unit (PIU) staff should be carefully selected, with an experienced project coordinator hired and PIU staff trained in project administration before implementation stars. Finally, detailed project preparation is required. (ii) Sustainability 14. The most important lesson (noted in several similar cases) is that inadequate conditions of pay and career development cause severe damage to the sustainability of institutional efforts. Attracting and retaining qualified staff depend on a compensation system that is competitive with the private sector (without necessarily entailing total parity), particularly by offering an attractive carrer path tied to training and objective performance measurement. Efficiency in the public sector requires that the compensation policy include both rewards for good perforrnance and sanctions for non- or poor performance. 15. Because revenue collection is the most visible function of tax administration, there is a tendency to over-emphasize computerization vis-a-vis other important areas like organization and management, human resources, training, and taxpayer auditing. Lasting improvements in tax administration performance will only be possible if all critical institutional weaknesses are addressed. 16. Similarly, the encouragement of voluntary compliance has not received sufficient importance. Only a small amount of the collection can be directly controlled, and the remainder depends upon the taxpayer's perception of the fairness and equity of the tax system, as well as of the costs of non-compliance. It is therefore necessary to improve the quality of services provided by the tax administration to the taxpayers, and introduce grievance redress and appellate mechanisms that enhance the transparency and accountability of tax administration. -vi- (iii) Best Practices 17. In the third category, a number of best practices have been identified. These include: (i) introduction of a unique taxpayer registration number for effective monitoring of taxpayer activity and compliance; (ii) organization of tax administration based on functions, instead of individual tax laws; (iii) delegation of responsibility and authority to middle management, accompanied by training and better management systems; (iv) prioritization of assessment, investigation and collection activities; (v) development of a single current account for each taxpayer for efficient collection of tax dues; (vi) strengthening of the legal framework to facilitate enforcement of tax laws and streamlining of procedures; and (vii) carrying out of an effective public relations campaign to educate taxpayers regarding their legal obligations. ARGENTINA TAX ADMISTRATION TECHNICAL ASSISTANCE LOAN (LOAN 3015-AR) PART I: PROJECT REVIEW FROM THE BANK'S PERSPECTIVE 1. Project Identity Project Name: Tax Administration Technical Assistance Loan Loan Number: 3015-AR Regional Unit: Latin America and the Caribbean Region Country: Argentine Republic Sector: Public Sector Management 2. Background 2.1 In the last half of the 1980s Argentina suffered from a long period of stagnation, mainly due to chronic public sector deficits and endemic inflation. After the return to constitutional democracy in 1983, public demands for the control of inflation were addressed by four successive stabilization programs each of which failed to redress rapidly and permanently the structural deficit of the public sector. The public sector deficit averaged 14 percent of GDP in the 1980-84 period and 8.9 percent in 1985-89. These endemic deficits had their origin in an erosion of revenue capacity and an inability to contain expenditures. 2.2 Poor performance of the tax administration during Total Revenues 1980-1989 these years was central to the (As Percent of GDP) erosion of tax revenues and 20 gave rise to rampant evasion. On the other hand, the 16 excessive fragmentation of the 12 tax system, inconsistency in legislation, and the complex, 8 -* and at times contradictory laws 4 U Social Security and regulations, produced a ' U. Tax Revenue legal framework difficult to o *.' enforce. The tax burden, which ° x N v LO CD @ co 0AC A_ 0c 0c 0 0C 0) 0 0 o) 0 had reache 1l8.7 percent of a- a) GDP in 1980, fell 6 percentage points to 12.7 percent in 1989. Revenues from the Value-Added Tax (VAT), which peaked at over 5 percent of GDP in 1981, had fallen to under 2 percent by 1989; their contribution to total revenues declined from 23 percent in the 1980-84 period to under 15 percent in 1989. The VAT became riddled with exemptions, primarily for industrial promotion. 2.3 In the fall of 1988, at the request of the Government of Argentina, the Bank developed a program of support for the stabilization plan, known as "Plan Primavera". This plan included a comprehensive tax reform to restore buoyancy to revenues, through a simplification and modernization of the tax structure. 2 To implement these reforms the Government asked the Bank to undertake, urgently, a program for substantial upgrading and reformn of DGI. 2.4 The Tax Administration Technical Assistance Loan (TATAL, Ln. 3015-AR), for US$6.5 million, was prepared in record time. It concentrated on strengthening DGI's collection and taxpayer auditing effort. Previous, mostly unsuccessful Bank projects in this area were made through the Public Sector Management Technical Assistance Loan (PSMTAL 2712-AR), approved in June 1986 and completed by April 1992. 3. PrQject Design 3.1 The project was identified in July 1988. Based on previous work done by the Government and the Bank', the project was prepared in only four months (between July and October 1988). A request to transform pre-appraisal into appraisal was approved by management in November 1988, and the project was approved by the Board on December 21, 1988. Total project costs were estimated at US$10.67 million, of which the Bank financed US$6.5 million. 3.2 A major influence on the design of the project was the government's urgent need to obtain maximum short-term revenue impact with minimum delay and to attain its revenue goals under the 1989 macroeconomic program. The primary functions of tax administration -- facilitating and monitoring taxpayer compliance and preventing taxpayer non-compliance -- had been operating very inefficiently. The tax roster deteriorated during most of the 1980s; additionally, data were collected manually with a high degree- of error. Low yield and high cost of taxpayer inspections and audits characterized Argentina's tax administration during the 1980s. 4. Project ObJectives and Description 4.1 The project's principal objective was to improve government finances by increasing the efficiency of the General Tax Department (DGI), mainly in the areas of taxpayer and internal auditing. The specific objectives of the project were the following: (i) Control of tax compliance of large taxpayers; (ii) Availability of additional human resources to carry out tax inspections; (iii) Training of DGI personnel; (iv) Improvement of internal auditing and development of indicators of internal efficiency to monitor the evolution of the agency's effectiveness; (v) Development of adequate tax inspection techniques, particularly for systems supported by mainframe computers and industry-specific fields; and (iv) Provision of computer equipment. I Argentina: Public Sector Investment Review (Report 6354-AR of November 17, 1986) and Argentina: Tax Policy for Stabilization and Economic Recovery (Report 8067-AR of November 8, 1989 3 4.2 Project Components. To obtain these objectives, the project was divided in the following components: i) DOSMIL. Develop a computerized system of "surveillance" of the 2,000 largest taxpayers in order to monitor their group's compliance with VAT, income, and capital taxes. Information from tax returns would be entered into a small multi-user computer (see EQUIPA) under the responsibility of a special group of taxpayer auditors. (US$0.09 million). ii) AUDINT. Review procedures and systems for internal auditing and control, and carry out comprehensive audit of DGI operations. (US$1.2 million). iii) METFIS. Review current techniques for inspection of taxpayers, and in particular the analysis of accounting systems in mainframes. Assess and improve, were necessary, DGI's tax inspection techniques as applied to specific industries. (US$2.3 million). iv) CAPPER. Train the newly hired inspectors and computer specialists and review the structure of DGI's training unit. (US$2 million). v) SELPER. Hire 600 more tax inspectors, representing a 50 percent increase over the existing workforce, and 50 computer analysts and programmers. (US$0.36 million). vi) EQUIPA. Acquisition of personal and multi-user computer equipment to improve DGI operations and encourage decentralization. (US$3.2 million). vii) The remainder consists of repayment of the PPF (US$0.73 million), price contingencies (US$0.78 million) and project administration (US$100,000). 4.3 In 1990 the project was modified and two components were substantially changed (DOSMIL and AUDINT). Greater emphasis was given to DOSNIL, reallocating funds accordingly. 5. Poec Imleenaton 5.1 The project was processed according to the timetable in Table 2. Loan effectiveness was delayed by 11 months because of changes in DGI management, who requested revisions of the project. This late start of the project resulted in two requests for extension of the Closing Date, to March 31, 1994. A Project Preparation Facility (PPF) funded activities between the beginning of the program in November 1988 and loan effectiveness 14 months later, February 28, 1990. About US$627,000 out of a total of US$773,000 was used (see Table 4) for additional project preparation as well as for the selection of taxpayer auditors and computer analysts, development of DOSMIL software, and undertaking a preliminary diagnosis of internal audit functions. 5.2 The project was implemented by through a Project Implementation Unit (PIU), which was composed of consultants hired for this purpose. A UNDP Management Services Agreement was prepared for equipment procurement (US$3 million); a UNDP cost-sharing agreement was set up for consulting services and training. 5.3 The MOP correctly identified as major project implementation risks the strong institutional challenge given to DGI's management, possible disruption caused by the May 1989 elections, and the expected change in DGI's Senior Management planned for November 1989. These fears were eventually 4 fulfilled, and implementation of the project was unsatisfactory until late 1989. Project execution thus took place in two clearly demarcated phases: the first phase covering the period from November 1988 to November 1990, and the second from November 1990 to the end of the project, in March 1994, which has been much more productive. 5.4 During the first phase, all components fell behind schedule despite the Bank's efforts through supervision missions to keep activities on track. By the end of FY90, cumulative disbursements only were 13 percent of estimates (see Table 3A). In particular, recruitment of inspectors and analysts (SELPER) was delayed due to the lack of acceptance by DGI of the examination methods used by the consulting firm and a general freeze on the number of public employees. This, in tum, delayed the beginning of the CAPPER component. Progress for the other components (METFIS, AUDINT and EQUIPA) was equally slow. This can be attributed in large part to the lack of agency ownership of the project, especially in the transition between two administrations,2 and was further aggravated by several changes in DGI's senior management. 5.5 The second phase began one year after the DGI Director General took office in November 1989. A new PIU was set up, with a new organization and staff. By the end of 1990 the project was redefined by DGI, with input from the IMF and the Bank. The original objective of the project, i.e. to increase revenues by improving DGI's efficiency, was maintained. The changes in the project were reflected in revisions to the DOSMIL and AUDINT components, with greater emphasis being given to the former. (a) DOSMIL. The simple monitoring system originally envisioned was converted into a full Large Taxpayer Collection System based on a single taxpayer account, processing tax retums and calculating liabilities, automatically issuing notices, and calculating fines. This system was developed using a relational data base operated in a decentralized fashion, using open architecture multi-user computers to ensure vendor independence. The EQUIPA component basically was merged with this component to provide equipment for the installation and replication of the system. (b) AUDINT. The improvement of auditing procedures was considered premature in view of the changes planned for the structure and functions of DGI, instead, the recommendation was to consider the development of a number of administrative, control, and operation systems for the following functions: * General accounting * Personnel administration * Taxpayer auditing functions * Legal functions * Acquisitions * Treasury 2 The Justicialist Party took over in May 1990, five months before term, from the Radical Party. The transition between two administrations contributed to the lack of continuous authority and clear-cut policies over both the agency and project. S Second Phase of Project Implementation 5.6 In February 1990, the Government implemented a major tax reform as conditions of effectiveness for the Public Sector Reform Loan (3394-AR). This reform broadened the base of VAT and extended coverage first to all goods and later to services in November 1990. The tax penalty law, adopted by the Congress in 1990, provided much-needed sanctions for tax non-compliance, and evasion was made a criminal offense. Another tax package of February 1991 improved the quality of the revenue mobilization effort, eliminating many of the so-called tax handles, easy-to-collect but growth-inhibiting taxes on financial transactions(such as export taxes), and several low-yield taxes. All these changes addressed the need to simplify the tax system and raise efficiency in tax collection. 5.7 DOSMTL. During the first phase of the project, this component had remained inactive due to a perception within DGI of its ineffectiveness and the lack of insertion of the concept with line personnel. When new leadership took over in November 1989, the agency tried to develop a mainframe-based software to create a taxpayer current account with the technical assistance of IBM. When this failed, the agency tried a decentralized approach, which resulted in the DOSMIL Large Taxpayer Collection System. 5.8 The operation of the Large Taxpayer Collection System (DOSMIL) began in January 1991 controlling VAT returns for the 920 largest firms in the country. The system began as a pilot in the "Grandes Contribuyentes Nacionales" agency, which deals exclusively with National Large Taxpayers. Subsequently a variation was developed for installation Customs in Buenos Aires to obtain VAT information on imports, (nearly 60 percent of customs revenues), for taxpayer auditing purposes. The customs collection system was implemented in August 1991. At this point, the Government gave highest priority to replicating DOSMIL throughout the country. By May 1992 the system was installed in another 36 agencies: 16 in the Buenos Aires Metropolitan Area, 14 in the provinces, and six in agencies considered important, controlling the aggregate contributions of the 50,000 largest taxpayers in the country. 5.9 Replication involved office refurbishing, provision of hardware, software and cabling, training of about 1,700 agents, obtaining data to prepare the database and informing large taxpayers the new tax collection procedures. Finally the implementation of the new system was linked to a follow-on Loan 3460- AR (Second Tax Administration Project, STAP). Under STAP, another 100 agencies received the system in the subsequent year, controlling a total of 260,000 taxpayers. The generalization of the DOSMIL system for medium and small taxpayers (SICOFI System) began under TATAL; implementation would follow under STAP. 5.10 AUDINT. Like all the components except DOSMIL, progress for AUDINT was slow and implementation delayed. A review by the Undersecretary of Public Sector Management of AUDINT elicited questions on DGI's strategy for reviewing and improving internal auditing. The scope of AUDINT was redefined (see para. 5.5). Development for four systems was begun under this loan (i.e., General Accounting System, Personnel Administration System, Taxpayer Auditing Functions, and Legal Area), development for the other two systems (i.e., Acquisitions and Treasury) remaining to be implemented under STAP. 5.11 METFIS. Studies of evasion trends in the industrial, commercial, and agricultural sectors were partially carried out, but did not reach the implementation stage. A more comprehensive execution of fraud-detection methods is being undertaken through STAP. The importance given to DOSMIL caused the taxpayer auditing component to be neglected during the second stage of the project. 6 5.12 CAPPER. The hiring of additional tax inspectors was delayed, and training consequently fell behind schedule. Training was eventually conducted at the public University of Buenos Aires, where 919 recruits received training in auditing and data processing.3 In May 1992 a Bank supervision mission noted the inadequacy of the courses given the new and future institutional demands of the agency. Regular training programs were observed to be restricted to short "one-shot" introductory courses taken by some officers who had not previously taken any other courses during their employment at DGI. CAPPER did not consider the MOP recommendation that the outline of a plan for future training activities must be linked to retraining efforts of existing and newly- recruited DGI staff. 5.13 SELPER. The recruitment of new personnel took place after the original deadline of June 1989, due to the Emergency Law of January 1990 that suspended all government recruitment. A special request made by the DGI director to the Ministry of Economy to waive the mentioned law, allowed the DGI to implement the recruitment plan from August 1990. Recruitment was completed, although in a costly manner for the agency: DGI paid a consulting firm (selected through competitive bidding) approximately US$1,000 per agent recruited, at a time when monthly salaries for inspectors were about US$300.4 Additionally, DGI management assuming authority in November 1989 repeated the entrance exam for applicants from within DGI before making the final selection, thus bringing costs even higher. 5.14 Eventually, however, these problems were overcome, and the number of DGI personnel engaged in taxpayer auditing increased from 1,400 agents in 1989 to 6,500 agents in 1992. This was crucial in increasing the number of monthly field audits. 5.15 EOUIPA. This component was merged with DOSMIL. Thirty-seven multi-user computers were bought in support of the automation of the Large Taxpayer Collection System. 6. Results 6.1 Empirical exanination of tax receipts over the 1983-92 period reveals that improvements in tax administration and tax legislation contributed significantly to tax revenue growth after 1990, from 15.2% of GDP in 1990 to 24.4% in 1992. Specifically for VAT revenues, the effects of the 1990 reforms were outstanding; the reduction of tax exemptions, increasing VAT collection from US$3.1 billion in 1990 to US$6.8 billion in 1991 (an increase of 120% in one year). The new collection system (DOSMIL), that began operating in January 1991, increased VAT revenues from US$6.8 billion in 1991 to US$13.2 billion in 1992 (94%). The use of this system substantially reduced collection and return verification, errors in returns (43% in 1990 to 4% by January 1992), late submissions (51% to 3%), and late payments (40% to 10%). During 1993 errors and delays were negligible. In addition, with the new automated system, a typical DGI branch in Buenos Aires, with 12 staff, can service four times as many taxpayers as before. 6.2 The availability and reliability of information for auditing taxpayer was greatly increased. This information is being used to conduct audit searches both globally by economic activity, and by taxpayer. DGI now collects monthly information on taxpayers5 which can be used to measure the performance of various economic sectors and permits a first selection of taxpayers for auditing. Another outstanding benefit is the new role which the Directorate of Studies has been able to assume. This Directorate has evolved from obtaining information and preparing simple tables to deeper analyses using software tools to 3 CAPPER provided a total oft35 specialized courses - 5 on personnel management, I on intemal auditing, 27 on taxpayer auditing, and 2 on communication techniques. 4 One monthly salary is normally charged by recruitment consultant firms. A subsequent contract renegotiation reduced the actual component cost by 10 percent. 5 This must be compared with the Central Bank information which is quarterly, collected based on surveys of economic activity and available with large time lags. 7 process DOSMIL-produced aggregate data. Using DOSMIL data, DGI has been able to monitor the performance of local agencies and measuring targets against monthly performance, a process that previously took four months and was plagued by errors. To summarize, the results of this component were highly satisfactory and beyond initial expectations. 6.3 The objectives of SELPER were achieved but in a relatively costly manner for the agency. The targeted number of 600 tax inspectors and 50 technical personnel (35 programmers and 15 analysts) was fulfilled.6 An important number of recruited personnel still remain employed by the agency, with a retention rate of 84 percent for inspectors, 77 percent for programmers, but only 20 percent for analysts. On the other hand, this effort resulted in an excellent methodology for the selection of staff that is now used by the Directorate of Human Resources. A conclusion to be drawn from using consulting firms for recruitment is the importance of close supervision by line staff personnel. 6.4 The objectives for CAPPER were achieved; all recruits were trained. Courses were given in blocks of three months through contracts with the University of Buenos Aires at a reduced cost compared to consulting firms. In addition to University teachers, higher-level DGI staff also provided instruction. A similar course of action is being followed to train staff outside the Metropolitan Area through the follow-on loan; moreover, closer ties between DGI and universities have resulted in more efficiently designed courses. As mentioned, through the SELPER and CAPPER component, the total personnel working on taxpayer auditing tasks increased by almost 300% between 1989 and 1992 and the audits per month increased from 360 in 1989 to 1,800 in 1992. 6.5 The AUDINT component was partially achieved, in that the analysis and development of control systems for four areas were begun. Results cannot be formally assessed at the present time, as the 117 systems (including systems finished under the follow-on loan) are just becoming operational. Nevertheless, the new systems are expected to make operations more efficient and provide infornation for decision- making in the effective allocation of internal resources. In one case these systems have already proved their worth, detecting a considerable fraud carried out by certain DGI employees. Benefits are expected in the following areas: (i) taxpayer auditing: control of inspectors, improved records management, reduction of losses and errors, and faster turnover of cases; (ii) human resources: integration of all personnel information (currently in three separate and inconsistent data bases) into a single data base; (iii) legal: an infonnation system taxpayer on suits, fines, etc. that have been applied. In parallel, DGI acquired a software package to register information on the 220,000 indictments in which the agency is involved; and (iv) treasury: budgeting of revenues including DGI's own budget and revenues collected from taxpayers. One important effect of the introduction of the systems is the extent to which they facilitate the collection and evaluation of information that can be used to establish agency performance indicators. 6.6 The METFIS component was partially achieved. The sectoral studies financed by this component were not used (although taken up later under the follow-on loan); however, METFIS also financed studies done by Secretariat of Public Finance that helped reduce industrial promotion exceptions and loopholes. The success of the industrial promotion scheme under PSRL was much enhanced by DGI's strong policy to enforce new rules that made tax evasion a criminal offense. 6 A discrepancy exists between the number of new recruits provided by the Bank and the number quoted by DGI. According to the Government, 550 tax inspectors, 18 programmers, and 12 analysts were hired. 7 Systems: General Accounting (Budget Programming and Execution, Financial Accounting, Accounts Payable and Receivable); Personnel Administration (Recruitment, Benefits. Transfers, Control); Target Selection and Control; Taxpayer External Auditing; Taxpayer Intemal Auditing; Judicial; Revision and Appeals; Acquisitions and Supplies: Treasury; Tax Accounting; and Assets Management. The last two were installed under Loan 3460-AR. 8 6.7 In conclusion, in spite of delays and modifications, we may consider this project to be a resounding success. In addition to the quantifiable results mentioned above, it is important to mention that during this time there was a marked improvement in DGI credibility and taxpayer mentality due to perception of strict enforcement which DGI created during this period. 7. Pro-ject Impact. 7.1 Project impact can be measured by an increase in revenue collection, which improved notably during the life of the project. Tax collection grew by 6.1 percent of GDP between 1990 and 1993 (see Table 5A). Even though macro variables, such as GDP growth, inflation, and political stability, influenced tax receipts during this period, empirical examination of tax receipts over the 1983-92 period reveals that improvements in tax administration and tax legislation also contributed significantly to tax revenue growth 8 after March 19908. Between the fourth quarter of 1989 and the first quarter of 1991, tax effort change is mainly explained by tax policy changes (its index increased by 54.8 percent), specifically the reduction in the number of exemptions (para. 5.6). Since then, tax policy contribution to the tax effort has been negligible, with increases brought about mainly by improved tax administration. 8. Sustainability 8.1 Control of the large taxpayers has continued under the follow-on loan; the DOSMIL system now controls 300,000 large taxpayers. Further increases in collections will now depend on the success of the taxpayer auditing component under the follow-on loan. Another measure of sustainability is the change in the image of DGI and of the public perception of need to pay taxes. Although no indicators are available to measure this change, anecdotal information (e.g., newspapers, informal interviews, etc.) strongly suggest that DGI is well accepted, and that paying taxes is far from the "fool's game" image that it had over the last decade and before. However, the long-term sustainability of the project will depend on giving more emphasis to human resources, particularly to the introduction of an adequate incentive system tied to performance and supported by training. This is being at present considered by DGI. 8 Effects of Tax Reform on Argentina's Revenues, by Jacques Morisset and Alejandro Izquierdo. World Bank, WPS 1192 of September 1993. 9 Figure 2. Summary of Accomplishments and Sustainability Component TATAL(Loan 3015) Follow-on Activities DOSMIL Successfully installed Large Replicated in 97 additional agencies; Systems Taxpayer Collection System for small and medium taxpayers developed in 36 agencies. and installed in 57 of these agencies. METFIS Studies of various sectors; Studies on the agriculture, meat-packer, results used in follow-on textile and metal-mechanic industries. project. Review of the industrial promotion scheme. Further studies on credit cards, stockbrokers, and prepaid health industries. SELPER 600 inspectors, 35 504 inspectors, 27 programmers and 3 programmers, and 15 analysts still with DGI. analysts hired. CAPPER Agents in Buenos Aires Training scheme is being applied to agents in trained. the provinces. Restructuring of the Training Directorate and introduction of courses relevant to the present and future needs of the agency are in progress. EQUIPA Multi-user computer Additional 1,200 personal computers, 130 equipment acquired. RISC machines, 300 laser printers, and 1,000 lap-tops acquired. AUDINT Systems in operation by Continuing development and installation of May 1994. systems. Selected modules to be replicated in agencies and regions. lo 9. Bank Performance 9.1 Monitoring and supervision were frequent and effective, and providing continuity through DGI's management turnover. The mix and quality of specialists were appropriated and contributed with the necessary guidance in the earlier years. In particular, the Bank established a framework and guidelines for planning, project execution, and procurement that have been repeatedly acknowledged for its importance to the success of the project. 10. Borrower Performance 10.1 After a shaky start, the Borrower executed the project satisfactorily. In 1989 a new PIU was set up, with direct access to the DGI General Director. The PIU place a strong emphasis on hands-on project execution, which resulted in dynamic progress in the latter stages of the project. However, as a consequence, two areas were neglected: the measurement of project performance indicators and the preparation of progress reports and actions plans. As a result, while loan covenants were complied, some of them were late (Table 4). 11. Project Relationships 1 1.1 The IMF's assistance, provided by the Fiscal Affairs Department, was crucial to the success of the tax administration reform. The Fund provided guidance for several years through repeated supervision missions, and most importantly substantially contributed to the idea of a Large Taxpayer Collection System. Relationships between the Bank and the Borrower were excellent. Project consultants and DGI management and staff were well integrated. 12. Consulting Services 12.1 Consultant performance was poor in the first phase of the project, mainly in the recruitment component. In the second phase, with new management and tighter Bank supervision, the performance of both individual consultants and consulting firms improved. 13. Project Documentation and Data 13.1 Government auditors found a steady improvement in project documentation. In 1991, around 6 percent of all documents were missing; in 1992, only 0.76 percent were missing. The auditors also noted a marked improvement in the preparation of terms of reference and bidding documents. 11 ARGENTINA TAX ADlMINSTRATION TECHNICAL ASSISTANCE LOA LOAN 3015-AR PART II: PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE Introduction and Conclusion Loan 3015 (Tax Administration and Technical Assistance Loan) was part of a structured set of activities through which The World Bank supported the efforts of the Government in its policies of reform and modernization of the public sector. Project 2712 (Public Sector Reform Loan), the antecedent of Loan 3015, included the strengthening of tax administration in one of its components, but failed to achieve the required results of this effort. Loan 3460 was conceived with the objective of further expanding and completing the structural transformation and results achieved under Loan 3015. Loan 3015 constituted the first organized effort aimed at strengthening the capabilities of tax administration and automating the process of collection. In December of 1989, the objectives of the project were accurately redefined through the reformulation of the DOSMIL, EQUIPA, AUDINT, METFIS, CAPPER and SELPER components: the last component being practically finished at that time. Impact Loan 3015 was the first organized effort to support the automation of tax administration activities, including the improvement in human resources required by the changes, using DGI staff, individual consultants, and consulting firms. As a result of the activities carried out through its various components, the project began the transformation of DGI's image before the Argentine people. DGI's computerization plan envisaged decentralizing control of activities and data capture and distributing operations in an integrated manner using a modem data base administrator. In addition, DGI decided to use UNIX-based, open architecture equipment. Among others, the following systems were developed: collection, taxpayer auditing, internal control, and management. In spite of its complexity, emphasis was given to the taxpayer collection system, for which the taxpayer roster was updated. A measure of its success was that in 1992 twice as many taxpayer returns as in 1991 were processed; in 1993 three times as many were processed. The acquisition of computer equipment was initially carried out through Loan 3015, then through counterpart funding and finally through Loan 3460. 12 During the early phase of the project, the organization was equipped with 45 microcomputers of the AT type and 31 data-entry machines, the latter already obsolete. At the end of Loan 3015 and with 60 percent increase in personnel, DGI had 150 multi-user computers with around 21 work stations and, in addition, 1,600 personal computers. This increase in equipment, together with an improvement of resources, continues today with the implementation of Loan 3460-AR. Every one of the components had, depending on its area, an important effect both internally and externally. The DOSMIL component developed and implemented the single taxpayer current account system, initially controlling the 2,000 largest taxpayers in the country. Subsequently, the installation of the system in 131 agencies nationwide (to the 10th of January 1994), controlling on line an average of 2,000 taxpayer per agency, permitted a close control of the 262,000 taxpayers who contribute 65 percent of the total revenues. Other effects of this system were the increase in the presentation of taxpayer returns and a strong reduction in the number of errors contained therein. The SICOFI component developed the taxpayer current account for the rest of the taxpayers. Although the same controls are used as for the DOSMIL system, a number of taxpayers preclude having an on-line solution and processing of tax returns is therefore carried out later in the day ("batch processing"). The SICOFI system was implemented as a pilot in Agency No. 8 of the Buenos Aires metropolitan area in May 1993. The effect of the success so obtained (increases in collection) developed the interest of the provincial tax administrations as well as the interests of various countries in Latin America. Today, seven provinces either have implemented or are in the process of implementing the system. Guyana, Nicaragua, Jamaica, and El Salvador have expressed an interest in adopting this technology. The AUDINT component focused on the development of control systems (for the operational areas excluding collections) and areas of the central administration. The operation systems (selection and control, taxpayer auditors, judicial, revision and appeals) were developed and implemented successfully in Region No. 5 in the metropolitan area and the administration systems at headquarters. The systems were implemented in three phases: First Phase: SICOGRA - General Accounting SIPREJE - Budget Programming and Execution SILIPER - Personnel Administration (Recruitment, Benefits, Transfers, Control) Second Phase: SISESCO - Target Selection and Control SIFEXT - Taxpayer External Auditing SIFINT - Taxpayer Internal Auditing 13 SIJURU - Judicial * SIRECU - Revision and Appeals SICOSU - Acquisitions and Supplies SITESO - Treasury Third Phase (installed under Ln. 3460): SICREC - Tax Accounting (Bank Collection, Returns, Collections, Accounting Register) SIPAMO - Assets Management In November 1993, the systems of phases I and 2 were implemented in normal operation. The AUDINT component thus substantially modified the internal work of the DGI. The METFIS component produced evasion studies of different sectors as follows: * Petrochemicals: plastics Metal/Mechanic: openings (doors and windows, auto parts) * Cotton/Textiles: sport clothes * Agro-industry: meats These evasion studies are being continued on other sectors through the METFIS component of Loan 3460- AR. In the first stage, the Government gave priority to collection over tax auditing, thus postponing this last area to the second stage. The DOSMIL received more emphasis in order to obtain the immediate results through the collection of recognized debts of taxpayers; for this reason the agencies were equipped with the DOSMIL system with a great psychological effect on evasion. The CAPPER component carried out training programs in information technology for taxpayer auditing personnel through an agreement with the University of Buenos Aires. This was of great importance given the complexity of the information system which must be analyzed during inspection. The SELPER component recruited inspectors to the different units of taxpayer auditing, most of whom are still with DGI. 550 of the 600 inspectors required were hired, as well as 12 out of the 15 analysts and 18 out of the 35 programmers required. Evaluation of the World Bank's Performance The successive missions of the Bank, nineteen in four years, permitted close contact with the project implementation unit (PIU). These missions permitted great flexibility in adopting a plan to changing conditions, providing opportune suggestions through the specialists in various areas who formed part of 14 these missions. In addition, the missions provided original solutions to problems encountered by the implementors of the components. Evaluation of the Borrower's Performance The continuity of management of the PIU was key to the success of the project. DGI responded to every project need with opportune political decisions which assured the feasibility of speed of execution, including two visits to the Bank by the Director General and the Project Coordinator. The first visit had as objective the redefinition of Loan 3015 and the second visit had as objective to begin processing of the second loan to consolidate in the medium term the success obtained thus far. In addition, DGI contributed all needed counterpart personnel, offices, and equipment as required. DGI carried out other activities which complemented the project objective. Some of these activities are: a) the creation of 13 new local agencies; b) the redefinition of the incentive scheme, including salary increases starting December 1991; c) the restructuring and strengthening of training area; d) the creation of support and maintenance units for the new computer applications; e) the hiring of systems staff for the administration and maintenance of new applications; and f) the hiring of around 4,000 students of economics, sciences, and law to assist in inspection tasks. Relationship Between the Borrower and the Bank Since 1989, the relationship between the Bank and Borrower firms up and missions become particularly fruitful. The supervision missions had a strong effect in increasing the motivation of the project work group, enabling DGI staff to accelerate the pace of their work and complete their tasks in a more timely manner. UNDPOS Pefomane UNDP: The office provided satisfactory services in the processes of contracting individual consultants and consulting firms. These funds were processed through a cost-sharing agreement. OPS: This office carried out all services related to the acquisition of goods. Financial information was not always provided in a timely manner, especially when carrying out the yearly audits. PIU Perfornance The PIU was an adequate tool for implementation of activities agreed between the Government and the supervision missions. However, due to urgent matters issued from the dynamic execution of the project, planning and monitoring received little attention. This situation was identified and corrected through Loan 3460. The PIU throughout maintained an efficient administration and control of all processes having to do with funds. Lessons of Experience The changes of leadership at the beginning of the project, added to the economic and social instability in the country, were factors of strong negative incidents. On July 8, 1989, Dr. Carlos S. Menem assumes presidency of the nation and in December of the same year, Mr. Ricardo Cossio takes charge of the DGI. Starting with this change in leadership and permanence of the authorities, a period of deep and effective modifications begins. Nevertheless, previous experience with the project resulted in resistance to change on the part of DGI staff. To this resistance must be added the natural resistance to changes in functions such as 15 that which would take through the pervasive introduction of information technology throughout this project. This situation was exacerbated by the lack of experience in planning and project monitoring at the beginning of the project, a lack which was addressed progressively throughout the implementation of the project. Lack of internal information dissemination caused resistance as well as wrong decisions throughout the early parts of the project, resulting in wasted efforts and lost time. A more thorough information campaign should have been carried out at the beginning. Most of the problems encountered in the integration of DGI staff during project implementation were caused by this initial lack of information. 16 PROJECT COMPLETION REPORT ARGENTINA TAX ADMINISTRATION TECHNICAL ASSISTANCE LOAN (LOAN 3015-AR) PART m: STATISTICAL INFORMATION Table 1: Related Bank Loans Project/Loan No. and Title Purpose Approval Status 2712-AR Improve Government's capacity to manage the June 1986 Completed economy through the strengthening of key agencies. PSMTAL In the case of tax administration, US$0.4 million were destined to studies on legislative reform, and US$1.4 million to the redesign of basic systems and training. 3280-AR Provide incentives for provinces to execute December Currently adjustment programs for revenue enhancement and 1990 in Provincial Development expenditure control through the Provincial Financial Project Action and Investment Plan (FAIP). The main operation objective of the FAIP was to produce a current account surplus that would enable provinces to finance new investments (US$200 million). 3394-AR Generate changes in revenue mobilization (US$325 July 1991 Completed million), which included the consolidation of VAT SAL:PSRL and the revision of the income tax law to eliminate loopholes. 3460-AR Obtain sustainable revenue increases and facilitate March 1992 Currently voluntary compliance through simplification of the in STAP legal framework and strengthening of the operation administration and in particular the collection function of the DGI (US$20 million). 17 Table 2: Project Timnetabk Activity Planned Date Actual Date Identification 07/88 07/88 Pre-Appraisal and Appraisal Missions 10/88 10/88 Approval of PPF 11/88 11/88 Negotiations 11/88 12/88 MOP 12/88 12/88 Board Approval 01/89 01/89 Loan Agreement 03/89 03/89 Effectiveness 03/89 02/90 First Loan Amendment 09/91 Second Loan Amendment 09/93 Loan Closing 09/91 03/94 Loan Completion 03/91 10/93 18 Table 3: Loan Disbursements A. Cumulative Estimated and Actual Disbursements (US$ Million) Bank Fiscal Year 1989 1990 1991 1992 1993 1994 MOP Estimate 3.6 5.8 6.5 - - Actual 0.0 0.77 1.5 5.0 6.2 6.5 Actual as percent of 0 13 23 77 95 100 Estimate Date of Final Disbursement: November 19, 1993 TATAP (Loan 3015-AR) CumulativeDisbursements (1989-1994) (US$ Million) 4- 1989 1990 1991 1992 1993 1994 Years r. Estimate O Actual '9 Table 3: Loan Disbursements (Continued) B. Actual Disbursement Dates and Amounts (US$ Date Disbursements 02/28/90(PPF) 773,442.22 07/26/90 103,058.00 08/28/90 627,000.00 03/27/91 150,000.00 04/12/91 75,000.00 08/14/91 500,000.00 12/04/91 490,312.00 12/31/91 375,591.00 01/31/92 1,134,097.00 06/29/92 1,000,000.00 07/28/92 478,707.00 04/06/93 500,000.00 06/04/93 278,993.00 11/19/93 215,887.93 TOTAL9 6,495,972.15 Actual Cumulative Disbursements TATAP (Ln.301 5-AR) 7,000,000 6,000,000- 5,000,000 US$ 4,000,000 3,000,000 2,000,000 1,000.000o 0xw -jl -- l- -r o F 0 0 0 cm 0m o 0 a a Date 9 The difference between the actual cumulative disbursed amount of US$6.49 million by November 19, 1993 and the total loan amount of $6.5 million was due to the effects of the currency putling system. 20 Table 4. Project Costs and Financing A. Allocation of Loan Proceeds (US$ Million) Category Original Actual Disbursements Allocation Consultant Services' 1.380 3.886 Training" 1.268 0 Computer Equipment 1.362 1.279 Computer Software .974 .594 Refunding of PPF .732 .670 Unallocated .780 0 UNDP 0 .07 TOTAL | 6.5! 6.5 Original Allocation Actual Disbursements (Categories as % of Original Loan) (Categories as % of Original Loan) Unaliocated UNDP Consnuant Services 12% 0% 21% UnallocatedComputerSoftware UNDP Reunding o1 PPF Computer Equipment 0% 9% 1% Refunding of PPF Reudngo1 P 20% 10% Training Trainring 0% 20% Computer Software 15% ~Computer 9flfllfll!lll = Equipment 21% C onsuftant services I_ _ __ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 0 10 Services pertain to consultants and consulting finrs. 11 Consists of teaching materials and consulting services. 21 Table 4. Project Costs and Financing (Continued) Project Costs (US$ Million) Components Estimated Costs ActualCost Local Foreign Total Local Foreign Total DOSMIL 0.09 0.0 0.09 1.44 1.44 SELPER 0.37 0.0 0.37 0.50 0.50 CAPPER 1.98 0.02 2.00 0.46 0.46 AUDINT 1.19 0.0 1.19 0.72 0.72 METFIS 2.27 0.0 2.27 0.53 0.53 EQUIPA 1.07 2.16 3.23 4.40 2.31 6.71 PIU (Project Director) 0.01 0.0 0.01 0.30 0.30 Repayment of PPF 0.0 0.73 0.73 0.04 0.04 Management Fees 0.0 0.0 0.0 0.04 0.04 (UNDP/OPS) Total Base Costs 6.99 2.91 9.89 4.40 6.34 10.91 Price Contingencies 0.38 0.40 0.78 0.0 0.0 0.0 TOTAL COSTS 7.37 3.31 10.67 4.40 6.34 10.91 '' C. Project Financing (US$ Million) Sources Planned | percent [Actual 1 percent l__ ___ ___ __ ___ ___ _ |__ (Planned/Total) [ | (Actual/Total) GOVERNMENT 4.17 39 4.4 40 IBRD Expenditures 6.501s 61 6.5 60 TOTAL PROJECT COSTS 10.67 100 10.9 100 12 The "Actual Foreign' column contains all Bank financed expenditures. The "Actual Local" column includes US$4.4 million of cash counterpart expenditures. 13 By November 19, 1993 22 Table 5. Pr.ject Results Fiscal Impact Tax Revenues as Percent of GDP Tax Average 1990 1991 1992 1986-89 Coparticipated 8.3 6.3 9.6 13.4 Taxes Non- Coparticipated 3.2 4.5 5.5 3.5 Taxes Social Security System 3.9 4.4 6.2 7.5 Total Revenues 5.1 15.2 21.4 24.4 Soure: Ministry of Economy, National Directorate of Budgetary Programming 23 Table 6. Use of Bank Resources Stage of Project Cycle Date No. of Day Staff/Weeks S/W per S/W Task Other persons mission at HQ Manager FY89 Preparation Jan23/Feb3 3 11 6.6 LB Preparation Aprl9/May4 3 7 4.2 LB Subtotal 10.8 9.8 FY90 Negotiations July 14-17 2 4 1.6 LB Negotiations Aug 21-25 2 20 8 LB Subtotal 9.6 1.1 I Supervision Oct24-31 1 7 1.4 EGF Supervision Dec 15-22 3 7 4.2 JFC Supervision Apr 2-5 3 5 3 JFC FY91 Supervision Sep 19/Oct2 1 5 1 LJM Supervision Sep26/Oct7 1 14 2.8 GF Supervision Apr 22-26 4 5 4 JVC Supervision May 16 1 2 .4 JVC LJM Supervision Jun 10-12 1 0 LJM FY92 Supervision Jul22/Aug5 5 14 14 LJM Supervision Nov 7-20 1 5 1 NVC Supervision May 10-23 5 7 7 NVC FY93 Supervision Sept 21-24 1 7 1.4 LJM Supervision Mar 22/Apr 2 5 14 1.4 LJM Subtotal 54.2 38.7 FY94 Supervision Ju; 12-20 1 10 2 LJM Supervision Feb 21/Mar4 3 15 9 LJM Subtotal 11 0 PCR Nov 10-24 1 15 3 LJM PCR May 25-29 1 5 1 LJM Subtotal 4 5 Total 89.6 54.6 Grand Total 144.2 EGF: Emilia Gomez (Co-task manager with JVC) GF: Guillermo Fino (Consultant) Allocation of Bank Resources JVC: Jaime Vasquez-Caro (Co-task manager with EGF) Stages as % of total staff time LB: Lucca Barbone (Task Manager) LJM: Luis-Jose Mejla (Task Manager) Evaluation Preparabon 6% 14 Negotiations = 7 7% Supervision 73% 24 Table 7: Status of Covenants Loan Brief Description of Covenant Compliance and Remarks Agreement 3.01b (i) Develop a computerized database for the 2,000 largest Complied. To date the taxpayers and supervise the implementation of collections of the 260,000 adequate measures to prevent tax avoidance and to largest taxpayers are promote tax compliance by such taxpayers through the controlled with this system. use of this database 3.01b (ii) Implement and monitor a new internal auditing system Modified during subsequent and prepare an auditing program for 1990 satisfactory reformulations of the to the Bank. program, these activities are contemplated in the follow-on loan (3460-AR) 3.0 b (iii) Provide computer equipment and training of DGI Complied. 80 multi-user and personnel in its use. 400 personal computers were bought with Bank and counterpart funds and are now fully operational. 3.01 b (iv) Provide semi-annual reports on the Progress of the Basically complied, although Project. late. 3.0lb(v) Appoint and maintain a coordinating commission Complied, although late. headed by a Project Director, and Directors for each project component. 3.02a Enter into contractual arrangements with an agent/s for Complied (UNDP). handling procurement of equipment. No later than September 1 st of each year, provide the Complied late, but Bank an Annual Action Program for the next fiscal supplemented by supervision year. missions. 3.03a (i) Provide not later than March 15, 1990, a list of key Complied, although late. indicators applicable to information available within DGI to be used in improving DGI's internal auditing and management. 25 Table 7: Status of Covenants (Cont.) Loan Brief Description of Covenant Compliance and Remarks Agreement 3.03a (ii) Every October 31 and April 30, until the closing date, Partial compliance. provide a written report with the evaluation of DGI Submission became through such key indicators for the preceding six progressively better and up- months and also for the period between the last report to-date towards end of the and the closing date. project. Upon completion of the auditing program for 1989, The reformulation of the provide to the Bank the results of such auditing project gave more emphasis to program. collections and to internal control; auditing was more properly considered to be addressed later, and will be implemented through 3460- AR. (ii) Prepare and furnish to the Bank audit reports within Complied, although late. four months of the end of each calendar year. Carry out studies to select priority areas for internal Not complied, see 3.04 above. auditing and to evaluate existing tax inspection techniques. 4.0lb Furnish to the Bank all studies' findings and Complied late, but recommendations for comment, taking said comments supplemented by supervision into account at implementation. missions. 4.01b Maintain qualified staff in adequate numbers as Complied counterparts 4.0lb Maintain records and accounts, and have them audited Complied every year by independent auditors acceptable to the Bank. I IMAGING Report No: 16031 Type: PCR