Philippines - Tax Computerization Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No: 20554 IMPLEMENTATION COMPLETION REPORT (36030; 3603A) ON A LOAN IN THE AMOUNT OF US$63 MILLION TO THE REPUBLIC OF THE PHILIPPINES FOR PROJECT ID: P004599 L/C NUMBER: 36030/3603A TAX COMPUTERIZATION PROJECT June 29, 2000 Poverty Reduction and Economic Management (PREM) Sector Unit East Asia and Pacific Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS (Exchange Rate Effective as of June 29, 2000) Currency Unit Philippine Peso (P) P1.00 = US$ 0.0232 US$ 1.00 = P43.17 FISCAL YEAR January I-December 31 ABBREVIATIONS AND ACRONYMS AAB - Affiliated Agent Bank ACOS - Automated Customs Office System ACPS - Andersen Consulting/Philippines Systems ADB - Asian Development Bank ASYCUDA - Automated System for Customs Data Management BIR - Bureau of Internal Revenue BOC - Bureau of Customs BOT - Build Operate Transfer CAS - Country Assistance Strategy COA - Commission on Audit DBM - Department of Budget and Management DBP - Development Bank of the Philippines DoF - Department of Finance EIRR - Economic Internal Rate of Return ILICR - Intensive Learning Implementation Completion Report FIRR - Financial Internal Rate of Return GOCC - Government Owned and Controlled Corporation GOP - Government of the Philippines ICB - International Competitive Bidding IFMS - Integrated Financial Management System IMF - International Monetary Fund ITS - Integrated Tax System IRR - Implementing Rules and Regulations LAN - Local Area Network NEDA - National Economic Development Agency NO - National Office PC - Personal Computer PIU - Project Implementation Unit QAG - Quality Assurance Group RDC - Regional Data Center RDO - Revenue District Office RO - Revenue Office RRO - Regional Revenue Office SGS - Societe Generale de Surveillance SAR - Staff Appraisal Report TCP - Tax Computerization Project TIN - Tax Identification Number UNCTAD - United Nations Conference on Trade and Development VAT - Value-Added Tax WAN - Wide Area Network WTO - World Trade Organization Vice President: Jemal-ud-din Kassum, EAPVP Country Director: Vinay Bhargava, EACPF Sector Director: Homi Kharas, EASPR Task Team Leader/Task Manager: Samia Melhem, CITPO  FOR OFFICL USE ONLY IMPLEMENTATION COMPLETION REPORT THE REPUBLIC OF THE PHILIPPINES TAX COMPUTERIZATION PROJECT Loan 36030/3603A CONTENTS Page No. PREFACE EVALUATION SUMMARY........ 1. Project Data .......................................................I1 2. Principal Performance Ratings .................... ....................................I 3. Assessment of Development Objective and Design, and of Quality at Entry ..... 2 4. Achievement of Objective and Outputs ........................................................ 9 5. Major Factors Affecting Implementation and Outcome .................................... 18 6. Sustainability ....................................................... 23 7. Bank and Borrower Performance ....................................................... 26 8. Lessons Learned ....................................................... 31 9. Partner Comments ....................................................... 34 Annex 1. Key Performance Indicators/Log Frame Matrix .................................... 36 Annex 2. Project Costs and Financing ....................................................... 37 Annex 3. Economic Costs and Benefits ....................................................... 38 Annex 4. Bank Inputs ............. 40 Annex 5. Ratings for Achievement of Objectives/Outputs of Components .......... 41 Annex 6. Ratings of Bank and Borrower Performance ......................................... 42 Annex 7. List of Supporting Documents .................................................... 43 Annex 8. Beneficiary Survey Results ..................................................... 44 Annex 9. Stakeholder Workshop Results .................................................... 45 Appendix 1: TCP's Financial Analysis of Project and related Spreadsheets ......... 59 Appendix 2: BIR's Implementation Completion Report ........................................ 72 Appendix 3: Beneficiary Survey .................................................... 96 Appendix 4: BIR's Lessons Learnt .................................................... 113 Appendix 5: BOC's Implementation Completion Report .................................... 126 Appendix 6: ILICR Workshop Participant List and Agenda ............................... 141 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not be otherwise disclosed without World Bank authorization.   INTENSIVE LEARNING IMPLEMENTATION COMPLETION REPORT THE GOVERNMENT OF PHILIPPINES TAX COMPUTERIZATION PROJECT Loan No. 3603-PH Preface The Philippines Tax Computerization Loan to the Government of the Philippines in the amount of US$63 Million was approved on May 11, 1993 signed on July 15, 1993 and made effective on December 1, 1993. The project was expected to close on June 30, 1999, however it was extended to close six months later, on December 31, 1999. This Intensive Learning Implementation Completion Report was prepared by Samia Melhem, (CITPO - Task Manager for the loan), Dana Weist (EASPR, Senior Economist) and Michael Engelschalk (PRMPS, Tax Administration Thematic Group Coordinator). This is one of the first Intensive Learning ICRs for the East Asia Region. Mr. Vinay Bhargava was the Philippines country director, and Mr. Homi Kharas the sector director for EASPR. Mr. Aloysius Ordu (Portfolio Manager), Mr. Bernard Funck (Country Economist) and Mr. Joven Balbosa (Economist) provided coordination and support from the field, and reviewed and commented on the report. Mr. Rene Ruivivar (OCS) reviewed and commented on the document and provided advice on the Intensive Learning ICR process. Mr. Homi Kharas (EASPR sector director) reviewed and cleared the report. The team also received useful input and advice from Barbara Nunberg (Lead Specialist, EASPR). The project team thanks all involved for their participation and review of the document. The production of the report would not have been possible however without the logistical support of Lily Tsang.   Intensive Learning Implementation Completion Report The Government of Philippines Tax Computerization Project Loan No. 3603-PH Evaluation Summary and Recommendations Project Status Today 1. The Tax Computerization Project was the largest agency modernization effort ever attempted by the Government of the Philippines (GOP). It entailed changing manual, paper-based tax administration to computerized, paperless ones, and training around 8,500 staff on new systems. It improved governmnent services, and affected all taxpayers, civil society and traders. Finally, it improved communication and information exchange in each of the beneficiary agencies. Some of the project's objectives are still unfinished as these involve capacity building, institutional learning, reduction in graft and other longer-term outcomes that have not been fully realized yet, and for which the GOP needs additional assistance to leverage existing accomplishments and newly acquired institutional skills. Project Objectives 2. The Philippines Tax Computerization project was designed in 1991 as the GOP wanted to increase the efficiency of tax administration of its two main revenue agencies: Bureau of Internal Revenue (BIR) and Bureau of Customs (BOC). Based on that objective the project had three components: (1) Computerization of the BIR Tax Administration System (2) Computerization of the BOC Customs Administration System (3) Institutional Strengthening for BIR and BOC staff to operate and sustain the new computer systems. Project's Origin 3. The project, conceived and initiated by the Department of Finance, was designed and gradually prepared by a joint Technical Assistance Team of the International Monetary Fund (IMF) and the World Bank The IMF recommended a 5-point reform program to the BIR and an 11-point reform program to the BOC. These were pre-conditions for the overall tax computerization loan, which was financed by the Bank and made effective on December 1, 1993.  - ii - Quality at Entry 4. Based on what can be concluded today, seven years after the SAR's approval, the project was much more complex and riskier than originally foreseen especially for BIR. The SAR: (a) understated the difficulties of customs software development and maintenance ventures for the BIR, (b) was too optimistic about overcoming broader institutional constraints such as the civil service salaries, (c) under-estimated the change management needed for an agency modernization effort of that scale, and (d) focused too much on Information Technology to the exclusion of tax policy and administration. The Bank's performance in project identification, design and preparation was satisfactory for the BOC component and unsatisactory for BIR. Achievement of Objectives an Outputs 5. Against all odds, the project's outcome was satisfactory for all of the project's three components outlined above. Both BIR and BOC have designed, developed and are currently using a functional tax and customs administration system. Manual procedures were redesigned and automated in a software package that tried to model as closely as possible the processes and workflow in each of BIR and BOC respectively. Some of these achievements were internationally recognized. For instance, the World Customs Organization gave an award to BOC in July 1998 for its Customs Computerization project. 6. BIR developed the "Integrated Tax Systems" (ITS) with Andersen Consulting. The ITS, a sophisticated fourteen module application, has been examined by many Information Technology and Tax Administration experts. One of the latest IMF reports of December 1999 finds it to be a "world-class" tax system, which many developed countries would benefit from. In order to utilize the ITS in all its territory, BIR installed a country wide network linking its Regional Offices, district offices and regional data centers all together. As of today the ITS is being used in 41 District Offices, 19 Regional Offices, three Data Centers and the Large Taxpayers Office. Substantial efficiency gains have been attained since the days of manual operations. However, as of today the computerization has not yet had significant impact on revenues. We expect the latter to happen once the ITS's auditing capabilities are fully used. 7. BIR Ratings: As part of the Intensive Learning ICR process, we looked at, and rated, each one of the smaller components of the projects. We rated as satisfactory: (a) development of the Tax Administration software,  -iii - (b) provision of communication infrastructure, (c) provision of hardware, (d) interface with the commercial Banks, (e) Large taxpayers services, (f) increase in number of registered taxpayers, and (g) improvements in the efficiency of tax administration in general. We rated as unsatisfactory some aspects of the ITS usage, such as some inability to use the system's Audit capabilities, and the fact we could not clearly demonstrate computerization did indeed increase collections. 8. BOC developed the Automated Customs Office System "ACOS" based on UNCTAD's off-the-shelf customs software called ASYCUDA. It did so under a turnkey contract with Unisys and UJNCTAD. BOC deployed the Customs system in 22 ports instead of the original seven planned in the SAR. BOC won a worldwide award from the World's Customs Administration in July 1998. As of today, BOC has attained enormous efficiency gains from both its intemal re-engineering prior to the computerization work and from the new system itself. Cargo clearance time, for instance, was reduced from 10 days to a few hours. BOC had a successful partnership with the private sector as traders, importers, affiliated banks, shippers etc. all got trained and had access to the Customs software and could locate shipping and taxation information whenever required. 9. BOC Ratings: We rated all the BOC components as satisfactory. These include: (a) Development of ACOS in partnership with Unisys and UNCTAD, (b) provision of a nation-wide network, (c) roll-out and usage of the customs system, (d) increase of collections sometimes above targets, (e) overall re-engineering of the customs procedures, and (f) partnerships with the private sector to improve trade and customs activities. 10. Institutional strengthening component was rated satisfactory for both BIR and BOC as an estimated 9,000 staff have been trained in both agencies, a number far above the SAR training estimates. Financial Analysis l1. An in-depth financial analysis of the project was undertaken using the same assumptions and methodologies of the financial analysis of the 1993 SAR, based on BIR  -iv - and BOC collections between 1993 and 2000. The project gave a total incremental collection benefits to the equivalent FIRR of 307%, whereas the SAR's forecasts of revenue increases amounted to a recomputed FIRR of 210%. Analysis also proved that the PhP 2.22 Billion present value of project expenditures could be recovered in just one year by a revenue increase of 1%. Analysis did not include intangible benefits brought in by automation, procedure simplifications, electronic mail, and overall computerization. Major Factors affecting implementation and outcome 12. Factors outside the control of government included: (a) the Asian economic and financial crisis which caused a significant revenue shrinkage for BIR in 1999, (b) the limited national telecommunications infrastructure, especially in the early 1 990s at design time, and (c) rapid changes in office technology equipment, which affected the software development and the rollout of the system to the regional offices. 13. Factors generally under the government control included: (a) inappropriate scope and complexity of the project over a short time span of five years, (b) frequent leadership changes in BIR, BOC and DoF, (c) inter-agency communications and coordination, (d) unstable tax policy environment, (e) lack of counterpart funds, and finally (f) lack of attention to the environment for a larger Public Sector Reform. 14. Factors subject to implementation agency control included: (a) an excellent transparent procurement process for the turnkey contracts, (b) cumbersome administrative processes, (c) lack of a data conversion strategy from legacy systems, and (d) need to enhance BIR and BOC's reputation in the public's eyes Sustainability 15. The project is likely to be sustained as the GOP has seen the benefits of the Tax Computerization as the Philippines is trying to change its government towards a modern,  delivery oriented one. Major obstacles are financial and human resource constraints. Most of the competent BIR and BOC IT staff generally end-up in the private sector where their wages are four to five times higher. The other major hurdle to sustainability are BIR's and BOC's reputation and poor perception by the public; as adverse public opinion may negatively influence the Congress and DBM's budgetary decisions and yearly allocations for the Bureaus. Funds to sustain the computerization are around 15% of the total agency-operating budget according to international standards. Both agencies have transition arrangements and long-term plans to sustain achievements accomplished thus far. Bank Performance 16. The Bank's performance in project identification, preparation and appraisal was rated unsatisfactory as the project focused almost exclusively on information technology instead of a tax administration project supporting capacity building in tax administration. The project also underestimated the difficulties that BIR would encounter in building its own tax system. However at that time there were no "off the shelf' tax system software packages and the Bank had no other choice than to recommend BIR contract out the writing of a customs-based system. Unlike the Thailand Tax Computerization project, the TCP yielded a satisfactory countrywide software system as was intended. Finally - and to the project team's credit - there were imminent signs in 1992 that a Salary Standardization law would be passed. This explains the SAR's optimism conceming the Civil Service laws and the fact that the SAR expected the human resource sustainability problem to be resolved in the short future. 17. The Bank's overall performance in lending and supervision is rated satisfactory as the Bank fought many challenges and succeeded in providing as adequate as possible software development and tax administration advice despite low budgets and lack of available expertise in the subject matter. The Bank can be credited for having designed an excellent procurement process, one of the best managed and supervised in the sector, again unlike similar projects in East Asia. The Bank has played a crucial role in facilitating communication and coordination between all affected agencies. The procurement process of the two turnkey contracts was one of the best managed and supervised in the sector, again, unlike many similar such projects in East Asia. Towards the project's end, the Bank increased its supervision frequency with a newly composed task team including tax administration and public sector experts. The latter was a catalyst to increased DOF involvement and interest, and accelerated the project's disbursements. It also raises GOP awareness and support to many of the modernization issues faced by BIR and BOC, who were pioneers in this field. 18. The TCP was the largest computerization effort ever launched in the GOP. Its effects are felt at all levels: by travelers filling out a simplified Customs entry form, by traders importing their goods, by the average taxpayers going to the banks and paying their taxes and by the large taxpayers who can today pay their taxes on the Internet. Both BIR and BOC delivered well at preparation and implementation stages, and hence their performance is rated as satisfactory.  -vi - Lessons Learned 19. The lessons learned are many for both the Bank and the borrower and are all listed and described in depth in the report (Section 8). Most important lessons for the borrower are the need to: (a) embed Tax Computerization within a broader public sector reform, (b) balance the complexity and the scope of the achievements, (c) select a solution to retain the skilled IT staff, and (d) maintain good communication with oversight agencies (DoF, DBM and NEDA). 20. The lessons leamed in software development and computerization are the: (a) need to have good functional specifications designed by users themselves, (b) essential role of a consistent software roll-out and training strategy, (c) crucial role of the pilot stage in large software development project, and (d) need to get taxpayers educated and committed to the new system. 21. Finally the lessons learned for the Bank were to: (a) avoid software development when an "off the shelf' package exists in the market, (b) simplify and train the borrower on the disbursement process, (c) include a Tax Administration specialist from the start in such a project, and (d) finance a larger, more comprehensive Change Management and training plan for the agencies' staff. Future Operations 22. Despite often adversarial circumstances and high turnover of skilled project staff the project's outcome is satisfactory as two world-class systems have been developed and deployed to administer tax and customs throughout the GOP territory. There is much to be done though in the area of training BIR and BOC staff on the system, notably training BIR staff on using the Audit capabilities of the ITS. BIR staff is still struggling to use the basic modules of the new system and need an additional push to master the Audit module. In turn, the Audit module needs more development to include statistical population information, profiles of taxpayers, expert system capability to identify fraudulent returns, etc.  -vii - 23. BIR, after the project ended, has a shrinking IT budget barely sufficient for minimum maintenance of its facilities. Identified solutions are costly: (a) Outsourcing tax administration IT operations to the private sector, a costly proposition which seems to give BIR the feeling of loss of control of its own system. (b) Creating an independent revenue authority for BIR and BOC but that also requires a total re-organization of both agencies and approvals of the process at the highest level. And there are major risks in this enclave approach. (c) A follow-up operation, for BIR only, presented by DoF in December 1999. 24. The objective of the follow-up operation is to help BIR roll-out ITS to all the district offices, further develop and expand the Audit module, and have all BIR staff use it in an effort to increase overall revenue collections. This program would also be accompanied by a series of anti-corruption training for all relevant BIR staff and possibly could fund a further agency reorganization. We think this would provide BIR the chance to leverage accomplishments made this far, based on an experienced management team who has already done the most difficult part of the journey. This follow-up operation would allow BIR to accomplish the difficult task of using the new system and the benefits of computerization to really increase collections by going after all the non- compliant or under-reporting taxpayers.   Project ID: P004599 Project Name: Philippines Tax Computerization Project. Team Leader: Samia Melhem TL Unit: CITPO ICR Type: Core ICR Report Date: June 29, 2000 1. Project Data Name: Philippines Tax Computerization Project. L/C/TF CPL-36030; SCL- Number: 3603A Country/Department: PHILIPPINES Region: East Asia and Pacific Region Sector/subsector: BF - Public Financial Management KEY DATES Original Revised/Actual PCD: 12/18/90 Effective: 12/01/93 12/01/93 Appraisal: 10/07/92 MTR: Approval: 05/11/93 Closing: 06/30/99 12/31/99 Borrower/Implementing GOVT. OF PHILIPPINES/MIN. OF FIN. Agency: Other Partners: STAFF Current At Appraisal Vice President: Jemal-ud-din Kassum Castillo Madavo Country Manager: Vinay K. Bhargava Kabir Ahmed Sector Manager: Homi Kharas Vineet Nayyar Team Leader at Samia Melhem Claudio Femandez ICR: ICR Primary Samia Melhem, Dana Weist, Author: Michael Engelschalk 2. Principal Performance Ratings (HS=Highly Satisfactory, S-Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N=Negligible) Outcome: S Sustainability: L Institutional Development Impact: SU Bank Performance: S Borrower Performance: S QAG (if available) ICR Quality at Entry: U U Project at Risk at Any Time: Yes  - 2 - 3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: I. Project's Objectives: The main objective of the Philippines Tax Computerization (TCP) Project, according to the 31 March 1993 SAR, was to support improvements in tax administration by providing technical assistance, staff training, software and priority hardware and facilities, to computerize the administration of the Bureau of Customs (BOC) and Bureau of Internal Revenue (BIR). The expected result of establishing and operating a modem, computerized and integrated revenue administration system throughout the Philippines was a more efficient revenue administration, greater revenue collections, and improved access and potential for audit of tax records. Based on that goal, the project had three objectives: 1. Computerization of the BIR Tax Administration System 2. Computerization of the BOC Customs Administration System 3. Institutional Strengthening for BIR and BOC staff to operate and sustain the new computer systems. 11. Background: The project was designed in 1991, when the Government of the Philippines (GOP) was seeking to increase revenue to enhance the country's infrastructure and economic development. The project is consistent with the 1993 CAS objectives of poverty alleviation and improved public services. Taxes are the main source of GOP revenue (accounting for 85 percent of revenue), and BIR.'s and BOC's performance is critical to GOP's revenue management. In 1990, the GOP requested technical and financial assistance from the Asian Development Bank (ADB), the Intemational Monetary Fund (IMF), and the World Bank (WB) to improve the buoyancy of revenue collections and enhance resource mobilization. In 1991, the IMF began providing technical assistance to improve the effectiveness of revenue processes and administration at the BIR and BOC. At that time, both agencies were planning future modernization and computerization activities. The IMF recommended a 5-point reform program to simplify tax administration at the BIR, including the creation of the Large Taxpayer Division, and the replacement of the Tax Account Number (TAN) with the Tax Identification Number (TIN). It assigned tax administration experts from the Fiscal Affairs Department to assist BIR in implementing aspects of the short and long-term computerization strategies. A second IMF mission assisted the BOC and developed an 11-point reform program for customs simplification. Both reform programs were declared as conditions to further Bank/IMF assistance. The ADB provided a short-term, technical assistance grant (US $600,000) to the BIR to prepare an information system master plan and to develop a pilot VAT computerization program. The Bank was asked to provide technical assistance and financing for overall tax computerization. The Bank anticipated that its assistance would extend over 4 to 5 years and that its financing would be approximately US $40 million. In 1991, the Bank secured an  $800,000 PHRD grant to prepare the project. The Bank began to articulate the project in a preparation/pre-appraisal mission in September 1991. The scope of the Project, as described in the 1991 BTOR, was to: "provide for enhanced resource mobilization activities by: (i) simplifying and streamlining tax administration procedures; (ii) instituting a unique tax identification system; (iii) using information technology tools to facilitate and improve tax filing, processing, collection and auditing; and (iv) developing adequate human resources to improve the efficiency and effectiveness of tax administration and sustain its computerization." In early 1992, the three donor agencies (WB, IMF and ADB) envisioned that the project's components would include: (i) the Integrated Tax System for BIR; excluding the VAT "pilot" that would be financed by the ADB grant mentioned above; (ii) the Integrated Tax System for BOC; (iii) an Information System for the Bureau of Treasury (BTr); and (iv) a Management Informnation System for the Department of Finance (DOF). As the project design was developed, components (iii) and (iv) were dropped. The proposed project and donor recommendations were received with considerable enthusiasm and support by BIR, BOC and DOF. In 1993, the TCP was appraised. The SAR was signed on 31 March 1993; the loan was signed on 15 July 1993; and the project became effective on 1 December 1993. III. The Beneficiary Agencies: The project's objectives were focused on modernizing, in parallel, BIR and BOC. The DOF also played a critical role in the project. Background on the agencies involved in the project is included below. Department of Finance (DOF) The project was overseen and supervised by the DOF. Both the BIR and BOC are bureaus within the DOF. The project management committee consisted of a steering group headed by a DOF Undersecretary, the commissioners and assistant commissioners of the two bureaus, as well as the project's national directors and the implementing agency (Crown Agents) management team. BIR BIR is responsible for collecting about 80 percent of GOP revenue. It is headed by the Commissioner of Internal Revenue (equivalent in rank to a DOF Undersecretary), and assisted by two Deputy Commissioners; one for assessment and collection, and one for legal issues and administration. The BIR has a three-tier structure composed of its Headquarters National Office, 19 Regional Offices (ROs), and 115 Revenue District Offices (RDOs). It also has four Regional Data Centers (RDCs). The National Office formulates national policies, planning programs, laws and procedures and directs the entire internal revenue service. It was reorganized in November 1999, at the recommendation of the IMF, to create a fully functional Large Taxpayers Service Division and an Excise Taxpayers Service Division.  - 4 - ROs set BIR's laws, policies, plans and regulations within a region. They coordinate with local government units and regional offices of government agencies setting regional standards. * RDOs plan and develop programs, methods and procedures for the efficient, effective and economical assessment and collection of taxes according to the standards set by the BIR commissioner and the regional directors. BOC BOC is responsible for collecting about 23 percent of GOP revenue (as per DoF's report to the Office of President Office, dated March 1, 2000). It is responsible for facilitating trade, assessing and collecting duties on imported articles (including VAT) and preventing smuggling. BOC is headed by the Comnmissioner of Customs, assisted by five Deputies responsible for administration, collections, assessment, intelligence and enforcement and information systems and technology. The Office of the Commissioner implements and coordinates BOC's policies and programs, which are in turn enforced and adapted by District Offices. The BOC includes a Central office and 14 collection districts: Port of Manila, Manila International Container Port (MICP), the Ninoy Aquino International Airport (NAIA), and the district ports of Sann Fernando, Batangas, Legaspi, Tacloban, Cebu, Iloilo, Surigao, Cagayan de Oro, Zamboanga, Subic and Davao. Each revenue district is authorized to have as many sub- ports as necessary to maximize revenue collection and minimize smuggling and fraud. 3.2 Revised Objective: The project's initial objectives were not changed. Rather, an additional objective - to improve the communications flow among DOF, BIR and BOC -- was introduced in 1996 based on rising communication needs. As a result, a Lotus Notes-based electronic communication system was established for the DOF to communicate with government agencies responsible for tax collection and fiscal management (BIR, BOC, BOT, the Central Bank, Land Registration Authority, etc.). The system - called FINLINK - was not included in the initial project objectives; although early Aide Memoires (dated 1990 and 1991) discussed briefly the need for linking DOF and the BOT to the new computerized tax system. Project funds were re-allocated to purchase, install and develop the FINLINK system. This activity was executed by DoF with the assistance of Project Supervision consultant Crown Agents. Hardware, software and technical support were supplied by BIR's supplier, ACPS, and BOC's non-computer equipment supplier Philips Electronics and Lighting. FINLINK cost approximately US$450,000 Except for this addition, the project's objectives were not revised during the six-year life of the project. 3.3 Original Components. The Tax Computerization Project was funded through a six-year, US $63 million World Bank loan, complemented by US $20 million of GOP counterpart funds, with the broad objective of improving the efficiency and effectiveness of revenue administration. It was organized around six main components: (i) BIR equipment, (ii) BIR software and hardware, (iii) BOC equipment, (iv) BOC Hardware, (v) BOC software (U7NCTAD's ASYCUDA system), and (vi) common consultancies. During the project's life, several reallocations were made to the project  - 5 - components. The initial and final allocations (approved by the Bank in March 2000, and based on the GOP's request to cancel $3 Million from Category 2) are included in the table below. The Project's 6 Component Categories in the loan agreement (Schedule 1) are detailed as follows Component Category Amount of the WB Loan Allocated (Expressed in Dollar equivalent) Original Revised (1) Goods under Part A of the project (BIR) $2,300,000 $2,078,481 (Vehicles, Video conferencing equipment, etc.) (2) Computer hardware $37,600,000 $34,539,657 (Hardware and software licenses purchases, site preparation, and software under Part A.2 of the Project (BIR) and Computerized Tax System software development) (3) Goods under Part B.4 of the project (BOC) $1,400,000 $2,505,000 (Vehicles, Video conferencing equipment, etc.) (4) Computer hardware $14,100,000 $11,885,000 ( Hardware and software licenses purchases, site preparation) under Part B.2 of the Project (BOC) (5) Computer software $2,200,000 $2,565,000 (ASYCUDA licenses and customization costs) under Part B.3 of the Project (BOC) (6) Common Consultancy Services $5,400,000 $6,426,872 (Crown Agent Services, Change Management & Training programs) Total $63,000,000 $60,000,000 3.4 Revised Components: During the project's life, several reallocations were made to the categories. The latest, approved by the Bank on March 2000, and based on the GOP's request to cancel $3 Million from Category 2 (BIR Turnkey system) and US$1,669,670.11 which remained undisbursed in the Loan Account, provides the latest components of the project as follows: Total Amount US$58,330,329.89 Amount canceled effective March 7, 2000 US$ 3,000,000.00 Amount canceled effective May 22, 2000 US$ 1,669,670.11 Original Loan Amount US$63,000,000.00 The BIR turnkey system used US $39.7 Million of the WB funds instead of the original amount of US $42.7 million.  - 6 - For details on the revised categories allocation please refer to the "revised' column in the category allocation table, section 3.3 3.5 Quality at Entry: I. SAR's Risk Assessment (1993): The March 1993 SAR identified the following two potential project risks (SAR, section 4.17): * Procurement Process: The SAR noted that awarding a large contract for computer hardware and software was technically complex, controversial and subject to political pressure. The SAR noted that these risks would be minimized during project implementation by the rigorous qualifications proposed for the two turnkey contracts. * Staff Retention: The SAR noted: "there is still a risk that the BIR and BOC may not be able to retain adequately trained and qualified staff to operate the much-expanded computer system. However this risk is lessened by: (i) the massive training programs included in the project; ii) the supplier initially operating the system and (iii) the potential to extend the assistance if needed." The SAR mentions in section 3.18 " Because of low government salaries a relatively high turnover of systems analysts is expected. Thus Congress is considering a bill to establish "confidential positions" (higher salaries not regulated by Civil Service rules) which would solve the problem. Nevertheless, the project includes provisions for replacing and training highly qualified recent graduates if they agree to stay with the Tax Administration for a couple of years". II. Retrospective Assessment (2000) Based on what can be concluded today, seven years after the SAR's approval, the project was much more complex and riskier than originally foreseen. The project's outcomes depended on several factors and events; some of which were outside of the beneficiary's control and jurisdiction, and most of which were not reflected in the initial design. These include: implementation of the Comprehensive Tax Reform, launched in 1997, as well as a broad range of legislation, policies and institutional improvements, and passage of various laws and legislation by Congress, assuring coordination and communication among many institutions directly involved (DOF, BIR, and BOC) and indirectly involved (Department of Budget and Management for budgetary approvals, the Banking community, BOT, the Judiciary etc.), providing continued, strong leadership by BIR, BOC and DOF, despite frequent changes in senior management, managing the dispersal of numerous project components over many areas,  - 7 - * overcoming the frequent turn-over of Information Technology (IT) staff due to impediments in the civil servant law, * managing an International Competitive Bidding (ICB) procurement process, * overcoming the limited capacity to absorb the project, new office automation, technology, processes and tax software, * changing the culture and behavior of Filipino taxpayers towards voluntary compliance, and - assuring operational budget for BIR and BOC to sustain and improve the developed system and infrastructure after the loan's closing date. Based on our reading of the SAR, it * understated the difficulties of custom software development ventures for BIR, and assumed that well-written procurement specifications would guarantee a good end product. Worldwide experience in systems development (including the WB's own experience) shows this assumption to be naive and simplistic. Software development projects are often highly risky and very difficult to achieve and sustain successfully. It is notable that the BOC's better outcome in comparison to BIR stem in large part from its use of an existing package (AYSCUDA++ developed by UNCTAD) rather than developing a system from scratch. * focused too much on information technology to the exclusion of tax policy and administration. For example, the project design did not require that re-engineering, process innovation and form re-design occur prior to computerization. Nor did it consider the impact of potential tax policy changes on the IT system. The project design also did not include incentives for using the IT systems as management tools. overestimated the GOP's ability to absorb and sustain a complex system. In retrospect, the project could have been sequenced better, with the BIR system development phased in tranches, fewer pilots conducted and the project extended over a longer time frame. Sustainability in terms of continuous leadership over the project life and the need for adequate budget to maintain and improve the system was also not emphasized in the initial design. under-estimated change management needs and the enormous training and facilities budget needed to train 14,000 staff over five years. too optimistic about overcoming broader institutional constraints, especially the limitations associated with non-competitive civil service salaries and standards for technical staff, and rigidities in the budget system. A modification of civil service compensation rules for BIR and BOC IT positions should have been a pre-condition to the loan. In summary, the Bank's performance in project identification, design and preparation (quality at entry) is judged unsatisfactory for the BIR component and satisfactory for the BOC component.  -8- The Bank provided significant and effective inputs into developing the BIR project but its objective was too ambitious and the project was too complex. In some ways, these problems reflect the Bank's limited experience at that point in time (1990) in providing technical assistance for large-scale software development efforts, and off-the-shelf software applications did not exist. If the project were to be appraised today, the risk assessment and project design would be substantially different. Notably, greater emphasis would be placed on an integrated program of policy, administration and technology reforms; the project would be phased in more manageable components; and greater effort would be spent on supervision. Based on the weaknesses noted above, and in light of the fact the BIR component financing constituted two- thirds of the overall project, we find the quality at entry to be unsatisfactory in this project. We rate quality at entry to be satisfactory for the BOC component as the appraisal identified an "off-the-shelf' solution for BOC and hence BOC was spared problems incurred when writing a large application from scratch. BOC could rely on the UNCTAD team of developers for assistance, and focused its efforts on deploying, training and using the Automated Customs Software system in a timely manner. BOC still faced staff retention problems but these were foreseen and discussed at length. III. QAG Review An August 1999 Quality Assurance Group (QAG) study found, in its rapid supervision assessment, that the project's design and planning were weak in their appraisal of the borrower's implementation capacity and unrealistic in the time frame to achieve its objectives. The project was rated poor for institutional capacity building and seen as an information technology project that was detached from any over-arching master plan for public sector reform. The project's supervision was also found unsatisfactory. We concur with some of these findings: the initial design was flawed and quality at entry was unsatisfactory for BIR. However, knowing what we know today, the project might not have been approved in 1993. The overall impact and effect of the project on the borrower was good, its outcomes were satisfactory and the project was well worth the investment according to our financial analysis. The project and its associated learning were critical for the GOP, which is better off today with a world-class tax software system essential to move the country to the next phase of economic development. The project's last supervision mission and the Intensive Learning ICR missions gathered the right mix of inter-disciplinary expertise and were quite successful. Cross-disciplinary teams should be required in the design and supervision of other similar projects. IV. Overall Public Sector Issues The QAG review found that the project was unlinked to broader public-sector reform initiatives, which is accurate. While problems of staff retention and poor compensation were highlighted, other issues related to the need to rationalize the expenditure management system, remove system-wide bottlenecks, strengthen judicial and legal institutions, and promote transparency and accountability were not. Addressing these systemic problems was beyond the scope of the project, and, while it was unintended at the outset, the project did serve as a pilot of sorts in modernizing selected aspects of the public sector and highlighting the need for more  -9- systematic reform. For example, the BOC's approach was seen to be the first - and most critical -- step in an effort to reduce corruption in public administration, and introducing information technology as a means to improve service delivery was piloted in the BOC and BIR. The limitations associated with the Salary Standardization Law were clearly highlighted in this project. Future loans of this kind should be approached in the context of an overall public sector reform, for reform-minded governments ready to take ownership of a complex project. 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: As noted, the project objective was to improve tax administration in the Philippines by providing technical assistance; hardware and software; priority equipment and facilities; and staff training to the Bureau of Internal Revenue (BIR) and the Bureau of Customs (BOC). The project included two large, distinct and parallel activities: (i) creation, development, testing and roll-out of an Integrated Tax System (ITS); for BIR's National Office, 41 Regional District Offices (RDOs), and 19 Regional Offices (ROs); and modernizing and strengthening BIR by training its staff on the newly created ITS and related office automation systems; and (ii) customization of the ASYCUDA Customs software for BOC; rollout to the main office; the Manila and Mactan airports and four main seaports; as well as modernizing and strengthening BOC by training its staff on the ASYCUDA software and other, related office automation tools. As mentioned in Section 3, two important activities were not explicitly mentioned in the project's SAR: (i) the need for each agency; to modernize, reorganize, and reengineer business processes prior to computerization; and (ii) the need to improve the quality and efficiency of communication among BIR, BOC, and DOF. As the project progressed, the GOP realized that these activities were crucial to the success of the project, and modified the project's work plan to include them. Although significant changes were made to the project design, implementation and schedule for BIR, the project objective did not change. In the light of the project outcomes -- notably, two functional computer systems for BIR and BOC, staff trained in their use, and enhanced awareness among the GOP, taxpayers and private sector -- we rate the project as satisfactory in achieving its objective. 4.2 Outputs by components: I. BIR Tax Computerization: The computer system for BIR was developed from scratch as no "off-the-shelf' tax software existed that fit the bureau's requirements. In addition to developing the software, it was necessary to purchase hardware, build communication infrastructure, and modernize buildings to accommodate wiring, network devices and computers. The Tax Computerization Project (TCP) was therefore a formidable project for both the BIR and the Government of the Philippines (GOP)- which had never before implemented a computerization project of that scope and magnitude. Before the TCP, most BIR processes - including collection, consolidation, audit and tax payer records -- were manual. Millions of taxpayer records were in paper form and stored at their respective RDOs. With the ITS, it is estimated that the BIR processes 5 million tax  - 10- payments yearly and receives about 8 million monthly, quarterly and annual tax declarations. Substantial efficiency gains have been achieved and several of the expected project outcomes have been attained. However, as of today, the computerization has not yet had a major impact on revenues. Development of Tax Software Application: The ITS was developed from scratch and has been completed as a comprehensive tax software system. The IMF commented in December 1999 that, "...many countries, including some developed ones, would welcome the opportunity to have such a computing infrastructure and a comprehensive suite of application systems to carry out their tax administration work." The ITS's 14 modules were developed by Andersen Consulting and Philippines System (AC/PS) under a turnkey contract. During the life of the project, 14 Steering committees were organized, staffed by BIR and AC/PS staff in order to develop and rollout the respective ITS modules. ITS has been perceived as too complex and feature-rich for the GOP tax administration process, but this is mostly due to insufficient training. Nonetheless, based on the successful development and implementation, we rate this outcome as satisfactory. Provision of Communication Infrastructure: The project created a private, Wide Area Network (WAN) for the BIR, which is the most modem communication infrastructure for a single agency and largest private WAN extending across the Philippines. At present, 5 Data Centers, 19 Regional Offices and 41 RDOs are computerized and running the ITS. These distributed sites had to be prepared, wired for communication and supported. Networking equipment (switches, hubs, routers, modems, cables, network monitoring and security software, etc.) and civil works had to be procured to make the above possible. We rate this outcome as satisfactory. Provision of Hardware: 20 SUN database servers were deployed; 56 SUN workstations, 850 PC workstations, 10 large system printers; 100 workstation printers; 50 WAN devices and 150 LAN devices were deployed, installed, configured and are running with appropriate software in the sites mentioned above. We rate this outcome as satisfactory. Rolling Out the ITS in BIR Offices: The ITS software system has been installed, configured and tested -- based on its standard architectural requirements (which piece goes on which server, and where) -- in all computerized BIR offices. At present, the roll out includes 5 RDCs, 7 ROs, 41 RDOs and to the National Office where the whole system is controlled via an advanced, nation-wide Network Monitoring, Command and Control Center. These computerized sites cover a territory representing about 90 percent of BIR's total revenue. A chronological review of this rollout reveals that several major difficulties were encountered since 1996: deferrals due to problems experienced during early production of the systems at their first pilot sites; changes in BIR organization, including the creation of new institutions such as the Large Taxpayers and Excise Taxpayers Services; and a very high 80% rejection (suspense) rate of transactions in 1996. However BIR spent the latter three years (1997-2000) correcting and addressing initial rollout problems. Based on current rollout and installation information of ITS, and on the reduced average suspense rate reported by BIR in December 1999 (10%) we rate this outcome as satisfactory.  - 11 - Application and Infrastructure Architecture: Significant strategic changes were adopted over the course of the project. The distributed processing approach was changed from total, RDO-based processing to partial, RDC-based processing. This was achieved by moving computer processing and storage of taxpayer data from ROs to RDCs. Today BIR is contemplating going back to a centralized approach by consolidating all RDCs. Similarly, data entry was transferred from the RDOs to the four RDCs and is now devolved back to ROs. Because of the several design changes and their accompanying changes in the rollout plans such as staffing, logistics, equipment, software design, etc. we rate this component as unsatisfactory. ITS Usage: At present, all computerized sites use between 4 and 7 ITS modules, even though the full ITS software (14 applications) is deployed and available for use in all computerized centers. While several of its unused modules are in place (i.e. loaded, installed and configured at the mainframe computers in the RDCs and the National Office) they are not being used. For example, the auditing functions are unused; hence the BIR is not reaping the benefits of computerization to capture and pursue understated income nor any type of tax evasion. Overall the ITS rollout was too slow and too cautious, even though the ITS usage is now paramount in all computerized RDOs. On the sole reason that BIR is still unable to utilize the whole ITS's suite of application and modules, we rate this outcome, today, as unsatisfactory. Increased Revenue: BIR has not yet succeeded in meeting its revenue targets despite the computerization of its operations. Some of the system's capabilities are not being used. For example, the system allows for instantaneous identification of stop-filers, and these non- compliance cases are followed through by BIR. However the volume of stopfilers and erroneous returns flagged by the system are often times beyond what can be followed through by BIR staff given the current communication methods with taxpayers. Nor is taxpayer information cross- checked with third-party data and BOC VAT data. BIR audit data indicate that only 35,000 taxpayers (out of 10 million) were audited. In general, even though the ITS facilitates the detection of non reported income or non filing of tax returns within the Tax Reconciliation System (TRS), the ITS is not used as a decision making tool to increase the number of audits nor to enhance their targeting, largely due to the lack of an audit strategy. BIR's research (noted in the BIR ICR) shows revenue increases in computerized RDOs (18.6%) as compared to non- computerized RDOs (8.5%). However, it is not obvious that these increases stem directly from the computerization of these RDOs. Consistent with outcomes for the rollout and ITS usage discusses above, we rate this outcome as unsatisfactory. It should be mentioned in this context, however, that the first revenue collection figures released in April 2000 indicate that the BIR exceeded its collection targets by PhP 4.5 billion over the period January through March 2000. Bank Interface: TCP has helped BIR realize considerable internal efficiencies that minimize the cost of tax collection transactions. Tax collection information is encoded in the 3,000 branches of the Accredited Agent Banks by Bank staff, and transmitted electronically to BIR (via RDCs) within 24 hours. The Limited Bank Data Entry (LBDE) application links 1200 affiliated Banks to the ITS. Taxpayers pay their taxes directly at their respective branch, where their returns and payments are encoded by Bank tellers based on the TIN, and sent to the ITS to update the tax records. A taxpayer can pay their taxes at any Bank branch in their RDO, achieving the concept of "file anywhere, pay anywhere." We rate this outcome as satisfactory.  - 12 - Large Taxpayers Service: The creation of the Large Taxpayer Services began slowly, as the reallocation of large taxpayers from RDOs to the LTS took a long time. Even though the IMF's target (a LTS with 2000 large taxpayers accounting for 80 percent of total revenue) has not yet been attained, the LTS's outcome is positive and shows visible improvements in efficiency and collected revenues over the pre-ITS era. One thousand companies are currently registered in the LTS. One of the five RDCs was selected to handle only the large taxpayers. We find this outcome satisfactory. Increased Base of Registered Taxpayers: The number of registered taxpayers in the REG module has expanded to 10 million (including 1.5 million corporations, self-employed and sole proprietors). This is an increase of 5 million taxpayers, and the total registered taxpayers account for one-third of the labor force. We find this outcome satisfactory. Development of the Information Systems Strategic Plan: BIR has a good strategy and planning team to prepare its ISSP and adapting it to its new technology and computing requirements. We rate this outcome as satisfactory. Enhanced Efficiency of Internal Administration: At present, the BIR receives tax payment information from the accredited affiliated banks within 24 hours as electronic data filling is sent directly to RDC from the Banks via the LBDE. Prior to the TCP, this took from 7 to 10 days before the TCP. The BIR now produces collection reports in 3 days at the end of each month as opposed to 20 days before the TCP. Registration of new taxpayers is instantaneous (20 minutes maximum as opposed to taking 2 to 3 weeks pre-TCP). A taxpayer can receive a Tax Clearance Certificate within 15 minutes of registration, versus two days in the previous setup. Introduction of the Intranet, video conferencing and e-mail has tremendously increased communication and outreach between the decentralized BIR branches and organizational units. Finally, reduction of the bank float to 5 days instead of 10 has increased BIR's yearly interest revenue by an estimated PhP 213 million. We find this outcome highly satisfactory. Increase of BIR's Auditing Capability: Whereas recent PREM studies indicate a serious neglect of audit features in newly built tax computerization systems, this was not the case for TCP which has a comprehensive suite of auditing tools, including case monitoring, profiling etc. However, attempts to generate audit cases from the system have been met with resistance by senior management, which suggests that further work is needed on setting selection criteria and on selling the concept to senior management. A March 1999 supervision mission assessed the ITS's audit module and noted the need for further development and tailor-made applications to be added, especially in terms of monitoring macro-economic data and aggregate tax indicators to estimate evasion for sectoral populations. We rate this outcome as unsatisfactory. II. BOC Customs Computerization This was also a complex project for BOC. It had a countrywide scope extending over 14 regions, involving multiple parties (World Bank, Crown Agents, Unisys, UNCTAD, DAP, DoF) and including cross-disciplinary areas such as facilities planning, civil works, software and hardware rollout, training, change management, etc. However, as noted previously, BOC did not have to develop customs application software, as a world-class application existed (ASYCUDA, written by UNCTAD). ASYCUDA did need to be substantially adapted to the GOP's customs  - 13 - procedures. While the BOC needed to acquire hardware and build communication infrastructure, it benefited from having its major sites (the Port of Manila, NAIA Airport, etc.) in the most optimum locations as far as telecommunication services were concerned. However for its ports and sites outside of the metropolitan area, BOC did face the same communications and facilities management challenges as the BIR. Development of the Automated Customs System (ACOS): Computerization of the BOC was based on the development of the Automated Customs Systems (ACOS) under a turnkey contract awarded to Unisys in June 1994. The initial contract price was US$8.2 million and was increased to US$11.9 million. The ASYCUDA-+ Customs software is at the core of the ACOS. A resident UNCTAD team was contracted to customize ASYCUDA++ for the BOC under a US$2.5 Million contract. At present, customs software has been adapted to the GOP's customs policies, procedures and laws and is operational in 21 sites. ACOS also includes an automatic payment program for importers' duty taxes; and a data-warehousing component used for intelligence and information gathering. The GOP received a World-Class Customs Award for its modernized customs system in July 1998 by the World Customs Organization. In December 1999, ASYCUDA was upgraded to its latest version, ASYCUDA 1.15. We rate this component as highly satisfactory. Provision of Communication Infrastructure: The BOC owns and operates a wide area network (WAN) that spans virtually the entire GOP territory, connecting the central BOC headquarters in Manila and all other sites via a frame relay network. We rate this component as satisfactory. Provision of Hardware Computing Platform: More than 600 Unisys PCs and 40 servers (including back-up and Email servers) have been purchased to operate ACOS. Other non- computing equipment such as printers, audio video conferencing equipment for training and conferences, generators, air conditioners; etc. were purchased to complement the BOC modernization program. Despite the increased number of rollout sites, the BOC purchased and upgraded its equipment in all its locations in 1999. We rate this component as highly satisfactory. ACOS RollOut: By 1999, ACOS was successfully rolled out to 21 ports instead of the 7 originally planned sites in the SAR. The additional 14 out ports were included in the roll out to extend the benefits occurring in the other operational ports without incurring major additional costs. We rate this component as highly satisfactory. ACOS Usage: ACOS has been in use for a few years now by trained BOC staff. There is much public awareness and recognition of the system as its targeted audience - traders -- benefited from the accompanying simplification of import and export processes, customs clearance, goods declaration, etc. We rate this component as satisfactory. Revenue Increase: BOC has consistently achieved and indeed exceeded its collection targets in 1993, 1995, 1996 and 1999. We rate this outcome as satisfactory. Trade Facilitation: The BOC TCP created a more favorable business environment for the private sector, which also generated additional revenue. The project accelerated the processing  - 14 - of goods and cargoes in Customs, which helped brokers, tax authorities and importers. The project was also a catalyst for successful private/public partnership as most data entry in the ports is outsourced to a private data entry service provider. We rate this outcome as highly satisfactory. Re-engineering of BOC Procedures and Processes: This work has occurred from the start of the project, based on the IMF's 11-point reform program and prior to the customization of ASYCUDA ++. Outdated, inefficient, paper-based, manual processes were canceled and the customs activities (release of imported cargo, selection of goods for inspection, inspection, reporting, clearances, computation and collection of duties and taxes on imports) were redesigned and then automated in the ACOS. We rate this outcome as highly satisfactory. Site Preparation and Facilities: Five sites were constructed from scratch (the Customs Computer Center, POM, MICP, NAIA, and Batangas), all other sites were renovated (including training center in the CRIC building, Cebu and Legaspi). We rate this outcome as satisfactory. III. Institutional Strengthening and Capacity Building Institutional Strengthening for BIR: Satisfactory Several types of training and change management programs were accomplished in BIR. These revolved around two different target audiences: rank and file staff and management, and inforrnation technology staff. Training the BIR rank and file staff, and management included (i) office automation (as most of these staff had never used a computer before the project), (ii) using the ITS and its new processes, and (iii) communications, values and ethics. IT training involved technical areas such as telecommunications, network administration, database development, database administration, office automation support, hardware support, and using and supporting the ITS itself (help desk function). ITS Support and Development Training: BIR IT teams received in-depth technical training on all aspects of the new technology introduced by ITS (LAN management, database administration, facilities management, software development, etc.) We rate this outcome as satisfactory. Change Management: Over 6,000 BIR staff (2,000 staff were specified in the SAR) participated in a change management training program where they learned new BIR processes and how to apply these in the ITS. This was the largest technical and user training program in BIR's history, and was accomplished over five years and across the GOP's territory. Some of the communication, leadership, ethics and values training courses were funded by GTZ, who assisted BIR's efforts to change its image and improve its internal values system. The change management program for the BIR rank and file was insufficient because of low funding for this activity. Based on all above we rate this outcome as unsatisfactory. Taxpayers' Education: The Taxpayer Assistance Service (TAS) was formed in 1998 after considerable delay. It is responsible for marketing, and educating citizens on BIR's mission and implementation of the taxpayer registration program. Ideally it should have been established in 1993 or 1994 to achieve its intended objective. However, based on the vast number of  - 15 - educational programs and out- reach seminars implemented by BIR in the last two years we find this issue satisfactory. Increasing Tax Administration Transparency: The project today provides better taxpayer service through the help desk kiosks and brochures at the affiliated agent banks, and specially created new Taxpayer Assistance Units for regular taxpayers and the LTD. BIR is also working on improving its web site. It intends to allow electronic filing and payments over the Internet, once legal barriers are overcome. Forms are generated and downloaded from the BIR's web site (www.bir.gov.ph); and taxpayers can obtain relevant tax information through the information kiosks and the web page. Today, most information is processed electronically with minimum human intervention; and collection reconciliation is done on-line by consolidating taxpayers' returns and the withholding agent's collection. We rate this outcome as satisfactory. b) Institutional Strengthening for BOC: Satisfactory Most of BOC staff also participated in a large-scale change management program and training on the ACOS. Training was developed by Development Academy of Philippines (DAP), then Unisys and in parallel SGS (Societe Generale de Surveillance) conducted Customs Clearance Training, Border Patrol Training, and much training related to the Customs Process. Information Technology, Support and Development Training: Over 2500 people (650 people were targeted in the SAR) were trained on the ASYCUDA++ software and on office automation, as well as some non-technical courses related to change management. All of BOC's IT staff received training with 17 courses offered 28 times. We find this outcome satisfactory. Change Management: DAP, UNCTAD and Unisys were the major providers of the change management training program. This program consisted of workshops, seminars, awareness sessions with traders and importers; and the transacting public at large. BOC produced 34 non-technical courses that were run 747 times. Two regional training centers were established in Cebu and Davao, including office automation and video conferencing equipment. The Manila training center was used for most of the metro Manila IT and change management training. We find this outcome satisfactory. Trader and Public Education: As part of the change management program outlined above, training and public awareness programs were also held for the transacting public, other agencies, stakeholders. We find this outcome highly satisfactory. Increasing Customs Administration Transparency: The BOC today provides better services to traders and allows electronic tracking of a shipment, its value, location; etc. The computer randomly assigns an inspector for suspect cargoes so there is less room for human intervention or corruption. We rate this component as satisfactory. 4.3 Net Present Value/Economic rate of return: Financial Analysis of project: The Staff Appraisal Report (SAR) prepared on March 31, 1993 for the Philippines Tax Computerization Project indicated attractive benefits on project returns and recovery of project costs using the Net Present Value (NPV), Financial Internal Rate  - 16 - of Retum (FIRR), and Benefit-Cost Ratio evaluations. There were likewise other non-financial benefits cited in the SAR. The project closed last December 31, 1999 and a financial evaluation was done in April 2000 as part of the Intensive Learning ICR by a team of Financial Analysts. The complete analysis and the collection data used are all listed in the annexes attached to the ICR and listed below. Objective: The actual benefits from the project were evaluated and examined in the light of the benefits cited in the SAR, with appropriate adjustments to be comparable, and of the economic conditions existing during project implementation. Specifically, the objective of the financial analysis was to evaluate benefits from the project using Annex 12 as well as paragraphs. 4.15 and 4.16 of the SAR. Coverage: The evaluation covered the period 1993- 2000 with the Bureau of Intemal Revenue (BIR) and the Bureau of Customs (BOC) as its major components for revenue collection agencies. It also covered the regional computerization/large taxpayers where data was available. Methodology: The evaluation mainly used the Discounted Cash Flow (DCF) evaluation tools and the Benefit-Cost Ratio based on DCF results. The DCF tools used were the Net Present Value (NPV) and the Financial Internal Rate of Return (FIRR). The framework of analysis used is that of the Annex 12 of the SAR. The figures in the SAR were deflated by the reduction in the GDP growth rates so as to be comparable with the actual growth rates used in the evaluation. Since rollouts only started in 1996, the project benefits started to be computed only at that year - which represents a 3-year delay on incremental benefits. As such the SAR figure on 3-year delay has been used. Overall observation: The economic benefits and incremental revenue could really not be directly identified with the project especially in times of economic difficulties of the country. Inputting, restating and deflating figures could have resulted to unfavorable results which might not even be attributable to the project, so this was minimized as much as possible in the FA assumptions. The non-financial benefits should be mostly emphasized at this point, rather than the purely financial ones. Annexes: Please refers to attached annexes as additional sources of information: 1 - Incremental Project Costs 2 - Consolidated Revenues 2-a Bureau of Internal Revenue - Schedule of Collection Target vs. Actual 2-b Bureau of Customs - Schedule of Collection Target vs. Actual 3 - Financial Rate of Return Calculations 4 - Assumptions for the Project's Financial Rate of Return 4.4 Financial rate of return: As can be seen in detail in the annexes, the benefits from the project were financial and non-financial. The incremental benefits could only be computed with a 3 year delay starting 1996 due to the fact that only in 1996 did the roll outs started. On the financial benefits, the  - 17- project gave a total incremental collection benefits to the government of about pH 2 trillion from 1996 to 2000 or an equivalent FIRR of 307 % from a total cost of PhP 3.35 billion. An attractive return indeed. The Benefit-Cost Ratio based on present values is about 9, indicating a highly beneficial undertaking. The NPV of the project's incremental revenues using interest rates of 10 % to 20 % averages to Php 16 billion. Comparing this to the project's average cost present value of Php 2 billion, the project earns more than what it costs by about Php 14 billion. The foregoing benefits were obtained despite the reduction of almost half of its assumed GDP growth rates from 7% to an average of 3.8% as a consequence of a sluggish economy mainly brought about by the Asian crisis. Comparison to the SAR could not be straight away be made on the peso figures since the GDP growth rates have been significantly different. Thus based on a revised SAR, imputing the GDP growth changes, the project attained 90% of the SAR incremental collection targets under a 3 year delayed benefits comparison. The FIRR of the project of 307 % is higher than the recomputed SAR FIRR of 210 % mainly because of the heavier initial cash outflow under the SAR. The benefit cost ratio and NPV of the actual figures and the SAR are not comparable due to the significant reduction in the GDP and lower collection. Despite all the reduction in GDP, the project managed to retain its attractiveness of return and cost recovery as a worthwhile project to sustain. The non-financial benefits are from the project were easily felt and clearly noticeable. At the BR, the computerization of key business processes such as receipts and reporting of tax payments, among others greatly enhanced employees' productivity. At the BOC, computerization has improved cargo release time, the selection and extent of physical and documentary inspection, and the banks' reporting and transmittal of importer's payment. Improved intemal and extemal communication has also greatly enhanced the efficiency of both agencies. However, the impact on manpower productivity in terms of head counts and time charges has not yet been established. Despite the deviations in the implementation targets and the generally low tax collection posted due to economic slowdown and other uncontrollable factors, the financial benefits during the project period is still highly satisfactory. It also proved that the P2.22 billion present value of project expenditures could be recovered in just one year by an increase in revenue of less than 1%. The project also showed an extremely high financial internal rate of return. This analysis did not include the intangible benefits that are difficult to quantify, more particularly the economic benefits brought about by automation and work simplification. In spite of the economic crisis and other difficulties encountered during its implementation, the TCP achieved reasonable financial benefits, although it did not meet the target set in the SAR. The benefit-to-cost ratio posted a respectable 39 to 43 times at discount rates ranging from 10% to 20%. Other financial indicators such as internal rate of return and net present values also showed high marks. This financial evaluation did not consider the non-quantifiable economic and social benefits that were inevitably brought forth by the project. In like manner, revenue losses due to deficiencies and inefficiencies of the manual system were not quantified. Considering the immediate and long-term impact of the project not only to the agencies concerned but also to the Filipino taxpayers as a whole, the overall benefit of the TCP more than compensates its investnent cost.  - 18 - It is important, however, that the full implementation and operation of the computerized system be sustained if the government wants to maximize the benefits on new technology. The planned linkage with other key government agencies is also worth pursuing. Proper and consistent system maintenance, both hardware and software must also be undertaken. The cooperation and support of the tax-paying public are also greatly needed to ensure continuing success of the computerization project. 4.5 Institutional development impact. In addition to outcomes outlined in section 4.2, other unanticipated outcomes arose from the TCP: Improved Fiscal Decision making and management. Based on new, detailed collections data, DOF developed a comprehensive and functional database to analyze collection problems (i.e., how were import patterns changing and where was leakage occurring) and to formulate tax policies Enhanced Government Expenditure and Cash Management. During the GOP's deficit period, it managed its cash flow on a day-to-day basis using FINLINK to assess collections and to communicate with DBM Engaged the Private Sector and enhanced their willingness to pay for selected public services (i.e., the super green lane in the BOC). Enhanced BOC's Reputation, especially in lessening bureaucratic graft and corruption - as viewed by the private sector, public at large, Congress and international observers. Experience in Agency Modernization and Electronic Service Delivery: BIR and BOC's experience has created awareness and sensitivity to large-scale modernization issue. The GOP can benefit from the TCP experience in any future E-govemment plans or strategies. Increased efficiency of communication: BIR, BOC and the World Bank greatly increased the efficiency and speed of project management by exchanging information via email as opposed to letters or faxes. Disbursement information, document reviews, procurement information etc. were all exchanged by email which considerable speeded-up approval cycles. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: The below-mentioned factors affected both BIR and BOC, unless specifically noted that only one agency was affected. Economic Crisis: the Asian economic and financial crisis caused a temporary shrinkage in all economic sectors and rising unemployment. The downturn in economic activity is a key explanation for revenue shortfalls in 1999, and BIR's inability to meet its annual revenue target  - 19 - of PhP 380 billion. Deterioration in the peso's exchange rate as a result of the crisis also had a significant impact on the project. When the loan was signed (1993), the exchange rate was 23 pesos per US dollar. In 1999, the exchange rate was 44 pesos per US dollar. Most loan disbursements occurred from 1997 to 1999, largely because of the staging of deliverables toward the end of the project. This caused enormous problems for BIR and BOC in obtaining the needed annual budget appropriation from Congress and the Department of Budget and Management (DBM) for the counterpart contribution to the project. Negotiating these budget allocations distracted many TCP executives from monitoring and supporting the project's implementation. Limited Telecommunication Infrastructure: BIR's ITS architecture was implemented based on a three-tiered, Regional Data Center (RDC) scheme. The original design was based on a two-tier architecture, with a National Office-centric taxpayer database replicated and updated at each RDO. The RDC-based design was introduced in 1995, when BIR realized that weaknesses in the telecommunication infrastructure in the Philippines would impede transmission of large files from the NO to RDOs. Hence the five RDCs were created with more powerful telecommunication infrastructure, and they were slated to be the "hub" for the main database. The ITS would have been easier to design, develop, maintain and use by taxpayers and BIR staff if the system were developed in the two-tiered structure that was originally planned. Rapid Changes in Office Technology Equipment: The project and the major procurement specification began in 1992 in an environment when the Disk Operating System (DOS) was widely used especially in developing countries. Between 1992 and 1999, hardware and software markets experienced rapid changes and life cycles. Hardware prices for servers, computers and networking equipment began declining rapidly around 1995, as computers became real commodities. At the same time, software shifted to Windows applications. Although the rapid change in technology is well known and will affect any project with a significant office automation component, BIR suffered at the end of the project from obsolete equipment that was purchased in 1993 and 1994; old software versions without provision for license upgrades, inadequate storage capacity for transaction archiving, and neither project funds nor as agency appropriations to replace obsolete equipment or software. BOC did not face that problem as most servers were upgraded and enhanced in 1999, giving them an additional life span usage until 2002. 5.2 Factors generally subject to government control: Inappropriate Scope and Complexity of the Project: As noted previously, the project was too complex to be implemented in only five years, and inappropriately narrow in excluding important issues of tax policy and administration. A better phasing of project components would have met the development objectives of the project and recognized the need to build capacity and manage change in the BIR and BOC. The project should have included a major business process re-engineering component for BIR's organization, processes and communication with other agencies. Many existing processes were automated without regard to whether they were efficient ones for a modem organization. BIR lost a good opportunity to streamline its processes, eliminate unwanted steps and procedures, and use the TCP as an application with built-in workflow mapped to the BIR's new procedures. Along the same line, fundamental simplification of tax filing and payment processes should have occurred before automation.  - 20 - Rollout Strategies: BIR's first ITS pilots were rolled-out in 6 RDOs, instead of just the Large Taxpayer Service as recommended by the Bank, IMF and consultants. Initial problems - e.g., the predictably high suspense rate for payments and returns -- were much more difficult to correct in six pilot sites than one, and significantly greater resources were required (e.g., six roll- out assistance teams). We contrast this with BOC's piloting of ACOS in the co-located Port of Manila: one pilot site within walking distance enhanced the efficiency of the pilot and reduced the "test and correct" cycle needed for pilot software to mature into production software. In addition, the initial strategy to roll out 14 modules was revised to concentrate on six core modules (registration, collections, returns, stopfilers, and bank interface) in the largest revenue regions (Makati, Manila, Quezon City and Cebu). The constantly changing roll-out strategy disrupted the project: the composition of roll-out assistance teams continually changed, scarce human resources were reshuffled, and attention to test and improve ITS's components diminished. Mismanagement of this roll-out process has significant costs: BIR can no longer hold the contractor accountable for bugs or development flaws and it must pay separately for any software change. Frequent Leadership Changes: By mid-April 2000, the project had witnessed several leadership changes: 6 Secretaries of Finance, 5 Under Secretaries in charge of the TCP, 4 BIR Commissioners, and 4 BOC Commissioners. Despite on-going GOP commitment to the TCP, each new management team needed time to understand the project and had their own agenda. These changes delayed some decisions and some~times led to totally different implementation strategies. Inter-Agency Coordination and Communication: The relationship between DOF and BIR was not clearly defined at the outset of the project, and DOF's role was envisioned only with regard to project supervision. This may stem from the initial design of a tax computerization project rather than one promoting better tax policy and administration. DOF appointed one (sometimes two) of its four Under Secretaries as an oversight manager for the TCP, which is organizationally equivalent to the BIR Commissioner. This sometimes caused friction in reporting and communication. The project would have faired better if the Secretary of Finance had defined a clear line of reporting for BIR's Commissioner. Similar lack of clarity characterized "who did what" at DOF regarding the TCP: at the end of the project, two Under Secretaries were in charge with different teams of advisors, countersignatures took a lengthy amount of time, and resentment grew between BIR and DOF. Efficient revenue collection requires the ongoing exchange of data between tax and customs administration, which can be facilitated by computerized systems. In the Philippines some progress has been made with FINLINK and the use of the same Taxpayer Identification Number by both BIR and BOC. However the exchange of information among agencies is still underdeveloped, and opportunities for using the TCP as a means to enhance overall fiscal decision making may be lost if no communication strategy is agreed upon amongst the different agencies. Indifferent Middle Management: BIR's top management included many dedicated managers and project champions. However, this commitment and support was not characteristic of all of BIR's middle management, whose lack of support hindered the achievement of some objectives and promoted project resistance among their own staff. Consequently, many unforeseen and avoidable problems occurred at the RDO level.  - 21 - Unstable Tax Policy Environment: The project witnessed several changes that were unanticipated as BIR underwent several policy and organizational changes. One of the most important was the Tax Reform Act of 1997, popularly called the Comprehensive Tax Reform Program. Every change in tax policy and administration (and related processes) directly affected the development of the ITS, since the software system had to emulate all BIR processes, legislation, policies, workflows and procedures. These changes delayed software development and rollout; and hence the timely usage of the ITS. Implementation of the CTRP distracted DOF management fromn daily monitoring of the project, and caused resentment among BIR staff who felt excluded from the tax policy decisions under discussion. This caused a widening gap between DOF and BIR until the end of the project. In addition, the frequent declaration and usage of tax amnesties contributed to the low degree of voluntary tax compliance in the Philippines. A continuing expectation of immunity from prosecution for tax evasion and non- timely payment of taxes seriously impedes effective tax collection. Lack of Attention to Environment for Public Sector Reform: The project was conceived, designed and planned as a tax computerization project. However, tax administration does not operate in a vacuum and depends significantly on the overall environment in which the administration operates, including issues such as civil service rules, predictability of budget funding, or the quality of the judiciary. To support and enhance efficiency gains in tax administration, additional public sector reforms will be necessary to professionalize the civil service (and overcome the difficulty in attracting and retaining qualified IT staff), enhance the legal basis for enforcement, and work toward stable funding of TCP activities. The need for civil service reform was raised in the SAR but it was assumed that the GOP would undergo a civil servant reform or other legislative changes (confidential positions bill) that would harmonize private and public sector wages. This issue is still unresolved as both BIR and BOC depend heavily on qualified IT staff to maintain the facilities and databases, produce management reports, enhance the system and adapt it new requirements, etc. Lack of Counterpart Funds: This budgetary issue became an overwhelming and distracting issue in early 1998 as the GOP, in response to the regional economic crisis, slashed non-personnel budgets by 25 percent across all agencies. Both BIR and BOC needed funding for their facilities maintenance, telecommunication costs, contractors' salaries, as well as to pay for the project's locally purchased goods and services. The TCP is based on a computing infrastructure that requires equipment upgrades. renewed software licenses, and increased server capacity for form processing and storage. Neither BIR nor BOC has succeeded in securing sufficient budget to improve nor sustain its operations. This will become a critical determinant of future sustainability, as loan proceeds are no longer available to fund these activities.  - 22 - 5.3 Factors generally subject to implementing agency control: Loss in Implementation Consulting Assistance: The contract with Crown Agents ended in November 1998 and was not renewed. This exacerbated inter-agency communication problems. During their contract, Crown Agents facilitated coordination among BIR, BOC and DOF. Crown Agents also did a good job in including taxpayers, accountants, users from BIRs and representatives from the Banking sector to participate in the design of the ITS; subsequent stakeholder input might have enhanced project outcomes. Cumbersome Administrative Procedures: In 1999 a new requirement of DoF countersigning of withdrawal application; which considerably delayed the loan disbursement process as withdrawal applications would sometimes sit for weeks at DoF. Multiple Change Notices with Contractors: Change notices are inevitable parts of any contract, however they consume managerial resources and require top management approvals from BIR, BOC, DOF and the World Bank, leading to considerable implementation delays in this project. For BIR, the ACPS contract grew from about US$32.8 million to about US$ 37.5 million, by about 14.3 percent of the original contract price. The BIR project had over 300 change notices. For BOC, the Unisys contract grew from US$8.2 million to US$11.6 million, equivalent to almost 42 percent of the original contract price. The BOC project had almost 300 change request notices. Transparent Procurement Process: The use of Crown Agents, a UK-based firm specializing in Public Sector Management, as the Project's procurement consultant allowed a transparent, efficient and high-quality procurement process that was lauded unanimously by all staff involved in the project. Large Informatics projects often experience major delays, misprocurements and disputes with bidders and vendors. This was not the case for TCP Need for Data Conversion Strategy: The manual systems used in BIR prior to ITS contained debt information that was not transferred to the ITS. When the ITS was installed in the Manila and Cebu offices, data about existing debt was not transferred to the new system so all taxpayers started with a zero balance. This is one of the many issues encountered during the migration from the legacy systems to the new ones, and could have been avoided if the implementation team had properly planned a data conversion strategy. Need to Enhance BIR and BOC Reputation: BIR and BOC do not generally have a reputation for professional administration, nor are taxpayers understood to pay their entire share of taxes and fees. The TCP-- with its potential to reduce opportunities for corruption and improve the detection of stop-filers and non-filers -- is an important element in initiatives to improve BIR's and BOC's image. A combination of a strong anti-corruption initiatives, efficiency gains in tracking tax evasion, and a public relations campaign are necessary elements of a strategy to improve the image of BIR and BOC. For BIR and BOC to reap the full benefits of the new systems, they must use the system to identify evaders and stop filers. Only consistent pursuit of fraud or evaders and the application of fair sanctions would exonerate the BIR and BOC in the eyes of their skeptical public. While some success has been achieved regarding the public's perception of BOC (especially among the community of traders), BIR will need to take  - 23 - further action and develop an effective public relations campaign. Voluntary compliance will be difficult to achieve as long as the image of BIR does not improve considerably. 5.4 Costs andfinancing. Cost changes were not significant as project costs were incurred as planned, and cost variations were small. However, changes in project scope and design (especially with regard to the creation of the RDCs, and the many change notices) took a toll on the project, especially for the 30% local counterpart funding required. BOC project savings were realized that allowed UNCTAD personnel to remain for another year to assist the BOC, free of charge. Disbursement delays were significant. The ICR quantifies these delays by comparing the number of years estimated in the SAR from loan signing to completion (or final disbursement for adjustment loans) with the actual performance. The estimated disbursements in the SAR were unrealistic given that service contracts were such that the payments to suppliers were programmed to be done in the last two years of the project. Also, unrealistic implementation schedule, incomplete project preparation, unexpected technical difficulties, changes in project scope, delays in selecting staff/contractors/consultants, delays in receiving counterpart funds all contributed to a slower than expected disbursement flow. In December 1999, BIR cancelled US$3 million from the total loan. The total cost of BIR's component (including counterpart funding) is PhP 2.1 Billion. The total cost of the BOC component (including counterpart funding) is PhP 900 Million. Loan reallocations occurred five times during the life of the project: in December 1996, in July 1997, in July 1998, in November 1999 based, and finally in March 2000 (reflecting the cancellation of US$4,669,670.1 lmillion from the total loan.) Such reallocations required BIR, BOC, DOF and World Bank approval and no objection, leading to project delays, sometimes as long as 6 months. For additional information on economic costs and benefits or disbursements, please refer to Annex 2 and 3. For Financial Rate of Return and Net Present Value information please refer to Section 4.3 and 4.4 as well as the Financial Analysis Annexes 6. Sustainability 6.1 Rationale for sustainability rating: The project is probably going to be sustained, as government counterparts have seen the benefits of the TCP and enhancing revenue collection continues to be a critical objective for the GOP. The GOP's commitment to the TCP is strong for both BIR and BOC, but obstacles such as financial and human resource constraints and poor bureau reputations must be overcome to assure sustainability. Continued leadership by the DOF and TCP senior managers will also be required for sustainability. Financial constraints reflecting shrinking bureau budgets and an annual budgeting framework. In the future, BIR and BOC must finance all of their future IT operations from GOP  - 24 - appropriations, which would require stable, multi-year funding commitments. For most advanced countries, the budget share allocated to computerization and information technology (excluding salaries) is around 15 to 20 percent, depending on the level of software development needed. Because of the effects of the economic crisis, BIR's budget has been declining (from PhP 790 million to PhP 389 million). In addition, the current, annual budget cycle runs counter to the multi-year cycles of IT projects. Insufficient funds to maintain the project and to enable nationwide rollout of ITS would substantially diminish outcomes. Human resource constraints have been present since the project's inception. BIR and BOC must hire and retain highly qualified IT staff. If no change occurs to the Bureaus' compensation structure, and if no exception is made to the Salary Standardization Law, BIR and BOC will have to outsource their computer operations. Senior managers have begun discussions with DBM and other officials about the possibility of outsourcing these functions. BIR and BOC are perceived poorly by the public. This is an obstacle to sustainability as adverse public opinion negatively affects DBM's decisions on budget allocations for the bureaus, and may antagonize Congress against the agencies. There is a need to scale up public awareness of progress accomplished, increased transparency, improved accountability, etc. The Tax Computerization Bill to be enacted in 2000 should provide more stable funding to the BIR since it assures that BIR will receives 0.25 percent of its previous year's collection in 2001 and 2002 and 0.20 percent of collections thereafter. Other future enhancements to the TCP include donor-assisted studies of business process reengineering and tax policy. Electronic filing and taxpayer service capacities will be available at the RDO level. Linkages to other government institutions such as BOC, Land Transportation Office, Land Registration Authority and local government units are expected to be completed by 2001. These linkages will help identify those under-declaring their taxes and non-filers. In 2001, the Computerization Bill is expected to be approved, which will expand taxpayer assistance facilities (i.e., customer call center and electronic filing capabilities). The TCP is expected to reap its full benefits once ITS rollout has been completed, i.e. at the end of 2002. One of the options to promote greater sustainability of reforms would be to establish an independent revenue authority that would combine BIR and BOC. International experience has shown that, when well designed, these authorities can enhance revenue collections by granting greater flexibility in budgeting, human resource management and public procurement to the authority in return for better revenue performance. The feasibility of this option would need further consideration by the GOP. However, we already received some feedback from the borrower, who thinks this option is not the right one for the GOP. Finally, consistent leadership and oversight of the project, regardless of cabinet changes will help stabilize the project and will speed-up implementation of existing strategies. 6.2 Transition arrangement to regular operations: Both bureaus have spent considerable time assessing their present state and planning for the future, culminating in transitional arrangements and long-term plans. For BIR, the short-term  - 25 - (2000-2001) action plan to assure the success and sustainability of ITS as a tool for BIR to enhance its revenue collections, as agreed with the IMF, includes the following critical activities: * deployment of the full ITS in the LTD and a key regional district office (RDO) such as Makati * usage of audit and case monitoring for the LTD and the new selected RDO * roll-out the six basic ITS modules to all RDOs * devolve encoding back to the RDOs * start a feasibility study on ITS outsourcing * initiate electronic filing and paying activities, pre-print tax payment vouchers, etc. * use third-party information to verify return and improve compliance * maintain and service the existing setup * train staff on ITS development and support * start a three-year rolling hardware replacement program BIR's longer-term (2002-2003) plan includes the following activities: * roll-out the full ITS to all RDOs * reduce administrative costs by 20 percent - establish a new IT strategy * outsource ITS * centralize all RDC processing in the National Office * improve the audit module and transform it into an expert system to match cases against tax behavior profiles * maintain and sustain the existing setup * enhance the IAS system BIR's rollout plan and medium-term strategies are listed as an attachment in this report. For the BOC, the short-term plan is to:  - 26 - * establish the super green lane * implement the data warehousing inherited from SGS * increase its analysis * connect the data warehouse with the ACOS * continue change management program for BOC staff 7. Bank and Borrower Performance Bank 7.1 Lending: The Bank's performance in project identification, preparation and appraisal (Quality at Entry) is rated unsatisfactory. As noted earlier, the project design was consistent with the GOP's development strategy and the Bank's Country Assistance Strategy. However, the project focused almost exclusively on information technology instead of a tax administration project supporting capacity building in tax administration. Due to this limited approach, important additional project objectives besides the support of improvements in tax administration by providing technical assistance, staff training, hardware, software and priority equipment and facilities were not identified. Also it was assumed that the IMF would take care of necessary tax administration and policy reforms without formally agreeing with the IMF on a sharing of tasks and responsibilities. Total expenditure on project identification, preparation, appraisal, and negotiation was $120,000, representing 51.3 staff weeks of effort. While the informatics team assigned to manage the project understood general computerization aspects of the project, a multi- disciplinary team - including specialists in tax policy, tax administration, public sector management, organizational change, and capacity building -- would have better served the needs of the project and the GOP. 7.2 Supervision: The supervision plan in the SAR allowed for 60 staff weeks of effort. Supervision intensity was planned to be heavier in the earlier stages of the project and lighter toward the end. The SAR (section 3.41) states: "Bank supervision of the project will require abQut 60 staff weeks. The most critical period was 1993 for procurement. After contract award, project supervision will require 3 missions per year during the first 2 years and 2 missions per year thereafter. These missions will include the Task Manager, and Computer and Tax administration specialists". The procurement was indeed managed successfully and the intensive supervision (as well as the choice of a competent project implementation agency) may have contributed to this success.  - 27 - Total supervision effort over the period 1994 to 1999 was 148 staff weeks or 2.5 times what was planned in the SAR. A review of mission Aide-Memoires (AMs) indicates a strong Bank presence from a consistent team with expertise in informatics and economics. Notably absent was expertise in tax administration expertise. The Bank assisted and coordinated a very complicated procurement process, which was successful. The Bank was also instrumental in reviewing and amending the turnkey contracts to BIR and BOC's benefit. After the signing of the two turnkey contracts the Bank helped BIR and BOC in the software development and rollout. BIR supervision was more challenging to the Bank's supervision mainly due to the problems with the QAE, which of course had consequences for supervision. Lack of systemic process re-engineering resulted in complicated form design for BIR for instance. Ultimately, with both BIR and the Bank's efforts, simplifications occurred and the development was completed in a satisfactory way. However, this absorbed a lot of the Bank and the borrower's time and resources; and could have been addressed and resolved earlier on in a more efficient way had the BIR QAE been satisfactory. Based on review of the supervision Aide Memoire the Bank tried to push for form simplification since 1995, however this was implemented towards 1997. The ICR team reviewed in detail the AMs generated over the period 1992 to 1999. These AMs advised the Borrower to correct implementation problems, to resolve supplier issues; etc. These AM clearly demonstrate that the project. despite some original design flaws, was correcting itself and moving in the right direction. Like most Bank projects, the project was supervised by several TTLs. The initial team (1990 to 1994) was composed of Messrs. Moussa and Fernandez. Mr. Fernandez's untimely death in 1994 resulted in Mr. Moussa assuming project management responsibility in collaboration with Ms. Khan (1995 to 1997). Ms. Khan left the project in 1997; Mr. Moussa reassumed responsibility for the project until November 1998, when he retired from the Bank. Ms. Melhem officially replaced Mr. Moussa in March 1999. The project had the usual matrix management issues that the institution is presently trying to eliminate. The project's TTLs belonged to EMTTI's Informatics Group, while the project's "owner" was EASPR, and other project initiatives and activities were coordinated and managed from the Manila-based Country Unit. The management and information flow across this three- unit matrix (EMTTI, EASPR and the Manila Resident Mission) was sometimes difficult to coordinate and suffered from inattention during some periods of the project's life. Changes in the task team leader (four over the course of the project) and EASPR sector leaders further impeded the effective supervision of this project. Notably, the TCP had no official task manager for four months as the region left the project without an official manager. These managerial weaknesses were noted in the QAG review conducted in July 1999. After a May 1999 supervision mission, the project's very low disbursements resulted in its performance being downgraded to unsatisfactory. The project was reviewed by a QAG panel in July 1999 (based on its unsatisfactory rating), and the panel found the project's supervision quality to be unsatisfactory. This was based on several perceived factors: (i) procurement delays; (ii) weak management and supervision by EASPR, (iii) poor coordination and  - 28 - communication across the tripartite matrix; (iv) delayed communication between the Bank and the borrower on important loan matters; and (v) delays in assigning a new task manager to the project in November 1998. Following the QAG review, project coordination and supervision quality improved, although only in April 1999. A more attentive (and new) sector leader provided critical support and a hands-on approach to supervision. Additional resources and adequate expertise were subsequently allocated to the project, including expertise in tax administration, tax policy and public sector management. Subsequently, in December 1999 a final supervision/pre-Intensive Learning ICR mission was led by EASPR specialists in Public Sector and Tax Administration, helped position the project within the broader public sector reform context and reviewed its status, achievements, and short-to-medium term prospects. Finally the project was designated as one of the 3 EA projects to undergo an Intensive Learning ICR process, with the objective of self-learning and disseminating awareness, lessons learned and best practices in the Tax computerization area. Based on the above we find the Bank supervision quality satisfactory. 7.3 Overall Bank performance: The Bank's overall lending and supervision performance is rated satisfactory. The basis of this evaluation is based largely on the assessments above. Despite the challenges described above, the Bank did work to supervise the project adequately, but its supervision was impeded by the project's relatively short time span, its extremely ambitious goals and its initial design flaws. On the positive side, the procurement process was managed extremely well and is used as a model for projects with high Information Technology investments. The Bank's unrealistic disbursement schedule in the SAR and loan agreement (where most disbursements were expected to occur in the first half of the project's life) ultimately led to its unsatisfactory rating. When the two turnkey contracts were awarded, both contracts stated clearly that disbursements for software development would occur mostly at the end of the project, when these services were delivered. This fact was not reflected in the disbursement schedule, which should have been amended in 1994 to reflect the procurement of goods and services, rather than the procurement only of goods. The Borrower's opinion of the Bank's performance is better than the assessment listed above (as noted in BIR's ICR.) The Borrower did mention that it would have benefited from better financial management training regarding the Bank's reporting requirements for the project. The Borrower also thought the Bank should have pushed for the project to be tax-free at the loan signing. That issue caused many problems during the project as 30 percent of local expenditures (for contractual, services and equipment) were subject to tax, to be paid from counterpart contributions. Finally, the Borrower thought that the Bank should have insisted on business process re-engineering prior to computerization. This would have considerably simplified the computerization of the registration, collections, returns, filing and audit and may have resulted in a better tax administration system. The Borrower also suffered from the absence of a tax administration expert in the project team; and was confused by the abundance of advice provided  - 29 - by both the IMF and the World Bank on the project. For related information please refer to Appendix 10, ILICR workshop minutes. Borrower 7.4 Preparation: The TCP was the largest computerization project the GOP had launched so far. Considering the lack of experience in preparing and managing such projects the quality of project preparation was reasonably high. BIR and BOC selected qualified management teams to handle the project and competent project directors were nominated. The main weakness in project preparation on the side of BIR was the attempt to import existing procedures in need of reform into the computerization process instead of aiming at simplification and streamlining of procedures before computerizing. However, this was partly a fault of World Bank project preparation, because BIR was planning to re-engineer and streamline operations prior to the launching of the project, but was advised by the Bank that such work should be part of the contractor's responsibility instead of a pre-loan activity. BIR's performance in project preparation therefore can be rated as satisfactory. BOC did not face the re-engineering problem for several reasons: re-engineering was mandated by the 11-point reform program imposed by IMF (which BOC executed entirely), and the ASYCUDA software already included procedure simplifications. BOC had a strong leadership, was ready for reform, and expended immense effort to reform its bureaucracy and improve its relations with importers. BOC's performance at preparation was highly satisfactory. DOF supported the project's preparation at the highest level. Its preparation for the project was therefore satisfactory. 7.5 Government implementation performance: BIR can be lauded for several key achievements that are often lacking in other tax computerization projects: * Successful development and pilot-testing of a new tax administration software, the ITS * Simplifying tax procedures and introducing a uniform taxpayer identification number * Successful management of project costs within the umbrella defined by the WB loan despite numerous Change Notices. Unlike other similar projects, the final cost of the turnkey system did not exceed the original estimate by more than 14 percent. * Incorporating major tax policy changes in the Comprehensive Tax Reform Program into the development of the ITS * Improving the suspense rate from 80 percent in initial pilot sites to less than 10 percent today  - 30 - One negative aspect on the software development side is that BIR did not assume sufficient ownership of the development activities, which were to a large extent simply left to Anderson Consulting. The fact that the Project Supervision consultant (Crown Agents) was at an early stage of project implementation no longer allowed to attend the BIR/Anderson Consulting management meetings may have contributed to the supplier delivering less than what was originally expected (such as providing tax experts, forms experts and top of the line database design). The Borrower lost valuable time by changing its encoding and processing strategy. It modified the original RDO-based processing proposal to the RDC-based one; and today is reverting back to an RDO-based processing scheme. DOF was the main catalyst behind the project's concept and preparation and provided support to the project, especially with regard to defending the project to Congress or DBM in order to ensure the needed government funding. We find DOF's implementation performance satisfactory. However, overall the borrower's implementation performance was satisfactory as the Borrower worked very hard to realize the project's objectives and succeeded in this task. 7.6 Implementing Agency: The implementing agency, Crown Agents, can be credited for having conducted a problem- free, highly transparent procurement process, which concluded smoothly with the award of two turnkey contracts. Since 1994 the TCP procurement has been hailed as a success by procurement experts. The implementing agency seems to have been excluded from crucial design meetings in the project's early phase (1994 to 1995). The PIU could and probably would have given proper advice to BIR on certain fundamental issues such as form re-design. Similarly the PIU could have assisted BIR in its struggle over the TIN and other data conversion issues. These problems seem to stem from an unclear evolution of CA's role, who started initially as a "procurement consultant" and whose role shifted to " project supervision consultant" after the procurement cycle was completed. FINLINK was an excellent project that was mostly conceived and designed by the PIU to improve communications among BIR, DOF and BOC. However, the agencies did not share applications as originally planned. FINLINK could allow BOC to crosscheck the top 5000 exporters with the income tax and VAT returns in BIR's system. BIR could have allowed the SEC and BOC to access the BIR TIN database to ensure that BOC and SEC importer records were filed under the same TIN. Despite limited past usage, FINLINK is witnessing renewed attention in 2000 as the new DOF management team is encouraging BIR and BOC to use third- party information and cross-checking mechanism to verify tax information reported by importers and some large taxpayers.  -31 - Despite the end of its contract in November 1998, the PIU continued to assist the project until April 1999 without remuneration. One of the PIU's project managers, Mr. Louie Chanco can be lauded for his continuous support to the project. For all above reasons we find the performance of the implementing agency to be satisfactory. 7.7 Overall Borrower performance: Based on 7.4, 7.5 and 7.6 we find the overall borrower performance satisfactory. 8. Lessons Learned On the basis of this evaluation the project provides the following key lessons: I. Lessons Learned for the Borrower: Embedding the Tax Computerization Project within a Broader Public Sector Reform Initiative: The Tax Computerization project should have been positioned, from the beginning, under a broader public sector reform perspective. The project was planned in isolation of its environment, with no strategy for sustainability and little recognition of the external constraints and challenges ahead. From the beginning, the World Bank and the GOP should have planned for sustainability post-1999 in three major areas: (i) budget resources, by committing the GOP to allocate yearly a certain percentage of the overall BIR budget to the computerization activities; (ii) staff resources by reforming civil service laws or seeking an exception from general civil service salary scales for BIR and BOC IT personnel; (iii) management and leadership for the project regardless of internal administrative changes. Complexity and Scope: The messages from the project's clients and reviewers are mixed: on the one hand the project was too complex for BIR to achieve its objectives in six years; on the other hand the project's scope was too narrow and should have been integrated with tax policy and administration reforms as well as accompanied by sectoral reforms of the judiciary, civil service, etc. These are not contradictory statements; they simply imply that BIR should not have been left alone to revamp its own procedures. Rather, reforms of BIR should have been part of a sweeping effort begun in parallel with the other agencies that BIR depends on to be the fully functional and credible revenue authority envisioned at project inception. All project managers (past and present) commented on the project's complexity, the need to have broken it into different, sequential sub-projects, of which the total duration would have been eight to ten years. To quote a BIR manager, TCP was "too much, too soon, too lonely." Retaining Skilled IT Staff: The need for legislative changes to civil service rules was recognized before the project started. Market-based salaries for IT staff would have allowed both BIR and BOC to retain their qualified IT staff, whose departure was very costly in terms of (i) system sustainability; (ii) ability to improve the system; (iii) the opportunity cost of managers' time selecting and recruiting new staff; (iv) training costs as new hires had to be retrained for the same vacated IT positions; (v) security of the computer systems; and finally (vi) lowering of the  - 32 - remaining staffs morale and motivation. Possible solutions to the staff retention problem are to: (i) reform the GOP's Civil Service system by introducing performance-based compensation that is on par with the private sector; (ii) create an independent revenue authority (to include both BIR and BOC) with specified performance targets for revenue collection over a defined time period; and (iii) outsource IT operations for BIR and BOC to the private sector (currently under study by BIR.) Communication and Coordination with DOF: Even though the project began with excellent communication among BIR, BOC and DOF, the last two years of the project (1998- 1999) saw significant degradation in communication. Collection targets set by DOF were seen as too high and unrealistic by BIR staff. DOF management required a countersignature of every withdrawal application; hence, slowing down the disbursement process. These problems could have been avoided if the original steering committee had continued its weekly meetings, Crown Agents had continued their role of facilitator and coordinator, and all three agencies had been better aware of what was happening across all project components. As of Spring 2000 we note that communications have improved and DOF is actively coordinating with BOC and BIR's new commissioners. II. Lessons Learned in Software Development and Computerization: Development of the ITS: As noted above, the ITS software was built from scratch. The developer's task was to understand the full complexity of how the Filipino tax process works and to model the process in a software package. This very difficult task requires excellent communication between the client and the developer. The BIR representatives in the developer's team to advise and test the software should have included the most innovative BIR users and some volunteer taxpayers. The development should have been done in stages. The system design depended on a certain database architecture outlined in the SAR that proved to be unfeasible. Finally, the database reporting module (NOMIS) should have been developed to be an executive decision system for both BIR and DoF to use intensively as a analysis tool for tax and collection trends per region, per tax type, per industry, etc. Functional Specifications: These are the guidelines given to the software developer by the BIR users, which serve as the blueprints of the new system. BIR's specifications were neither geared toward a new system nor a computerized environment, and most of what they accomplished was to automate existing manual processes. This could be partially attributed to relative inexperience of the BIR users involved in the functional specifications; some misunderstanding about the contractor's mandate; and certainly lack of BIR process re- engineering prior to designing the system's functional specifications. Re-engineering of any agency's processes should be conducted prior to their computerization. BIR's processes in tax administration, intemal communications and public communication should have been reviewed, streamlined and simplified prior to the development of the ITS software. [I don't really think we need all these details here] Site Preparation: Both BIR and BOC had local offices that were inadequate to run a computerized system. Both the tax and customs systems were planned to run in a largely  - 33 - distributed environment (i.e., in several sites). These sites had to be prepared for the installation of major hardware; wiring offices; connecting them for data communication; ensuring adequate temperature, space, security and power conditions for the computers running the system; and containing crucial taxpayer and revenue information. The site preparation and facilities management were very costly to both BIR and BOC. A facilities upgrading plan should have been prepared prior to the project. The SAR did not examine that issue nor did it visit with sample RDOs to assess the potential costs of site preparation. Regional Data Centers: The project may have benefited from increased efficiency and economies of scale had there been one data center number rather than seven. This consolidation would have solved some of the IT staffing and training issues as well as some of the facilities management issues. The GOP is considering going back to a one data center architecture once the ITS is rolled-out entircly to the 41 RDOs that are currently computerized. Pilot Strategy: The pilot stage, which spanned more than two years, should have been conducted in one-year or only two sites, preferably co-located in the vicinity of the contractor and of BIR's IT support team in the National Office. BOC's successful experience in using its adjacent Port of Manila site as its pilot site benefited from easy access and hence faster problem solving for all of its ACOS pilot rollouts. Pilot testing was always staged and smoothly handled at BOC, hence its staff morale was quite high during rollouts. Taxpayer Education: Taxpayers should have been informed at an early stage about the impact of the project on administrative procedures. Also, the awareness of the population about the increased possibilities to detect stop-filers and non-filers with the use of IT should have been raised. Such an initiative should have been launched before or in parallel to the roll-out of the system. BIR started to create kiosks for taxpayer assistance only in 1999 as it realized its services needed improvement. This was neither sufficient nor timely. BIR had already launched in 1997 a special assistance desk only for the large taxpayers that was quite successful and well received by the target group. III. Lessons Learned for the Bank Custom Software Development: The ITS was developed from scratch. In 1993 there was no other choice as no "off-the-shelf' package existed for tax administration. Today, many options exist as with the advent of special tax administration software, such as Tax Solutions (developed by CIAT), ESKORT (developed by WM-data) various products, which have already been pilot-tested and used in other countries, are available. Assembly and customization still require time and effort especially to integrate multi-vendor products. Yet a clear advantage of buying off-the-shelf software is that the customer is guaranteed support, and timely improvements as the software maker improves and enhances its core products. BOC for instance, has already received seven upgrades of the ASYCUDA software. BOC and other ASYCUDA client countries can request system enhancements, and UNCTAD developers incorporate the client requests in their new releases of the core software. In future tax computerization projects the Bank should, to the extent possible, encourage its clients to purchase and customize standard tax administration software.  - 34 - Simplification of the Disbursement Process: Both BIR and BOC stated that the Bank's disbursement rules were complicated when applied to the project; where 30 percent of local expenditures were financed by local counterpart funds. This 30 percent was to include taxes. Until 1996, BIR and BOC were still making some mistakes in their withdrawal applications by including taxes in amounts requested for reimbursment. Training on the Special Account Management was insufficient as BOC did not report according to Bank procedures on the BOC Special Account (SA). All of these issues delayed both agencies and the Bank supervision team, and sometimes distracted them from other fundamental issues. Also, the Bank's disbursement estimates supplied in the SAR did not match the signed contract agreements for the two turn-key contracts, which stipulated that payments for services delivered (software development) would occur towards the end of the project and after the client has satisfactorily used the delivered software applications. In the future the Bank should also avoid to the extent possible multi- currency contracts as the ACPS turnkey had provisions for four currencies. Matrix Management: Project supervision was complicated by the fact its TTLs belonged to a different anchor unit (EMTTI) from the team formally responsible for the project (East Asia's PREM). This reduced the efficiency of various Bank processes and sometimes delayed decision-making. Need for a Tax Administration Specialist: The Bank relied on the IMF to provide tax administration and policy advice instead of having this expertise as part of the Bank team. This caused considerable confusion to the borrower who -(as stated in the ILICR workshop) "was torn between the WB and IMF's advice" and likely delayed the implementation of important administrative reforms. Change Management: The Bank underestimated the magnitude of the changes introduced by the TCP on all stakeholders: BIR, BOC, all oversight agencies, and the taxpayers themselves. The content and duration of the training programs were insufficient for the GOP and its stakeholders to absorb all of the changes introduced. Trainers, facilities, time, budget and incentives were inadequate. In approving large agency modernization loans the Bank should emphasize training and change management, which are critical success factors. Project Planning: The project's implementation plan was simplified. It did not include specific benchmark performance indicators nor success criteria. Both agencies had to tailor their own implementation plan to that of their supplier. Existing methodologies for software development cycle should be followed to enhance implementation efficiency systematically. 9. Partner Comments (a) Borrower/implementing agency: We have received comments on the draft ICR from the three major stakeholders, DoF, BIR and BOC. DoF lauded the Bank's candor and stated that it was encouraging to the borrower, and motivating for increased efforts to resolve remaining issues with the project. DoF agrees that the  - 35 - project preparation needed more focused and detailed evaluation and a planned re-engineering and streamlining of BIR procedures before project start. DoF noted that all the major observations and assessments in the report concurred with those voiced during the ILICR workshop by DoF, DBM, BOC, NEDA, BIR, etc. participants. DoF commented that the shared common views were a good indicator of sustainability since all the major stakeholders agreed on the project's successes and unresolved issues. DoF disagreed on two points mentioned in the report: a) On the weak commitment of government to support/sustain gains when administrations changed, and that the current administration did not place the project on their top priority list. b) On the feasibility of establishing an independent revenue authority that would combine BIR and BOC, in the short to medium term time frame. This is very much in line with our view as we think the Bank or other donors could assist DoF in financing a feasibility study for the independent revenue authority model merging BIR and BOC, today. BIR: BIR had some editorial comments and some reservations about some of the "unsatisfactory" ratings in section 4. However, BIR thought the ICR was an excellent reference document to be used for similar projects as it contained the essence of some replicable, universal, best practices in the sector. BOC: BOC had some editorial changes but was overall very appreciative of the thinking and analysis effort that the World Bank invested in the report. (b) Cofinanciers: N/A (c) Other partners (NGOs/private sector): N/A  - 36 - Annex 1. Key Performance Indicators/Log Frame Matrix Outcome I Impact Indicators: . d i0 dI i;j00toInd5i1x 0 0 1 uProjectedin last PSR Actualtest Estlmate Development of a 14 Modulle ITS System Completed Completed Rollout to BIR Offices: The National Office, Has to be completed by end of 2000 19 ROs, 44 RDOs Increase of internal efficiency of Tax On-going effect On-going effect Adminsitration: Cost of generating vs Revenue generated, for BIR and BOC Number of BIR and BOC staff that have 6000 BIR staff, 650 BOC staff 6000 BIR staff, 2500 BOC staff been trained on the system Development of the Customs ACOS and ACOS system developed, rollout to five last ACOS system totally developed, Rollout to rollout to seven sites sies (out of 21) have been iniated 21 sites have been completed Output Indicators: 1nSctrI4ti Pjete In las PSR A.ulLts Etat Increase in Number of TaxPayers correctly 3 Millions 10 Millions registered taxpayers, broken regsitered in all groups (regular Taxpayers, down as LTD, Industry Groups) 1.5 Millions self-employed and sole-proprietors 8.5 Million employees 1000 Large TaxPayers Increase in Revenue collection for all groups Not projected, at that time the revenue In Jan, Feb and March 2000 BIR exceeded (regular Taxpayers, LTD, Industry Groups) collected was dropping due to the economic its collection targets by PhP 4.5 Billion, a first crises, increase of non-compliance and time since 1992. prospect of tax amnesty. However revenue increase from the LTD was 24.65% between Sept 98 and April 99; a much higher increase compared to previous period and prior to ITS. Increase of collections in computerized The earlier computerized RDOs experienced RDOs versus non-computerized RDOs an average growth rate of 18.6% versus 8.4% for non-computerized RDOs. Increase of identification of stopfilers or tax 34 RDOs had to have the StopFiler Module Between May and Dec 99, BIR collected evaders. Prosecution of stopfilers and rolled-out and functional PhP 60.5 Million from the resolution of evaders 38,512 cases. Decrease in sumggling (for BOC) End of project  - 37 - Annex 2. Project Costs and Financing Project Cost by Component (in US$ million equivalent) Appraisal ActualSLatest Percentage of Estimate Estimate Appraisal Project Cost BY Component US$ million US$ million Bureau of Customs 27.40 16.20 Bureau of Intemal Revenue 54.90 35.30 Total Baseline Cost 82.30 51.50 Physical Contingencies 3.70 Price Contingencies 6.30 Total Project Costs 92.30 51.50 Interest during construction 13.80 Total Financing Required 106.10 51.50 Project Costs by Procurement Arrangements (Appraisal Estimate) (US$ million equivalent) Procurement Method Expenditure Category IC NC Other' N.B.F. Total Cost 1. Works 0.00 0.00 0.00 4.90 4.90 (0.00) (0.00) (0.00) (0.00) (0.00) 2. Goods 56.60 0.00 3.70 0.00 60.30 (54.40) (0.00) (3.20) (0.00) (57.60) 3. Services 0.00 0.00 5.40 0.00 5.40 (0.00) (0.00) (5.40) (0.00) (5.40) 4. Miscellaneous 0.00 0.00 0.00 21.80 21.80 (0.00) (0.00) (0.00) (0.00) (0.00) 5. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 6. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) Total 56.60 0.00 9.10 26.70 92.40 (54.40) (0.00) (8.60) (0.00) (63.00) Project Costs by Procurement Arrangements (Actual/Latest Estimate) (US$ million equivalent) Procurement Method ExpenOur IC2NCB er N.B.F. Total Cost 1. Works 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 2. Goods 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (45.60) (45.60) 3. Services 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (5.90) (5.90) 4. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00)  - 38 - 5. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 6. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) Total 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (51.50) (51.50) ' Figures in parenthesis are the amounts to be financed by the Bank Loan. All costs include contingencies. 2' Includes civil works and goods to be procured through national shopping, consulting services, services of contracted staff of the project management office, training, technical assistance services, and incremental operating costs related to (i) managing the project, and (ii) re-lending project funds to local government units. Project Financing by Component (in US$ million equivalent) Percentage of Appraisal Appraisal Estimate Actuat/Latest Estimate Bank Govt. CoF. Bank Govt. CoF. Bank GGovt. CoF. Bureau of Customs 31.52 20.97 16.20 51.4 0.0 0.0 Bureau Internal Revenue 31.48 22.23 35.30 112.1 0.0 0.0 0.0 0.0 0.0 Comments: The Total WB budget for BIR's tax Computerization project was USD 42 Million. The peso counterpart amounted to PhP I Billion (equivalent to roughly USD 25 Million based on 1999 exchange rate) The World Bank project loans of USD 42 Million was allocated to four major items, namely: 35% (or IJSD 14.8 Million) for Hardware Implementation and site preparation 31% (or USD 13 Million) has been allocated for software development, training of users on ITS and Technical assistance 21% (or USD 8.9 Million) has been allocated for roll-out implementation at national office, RDCs, RDOs, and the LTD. 13% (or USD 5.3 Million) has been allocated for non-computer equipment (video-conferencing equipment, information kiosks, vehicles, etc. and project management (Crown Agents) )  - 39 - Annex 3: Economic Costs and Benefits Please refer to the formal Financial Analysis of the project's IRR, summarized in both section 4.2, 4.3. The complete financial analysis is listed in the separately attached document and spreadsheets in Appendix 1.  - 40 - Annex 4. Bank Inputs (a) Missions: [Stag:e of Project Cycle No. of Persons and Specialty Perfoma nce Rating (e.g. 2 Economists, I FMS. etc.) Implementation Development Month/Year Count Specialty Progress Objective Identification/Preparation 2 IS 1991 1992 2 IS Appraisal/Negotiation 1993 6 IS, FE, PE Supervision 1994 4 IS S S 1995 4 IS S S 1996 4 FE, IS S S 1997 4 FE, IS S S 1998 3 IS S S 1999 10 FE, IS, PS, TA, TP S S ICR 2000 3 FE, IS, TA S S Data in table derived from the old FACT system (Financial and Cost Application, Task management Project. Project preparation and appraisal data do not reflect work output. Specialized staff skills FE: Financial Economist IS: Informatics Specialist PE: Procurement Expert PS: Public Sector Specialist TA: Tax Administration Specialist TP: Tax Policy Specialist U: Unidentifiable (b) Staff Stage of Project Cycle Actual/Latest Estimate _______________________ No. Staff weeks US$ (,000) Identification/Preparation 40.7 89.3 Appraisal/Negotiation 11.6 31.3 Supervision 148.4 399.1 ICR lX.l 92.5 Total 218.1 612.2  - 41 - Annex 5. Ratings for Achievement of Objectives/Outputs of Components (H=High, SU=Substantial, M=Modest, N=Negligible, NA=Not Applicable) Rating Macro policies O H O SU * M O N O NA @Sector Policies O H OSUOM O N O NA E Physical OH *SUOM ON ONA Z Financial O H OSU*M O N O NA M Institutional Development 0 H * SU O M 0 N 0 NA ZEnvironmental O H OSUOM O N * NA Social E Poverty Reduction O H O SU * M O N O NA @ Gender O H OSUOM O N * NA M Other (Please specify) O H O SU O M O N * NA @ Private sector development 0 H O SU * M 0 N 0 NA @ Public sector management 0 H 0 SU * M 0 N 0 NA Z Other (Please specify) O H OSUOM O N * NA  - 42 - Annex 6. Ratings of Bank and Borrower Performance (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HU=Highly Unsatisfactory) 6.1 Bank performance Rating C Lending O HS Os *U OHU O Supervision O HS * S CU OHU O Overall OHS OS C U O HU 6.2 Borrowerperformance Rating O Preparation O HS * S C U O HU O Government implementation performance 0 HS 0 S 0 U 0 HU O Implementation agency performance O HS Os O U O HU O Overall OHS OS G U O HU  - 43 - Annex 7. List of Supporting Documents Appendix 1: TCP's Financial Analysis, and related Annexes Appendix 2: BIR's own Implementation Completion Report Appendix 3: Beneficiary Survey Appendix 4: BIR Lessons Learnt Appendix 5: BOC's own ICR Appendix 6: Intensive Learning ICR Workshop on 4/12/00 a) Participants List b)Workshop Agenda  - 44 - Annex 8. Beneficiary Survey Results Please refer to Appendix 3, which contains the Beneficiary Survey results.  - 45 - Annex 9. Stakeholder Workshop Results Notes from Tax Computerization ILICR Workshop I. Introduction and Launching of workshop: The intensive learning implementation completion workshop was held on 4/12/00 at the Resident Mission. The workshop's agenda and participant list are listed in Appendix 8's attachments The workshop was attended by almost all invitees (27 participants out of 30 invited). Notable in their attendance were DoF Undersecretaries Castello and Gison, DCIR Guillermo, DCBOC Reyes; DBM's Ed Campos and former project leaders such as DoF Undersecretary Guevara and BOC Commissioner Parayno. DoF, DBM, NEDA, BIR, BOC. The TCP's consultants (Andersen Consulting, SunPhil, Unisys and UNCTAD) were also well represented. The Bank ICR Mission team consisted of Dana Weist, Michael Engelschalk and Samia Melhem. The conference was coordinated and facilitated by Louie Chanco. The workshop was introduced and concluded by Mr. Aloysius Ordu, Acting Country Director. Aloysius described the implementation completion exercise, its importance for the World Bank and its borrowers, and the value of the lessons learnt distilled from the project. The first section of the workshop was based on presentations of the project's outcome for BIR and BOC, followed by a World Bank presentation. BIR Presentation: DCIR Guillermo presented the BIR achievements as of today. To summarize these (listed in detail in BIR's ICR and in the document "Making every Peso count"): a) Development, completion and deployment of the Integrated Tax System b) Building of a large communication network connecting all RDOs, Data Centers, National Office, Large Taxpayer/Excise Division and 3000 Agent Authorized Banks to BIR's systems c) Massive training and Change Management program: Trained 6000 BIR staff d) Increase of revenue in 2000 by PhP 4.5 Billion above target (for Jan, Feb and March) e) Setup BIR Web page, Information kiosks and Taxpayer assistance units. i) Enhanced voluntary compliance as collections in computerized RDOs grew by 33% in contrast with 14% for non-computerized RDOs g) Total project costs (Bank financing plus counterpart funding) were PhP 2.1 Billion BIR concluded by stating computerization had enormous benefits on BIR, and would even have greater benefits if the agency was provided with the necessary resources to improve itself. Finally BIR stated that a mere 1% increase in collections is enough to recoup the BIR's investment in TCP.  - 46 - BOC presentations: The presentations were made by UNCTAD and Unisys UNCTAD presented the development of ASYCUDA++, its testing, its phased development and rollouts, its numerous upgrades (Currently the BOC has ASYCUDA 1.15) its core modules (Selectivity, Manifest Declaration, Assessment, Import entry, etc.) and its extensions (EDI, Direct Trader Input, On-line Release system, etc.) Unisys explained the development of the ACOS system. ACOS consists of three components: ASYCUDA++, (described above); Collection Systems (Project Abstract Secure, Automatic matching and payments and Payables and Project Reconcile) and Data Warehousing (statistics and comparative analysis). Unisys discussed also the implementation of the wide area network in 21 BOC ports, and the modernization of the agency in general. The World Bank concluded the first session with two presentations. The first was "Lessons learnt in Tax administration projects" (by Michael Engelschalk) and the second one discussed the importance of Public Sector Reform to backup Tax administration efforts (by Dana Weist). These introductory presentations got everybody ready for the second agenda item, undocumented outcomes and benefits that stemmed from the project. The following are most of the participants' comments on the subject. II. Undocumented Outcomes and benefits of the project Usec Guevera, DOF BOC components have done wonders: "overwhelmed with success" * DOF developed comprehensive and functional database to formulate policies and analyze problems based on TCP data. * TCP helped fight corruption, introduced transparency and reduced transaction costs. Key to BOC's success was (i) engaging of the private sector; which is even willing to pay for BOC's services; and mostly, (ii) trade facilitation. * Enhanced government expenditure and cash management; during deficit period, managed cash flow on a day-to-day basis (FINLINK to DBM). Daily forecasts also used to make decisions for next month based on TCP data. * Sustainability could also be insured from the private sector: If project is accepted, other resources could be made available (A certain Senator contributed his pork barrel to build a Customs center)  - 47 - * Taxpayers would be willing to pay for less staff discretion and better, honest services. UNISYS, Tony Tissington * commitment of people, and real team work between BOC and its suppliers (UNCTAD, Unisys) * packaging of services - not just hardware, but also better offices, equipment, processes and even Change Management. * workplace modernization benefits in general * prepayment of shipment was an excellent innovation in collecting duties and taxes from importers in a timely manner * selectivity allowed increased efficiency in cargo clearance processing * computerization - ACOS, (ASYCUDA ++, Collection System and data warehousing) other systems, interfaces to DOF and other government agencies and banks * change management achievements * Changing attitudes and workplace environment and behavior was a significant outcome of the project's implementation for BOC. This was mainly due to working with Development Academy of Philippines in developing change management program; and closely working with DoF as most directives came from top of DOF and BOC. CM included awareness surveys, questionnaires and workshops * lhuge increase in rollout without any other resources; roll out teams were extraordinary in gaining acceptance Efficiency gains BOC has achieved less paper (fewer opportunities for bribes, streamline procedures) improved cargo release times (few hours instead of up to 20 days) * selectivity improvements * cargo inspection improvements * pioneers in new frame relay network technology - won awards for using this technology (led technology development - commercial use) to build a first-class Wide Area Network (WAN) for the 21 modernized ports * Contractor focused on project outcomes (need to get job done) rather than sticking to the text of the contract's TOR, flexibility and responsiveness to client's need  - 48 - Comm. Parayno, BOC * much of success due to willingness of UNISYS to establish itself as a leader in customs project integration (willingness to view project implementation as an investment rather than a source of gains) * partnership among DoF, BOC, UNCTAD and UNISYS extraordinary and key to success * Project Enhanced reputation of BOC * Level of corruption measured by SWS in 1991: police and customs were tied as most corrupt agencies (-48 percent) * Comments from Federation of Philippines Industry: major change in satisfaction with how BOC is performing its job especially with regard to graft and corruption (bureaucratic) -- not as much impact on smuggling * International observers have found greater professionalism in BOC staff (belief in themselves and BOC's mission, and ownership of the Customs modernization project) relative to 5 years ago (failure to identify with organizational goals) Ed Campos, DBM - Focus on sustainability now that the main achievements have taken place: Involving private sector may provide some elements of sustainability - private sector demands may impede political interference (should perhaps be extended to BIR). * Need for public sector reform: cannot address institutional issues (possible misuse of computers - example of property tax payment on and off the table) despite the systems' built-in audit trail. The reform has to include Civil Service and Judiciary reform and change many related government processes such as budgetary allocation and overall agencies' performance management. NEDA staff * Broadening of taxpayer base enhances fairness of tax system, and the TCP has contributed to increasing the number of registered taxpayers. * Need to train government auditors, such as COA staff to work with computers to understand BIR and BOC's computerization issues. BIR * AC had excellent Filipino partners and that was one of the factors behind the project's success.  - 49 - Discussion of Project Outcomes (all participants) * Role of Crown Agents provided continuity between DOF, BIR and BOC as CA established communication standards and change notice management between both Bureaus and their supplier. Crown Agents was key to a successful and smooth procurement process. * Recognition by Congress, as Legislators' testimonials (both Houses in Congress) state seeing real improvements in BOC since the TCP was implemented Trade and export promotion with shorter cargo release times Simplified immigration and customs forms (1 short form similar to the US entry Customs form) * Leadership in Information technology implementation: BOC pointed the way in how IT can be used to improve and enhance efficiency of processes, while addressing: * Challenges of computerization, decentralization Information effect on reengineering processes Need to move fast on technology projects Seeking partnership with private sector by outsourcing some IT functions Training and capacity building to BIR and BOC staff in general Creation of BOC and BIR training centers and programs Training in tax and customs is essential element of capacity building * Technology literacy * Change in work processes towards knowledge and analytical work from clerical functions (BIR and BOC) as: * Electronic data transmission greatly enhances timeliness and possibilities for trend analysis and forecasts (BIR) * Previous work was computational rather than analytical (BOC) * Future need to make staff analysts and data miners. (However need for further analysis and intelligence work. Still much has to be done; BOC would be happy if 3 percent of existing data in its data warehouse were analyzed systematically towards intelligence and fraud detection purposes) Cross checking with other institutions (LRA, BOT, etc.) much easier  - 50 - Enhanced taxpayer awareness regarding filing and compliance. III. Challenges and Difficulties Encountered in the project The third section focused on the Challenges and Difficulties encountered in the project. Participants filled out what they experienced as challenges in the following 13 categories. After that, volunteers reported on each category to the rest of attendees, and some additional feedback and brainstorming was generated by the attendance. Reports on Challenges Encountered (ranked from highest to lowest priority) 1. Challenges in GOP support, coordination and ownership * Changes in leadership caused discontinuity to policy and direction * New BIR Commissioners had their own agenda and one did not support the Large Taxpayer Services and changed the TCP specifications repeatedly * Weak commitment of Government to support/sustain project gains when administration changed. For example, the Ramos administration was an enthusiastic supporter of the technology effort but the succeeding administration did not place the project on top of their priority list. Wavering political support over changing administrations, lobbying for the project's benefits and sustainability absorbs agency managers' time. Administrative arrangements became more difficult with new leaders Cabinet members place the tax administration budget among lowest priorities (need for better awareness of sustainability issues with DBM, DoF and Congress) Leadership and change management functions - There is a real need to create organizational ownership; however, this is still perceived as external imposition (which requires top level political leadership) by BIR and BOC Very unclear division of responsibilities at DOF - confuses staff; DoF had sometimes several Usecs overseeing TCP, creating confusion and inefficiency No clarity with respect to authority of DOF over BIR or BOC Commissioners all are in grades equal to DoF undersecretaries, and this created authority problems and sometimes minimal teamwork. FINLINK, one-stop shop should be in the same DOF office that coordinates the TCP Higher salaries for technical staff in GOP New paradigm for IT personnel; consider privatizing or out sourcing options  - 51 - * Head of BIR/BOC not protected from harassment, lawsuits, by the private sector; so the managers may waste their on trivial civil action suits. Disbursement procedures are long and inefficient between the agencies Voluminous documents needed for the disbursement process * Above two points results in slow release of funds, slow disbursements and hence Bank rating of the project as Unsatisfactory (June 99) * Need for stronger support for Customs Modernization Bill (especially section on electronic submissions and paperless processing) crucial for adoption of Ecommerce. 2. Challenges in Project Financing * Difficulty in requesting funds from DBM. Inadequate counterpart funding, no guarantee of availability of the Pesos counterpart. Problems with the exchange rate fluctuations between 1993 and 1999. * Sustainability problem: Need additional funds beyond budget for regular operations. * No creative financing scheme or strategy (perhaps using private sector BTO schemes) to fund the operation of project beyond closing date. 3. Challenges in Legislation and Tax Policies * E-commerce bill has not been approved yet. For BIR to be able to process electronic filing, the e-commerce bill approval is crucial. Today electronic filing and signatures are not recognized as legal document. BIR was not consulted when new Government placed emphasis solely on revenue collection CTRP developed separately from project - some systems not applicable to new reform and tax policies. The tax laws should have been reformed before developing the new tax software system. The approval of CTRP in the middle of the project caused major disruption. Legislators impede tax agency's ability to administer taxes - unfunded mandates, more work for BIR staff to manage new taxes with no increase of personnel. Whenever laws are changed, lawmakers should consider the effects on current systems (including consultation with IT professionals). The lack of thereof consultation reflects a management problem in bureaus - bureau staff always involved in discussion of changes but IT staff are not included at outset. This lack of communications makes it very difficult to manage the project.  - 52 - 4. Challenges in Tax administration, organization and procedures * The main issues revolve around not having made major changes of policy and organization nor any procedure streamlining before the Project. Process re-engineering was needed and was not done. - Changes in Tax policy affected Tax administration, which affected computerization. All activities (policy, administration and computerization) should be cohesively planned to work together. The continuous change of people and positions was very disruptive. The lack of leadership continuity lessened the project's urgency and importance to the GOP. New administration meant new policy and strategies, which in turn has an implication on development; which had to be modified for the new administration. Form redesign and restructuring of manual processes and procedures occurred after computerization Many programs were launched on spiritualism and values that were unrelated to computerization. In fact, too many things launched at the same time; staff received mixed signals and too much to learn. While there were many initiatives, these were not part of a comprehensive overall program BIR staff were pressured to roll out despite problems (staff performance was benchmarked on rollouts). That contributed to the loss of sight of the enhancing revenue collection objective Too many changes occurred in IT architecture and structure. BIR toggled between decentralization and centralization. TCP's implementation had a great impact on organization structure and procedures on the agency. No adequate implementation infrastructure was created for the reform process. BIR's organization is old and outmoded, and its staff skills do not sometimes match its organizational needs. BIR was not properly organized to assimilate the TCP. BIR and BOC have no structure to do administration of tax cases. BIR should increase its audits by immediately making audit selection based on objective criteria, via computer data, or if not available, via sampling. Efficiency and fairness of examiners should be rated by computers given some objective, quantitative criteria complemented by qualitative factors  - 53 - 5. Challenges in Human resources and training Recruitment and retention of highly skilled personnel: Extremely difficult under today's defective HR system. * Large gap between government and private sector compensation level (contractor paid staff average pH 50,000/month, GOP paid staff average PhP 14,000/month) * Issuance of administrative order 70 and 100 prevents hiring of new personnel Unreasonable standards for technical (IT) staff defined by Civil Servant Commission such as having a Master's degree in an MIS-related field from an accredited Filipino academy or university Need for continuous technical training program, which was not possible in the project due to insufficient funding. The continuous training program was needed due to the several new tools to learn, to the long learning curve, slow staff absorption and to the need to constantly retrain new staff because of the drain in the IT personnel. * Training needed to be hands on, not only CBT or video-based. * Retention of Change Management training was needed. BIR and BOC turnover was a major issue on training (not just technical staff) as well as RDO heads rotation. Government personnel stalled from assuming more active role as long as consultants were around. This deepened the Bureaus' dependencies on their contractors and lack of ownership for the software developed. 6. Challenges in Change management and organizational culture Resistance of personnel in reforms resulted in manual procedures being added to process as opposed to being changed. The TCP was seen as a radical change instrument and was vehemently rejected by some. Agency staff should prepare staff in top and middle management by holding extensive discussions to accept changes first and then implement them. Lack of a systematic Change Management (CM) plan. CM should be a critical and significant component of reform projects such as the TCP. The program would include values orientation, formation seminars on regular basis and on-going refresher programs or orientation for new comers. Cultural issues: There is a tendency to reject what others have accomplished, the "crab" mentality, big threat to sustainability. Outcome not linked to financial performance; lack of systematic change management plan (success measured in number of rollouts and not in broader  - 54 - institutional objectives). Computerization should not be perceived as the main solution to improve revenue collections. * Assuming that IT will solve problems of corruption is prescribing "aspirin for cancer". * Ability to adapt to the system and use it as a tool for graft and corruption should apply to both government and private employees. 7. Challenges in Changing Taxpayer attitude and reaction * Focus from the beginning on the taxpayer and his education. Adequate program for taxpayer education came in late in BIR. Had they have been in place in 1996, the 95 percent in suspense rate may have been a lower figure. Taxpayers did not clearly see TCP's benefits, actually the public reaction was inconvenience and annoyance towards it. * Taxpayers should have benefited from system - system could have helped them calculate their taxes once they had entered all their family and work information Taxpayer attitude toward BIR and BOC determine success of any computerization project. Unfortunately BIR and BOC are still viewed as corruption and graft ridden. 8. Challenges in Relation with contractors and consultants Adversarial relationship between AC/PS and CA Crown Agents (CA) got lost in the shuffle, as Andersen Consulting (AC) did not want it included in the BIR steering committee meetings. * Not much trust between AC, Philippines Systems (PS) and users, due to cultural insensitivity of AC's project director at that time. AC refused to participate in DOF meetings since they said reported for BIR. * CA became too close to BOC and BIR to provide objective advice to DoF. 9. Challenges in World Bank Assistance Reengineering or transformation program should be critical component of program. The World Bank (WB) did not insist on having a transformation program happen before the project as it was in a rush to move the project to the Board of directors for approval. There was no provision for the development of a business plan before the development of the IT Plan for BIR. * No tax administrators on any of the supervision missions, just IT specialists. Project monitoring was IT-based and not Tax administration-based.  - 55 - * WB was too optimistic about BIR's absorption of changes and the amount of time needed to finish the project. * BIR discussed issues separately with AC and CA; at the same time, CA was shut off from the consultative process. The WB did not attempt to correct the situation; nor to hold joint meetings including CA, ACPS, BIR and DoF. * Lack of coordination between World Bank and IMF. Sometimes both organizations would give conflicting advice and that caused some confusion to BIR. * WB did not provide project plan nor adequate indicators and milestones to evaluate project progress * WB should have spent more time pointing out slippage, problems and regressions and things not turning well in program. * WB was too politicized in appraising projects 10. Challenges in Project Design, Planning and Procurement * Tradeoff between simplification (simplification of forms suggested by AC knowing that it would result in high suspense rates) * Sustainability issues - budget, staff, leadership, political will and ownership un- addressed for a long time and only considered towards closure of loan. There should have been a comprehensive, front-end commitment for sustainability as a pre-condition for loan approval. 5-year turnkey approach was too ambitious, unpractical and sometimes impossible for objectives to be achieved over such a short-time span given allocated budget. Underestimation of complexity of project Need to reduce project into more manageable identifiable sequential phases, or "chunks" instead of binding the agencies with one large big design. This issue was raised at design stage however the World Bank pushed for a single turnkey. A phased approach (over eight or ten years) would have resulted in simpler smaller pilots, and improved roll- out cycles (build small modules, test, pilot and roll out) DOF's need for information for policy and planning was not prioritized in system; DOF had to do separate project to build database; DOF should have been involved in planning of system and should be now a user of the ITS. Did not openly address up front how the new system could be used to identify, address and eradicate graft and corruption Government procurement systems impeded system development and improvement.  - 56 - 11. Challenges in Software Development BIR's Process of re-engineering and forms redesign should have preceded software development. Forms should have been simplified for taxpayers. * Integration of new systems (BOC and BIR) with existing legacy systems. * E-filing and e-payment, as well as automating data capture would have been very helpful. This could have been done on a BOT basis (Build Operate Transfer). No computerization strategy, nor plan for the future beyond the loan closing date * Changes to UNCTAD core software had to go through Geneva; which slowed down the process. BIR significant process changes were made difficult due to contract terms, and conservative project owners. The contractor' simplification strategies being viewed suspiciously by client, interpreted as an attempt to do less work and hence rejected. Now would purchase packaged tax software rather than develop our own Too little attention was paid to data migration Establish acceptable procedures for acceptance and payment of goods and services 12. Challenges in Rollout and Usage of New Systems * BIR's roll out pilot was too ambitious, not focused on large taxpayers, and scattered geographically in too many sites BIR's lack of trained personnel before roll out (roll out was fast tracked to meet benchmarks) BOC's personnel resistance to change; low confidence on new system Rolling-out the software to decentralized agency staff and to banks, while combining the roll-out with a harmonious training plan and sustaining normal operations Learning curve of personnel 13. Other Challenges Well thought of contingency plan for hand over of project not made COA auditors not knowledgeable in information technology issues Policy changes with every new manager  - 57 - Establish accepted, universal, change control procedures for project IV. Lessons Learnt Paradigm for continuity despite administration changes is a must. Private sector could have a role, in a BOT type relationship with the Bureaus to improve Service Delivery. * Automating Manual procedures is insufficient. A Complete examination of agency processes, policies and linkages ought to be examined, improved and changed before any modernization effort is undertaken. * Sporadic training is not sufficient. Training should be part of an on-going Change Management program funded and supported at the highest level. * People at the working level must have incentives to keep system. Need a strong, professional, committed middle management. Need internal enthusiasm, vision and leadership to sustain the project Complexity is also a factor of the geographic dispersion and level of decentralization. The project would have been much easier if it had to be implemented in the national office only. V. Future/Next Steps Need for BIR to implement the IMF's recommendations Need For Immediate ownership, Leadership and Roadmap Direction will depend on managing expectations; computerization of both bureaus can only go so far. The rest is a matter of agency absorption. Focus Large Taxpayers are the focus today not the underground economy. The latter should be officially confronted and eradicated as much possible with the help of the computer system, and with legislative support. Use of Third-party Tools Commissioners have been granted broad powers to use 3rd party information and this should be fully utilized (change introduced by 1997 tax reform). Need to enhance 3rd party information (social security, etc.)  - 58 - * As a starter for above point, a realistic strategy is to focus on existing databases and crosscheck information in both (link BIR and BOC's VAT and import information for instance; for under-declarations). Solving the IT staff turnover problem Must find a creative and immediate solution to retaining and improving the training of competent technical (IT) staff * Need for financing of feasibility study of outsourcing (BOT), especially of IT functions Need to change the Civil Service laws as the law these cause graft and corruption Improve Agency Image and Communication to the citizens GOP must fund and develop a Communications and outreach plan for BIR. Attrition and redundancy program to get rid of staff with unsatisfactory performance in both BIR and BOC. Having good staff improves the Bureaus' image. Find Creative solutions for budgetary sustainability Creativity is an answer to concerns of sustainability New program: super green lane program arose from analysis of data from system (and desire to loosen controls on "good guys") - expectations are very high. Proposing to charge a fee (P2500 per transaction) for this program which may address some of sustainability concerns. Rewarding internally agency staff who comes with innovative proposals on bettering the agency's organization, systems or procedures. Making further efficiency improvements Move to electronic processing of export clearances - need funding (UNISYS available for one more year; annual budget has already been expended)  - 59- Appendix 1 WORLD BANK OFFICE MANILA REPORT ON THE ICR FINANCIAL EVALUATION OF THE PHILIPPINES TAX COMPUTERIZATION PROJECT (TCP) MAY 11, 2000 Manila, Philippines   - 60 - ICR FINANCIAL EVALUATION OF THE PHILIPPINES TAX COMPUTERIZATION PROJECT (TCP) Contents I Introduction I 1. Background 2. Objective 3. Coverage 4. Methodology II Executive Summary 2 III ICR Financial Analysis 3 IV Conclusions and Recommendations 3 V. List of Annexes 4 1 - Incremental Project Costs 1993 - 2000 1-a Incremental Project Costs 1993-2000 (Actual VS. SAR) 2 - Consolidated Revenues (Actual Collections VS. Target) 1993-2000 2-a Consolidated Revenues (Actual Collection VS. SAR Target) 1993-2000 2-b Bureau of Internal Revenue - Schedule of Collection Target VS. Actual 2-c Bureau of Customs - Schedule of Collection Target VS. Actual 3 - Financial Rate of Return Calculations 4 - Assumptions for the Project's Financial Rate of Return   - 61 - FINANCIAL EVALUATION OF THE PHILIPPINES TAX COMPUTERIZATION PROJECT I Introduction 1. Background The Staff Appraisal Report (SAR) prepared on March 31, 1993 for the Philippines Tax Computerization Project indicated attractive benefits on project returns and recovery of project costs using the Net Present Value (NPV), Financial Internal Rate of Return (FIRR), and Benefit-Cost Ratio evaluations. There were likewise other non-financial benefits cited in the SAR. The project closed last December 31, 1999 and there is now the need to prepare the project's Implementation Completion Report (ICR). This financial evaluation is part of the ICR. 2. Objective The actual benefits from the project are to be evaluated. Likewise, these benefits should also be looked at in the light of the benefits cited in the SAR, with appropriate adjustments to be comparable, and of the economic conditions existing during project implementation. Specifically, the objective of this study is to evaluate benefits from the project using Annex 12 as well as paragraphs. 4.15 and 4.16 of the SAR. 3. Coverage The evaluation covered the period 1993- 2000 with the Bureau of Internal Revenue (BIR) and the Bureau of Customs (BOC) as its major components. It also covered the regional computerization/large taxpayers where data was available. 4. Methodology The evaluation mainly used the Discounted Cash Flow (DCF) evaluation tools and the Benefit-Cost Ratio based on DCF results. The DCF tools used are the Net Present Value (NPV) and the Financial Internal Rate of Return (FIRR). The framework of analysis used is that of the Annex 12 of the SAR. The figures in the SAR were deflated by the reduction in the GDP growth rates so as to be comparable with the actual growth rates used in the evaluation. Since rollouts only started in 1996, the project benefits started to be computed only at that year, which represent a 3-year delay on incremental benefits. As such the SAR figure on 3-year delay has been used.  - 62 - II. Executive Summary The benefits from the project were financial and non-financial. The incremental benefits could only be computed with a 3 year delay starting 1996 due to the fact that only in 1996 did the roll outs started. On the financial benefits, the project gave a total incremental collection benefits to the government of about Php 2 trillion from 1996 to 2000 or an equivalent FIRR of 307 % from a total cost of Php 3.35 billion. An attractive return indeed. The Benefit-Cost Ratio based on present values is about 9, indicating a highly beneficial undertaking. The NPV of the project's incremental revenues using interest rates of 10 % to 20 % averages to Php 16 billion. Comparing this to the project's average cost present value of Php 2 billion, the project earns more than what it costs by about Php 14 billion. The foregoing benefits were obtained despite the reduction of almost half of its assumed GDP growth rates from 7% to an average of 4% as a consequence of a sluggish economy mainly brought about by the Asian crisis. Comparison to the SAR could not straight away be made on the peso figures since the GDP growth rates have been significantly different. Thus based on a revised SAR, imputing the GDP growth changes, the project attained 90% of the SAR incremental collection targets under a 3-year delayed benefits comparison. The FIRR of the project of 307 % is higher than the recomputed SAR FIRR of 210 % mainly because of the heavier initial cash outflow under the SAR. The benefit cost ratio and NPV of the actual figures and the SAR are not comparable due to the significant reduction in the GDP and lower collection. Despite all the reduction in GDP, the project managed to retain its attractiveness of return and cost recovery as a worthwhile project to sustain. The non-financial benefits are from the project were easily felt and clearly noticeable. At the BIR, the computerization of key business processes such as receipts and reporting of tax payments, among others greatly enhanced employees' productivity. At the BOC, computerization has improved cargo release time, the selection and extent of physical and documentary inspection, and the banks' reporting and transmittal of importer's payment. Improved internal and external communication has also greatly enhanced the efficiency of both agencies. However, the impact on manpower productivity in terms of head count and time charges has not yet been established. III. ICR Financial Analysis The actual incremental collections from 1996 to 2000 was about Php 2 trillion which is 90% of the adjusted SAR figures primarily due to the Asian crisis resulting to sluggish GDP growth (Annex 2-a). This represents an FIRR of 307 %, an NPV of about Php 19 million compared to cost NPV of about Php 2 billion with a Benefit-Cost Ratio of 9 (Annex 3). This analysis did not include the other intangible benefits, which are difficult to quantify, more particularly the economic benefits brought about by  - 63 - automation and work simplification. By the closing of the project on December 31, 1999, both agencies met their respective rollout objectives. IV. Conclusions and Recommendations In spite of the economic crisis and other difficulties encountered during its implementation, the TCP achieved reasonable financial benefits, although it did not meet the target set in the SAR. The benefit-to-cost ratio posted a respectable 9 times at discount rates ranging from 10% to 20%. Other financial indicators such as internal rate of return and net present values also showed encouraging marks. This financial evaluation did not consider the non-quantifiable economic and social benefits that were inevitably brought forth by the project. In like manner, revenue losses due to deficiencies and inefficiencies of the manual system were not quantified. Considering the immediate and long-term impact of the project not only to the agencies concerned but also to the Filipino taxpayers as a whole, the overall benefit of the TCP more than compensates its investment cost. It is important, however, that the full implementation and operation of the computerized system be sustained if the government wants to maximize the benefits on new technology. The planned linkage with other key government agencies is also worth pursuing; proper and consistent system maintenance, both hardware and software must also be undertaken. The cooperation and support of the tax-paying public are also greatly needed to ensure continuing success of the computerization project. V. Annexes 1 - Incremental Project Costs 2 - Consolidated Revenues 2-a Bureau of Internal Revenue - Schedule of Collection Target VS. Actual 2-b Bureau of Customs - Schedule of Collection Target VS. Actual 3 - Financial Rate of Return Calculations 4 - Assumptions for the Project's Financial Rate of Return   - 64 - A N N E X E S   - 65 - Annex 1-a TAX COMPUTERIZATION PROJECT ICR Financial Evaluation Incremental Project Costs 1993 - 1999 (Actual VS. SAR) (Billion Pesos) Year Total Expenditures SAR 1992 1992 % of Total BOC BIR Total Prices(1) Prices Expenditures 1993 0.1295 0.000 0.130 0.120 0.239 50.421 1994 0.210 0.164 0.374 0.438 0.528 82.966 1995 0.218 0.391 0.610 0.481 0.470 102.306 1996 0.182 0.525 0.706 0.513 0.339 151.466 1997 0.131 0.293 0.424 0.293 0.365 80.236 1998 0.050 0.328 0.378 0.239 0.299 79.705 1999 0.091 0.410 0.501 0.296 0.299 99.015 2000 0.105 0.123 0.228 0.127 0.299 0.423 Total 1 117 2.233]_ 3.350 28 2.840 8.318 Note: 1. Current prices divided inflation factor based on 1992 prices.  - 66 - TAX COMPUTERIZATION PROJECT Annex 2 ICR Financial Evaluation Consolidated Revenues (Actual Collection VS. Target) 1993-2000 (Billion Pesos) Target Actual Target Year BIR BOC Total BIR BOC Total Attainment 1993 143.34 77.55 220.89 145.93 81.97 227.90 103 1994 185.49 84.52 270.00 187.44 81.61 269.05 100 1995 214,11 93.35 307.45 211.46 97.60 309.06 101 1996 258.79 101.35 360.14 256.49 104.57 361.06 100 1997 333.18 110.80 443.98 314.54 94.80 409.34 92 1998 335.06 78.28 413.34 337.64 76.01 413.64 100 1999 353.63 83.64 437.27 341.16 86.50 427.66 98 2000 397.80 91.88 489.68 414.00 93.68 507.68 104 Totals 2221.39 721.35 2942.74 2208.67 716.73 2925.39 99 Note: The collection data for 2000 was based on annualized first quarter collections of P82.8 billions for BIR and P28.11 billions (as of April) for BOC.  - 67 - TAX COMPUTERIZATION PROJECT Annex 2-a ICR Financial Evaluation Consolidated Revenues (Actual Collection VS. SAR Target) 1996-1999 (Billion Pesos) Actual | _ SAR Year BIR BOC Current P 1992 P(1) 1992 P(2) % Attainment 1996 256.493 104.566 361.06 262.59 259.06 101 1997 314.542 94.800 409.34 283.28 284.17 100 1998 337.637 76.005 413.64 260.97 293.92 89 1999 341.158 86.497 427.66 253.05 315.02 80 2000(3) 414.000 93.667 507.67 282.04 343.59 82 Totals 1663.830 455.535 2,119.36 1,341.93 1,495.75 90 Notes: 1. Collection at current prices divided by inflation index. 2. Tax effort with the project multiplied by GDP based on 1992 Prices 3. Estimated collections based on first quarter data or P82.8 billions for the BIR and P28.11 billion for the BOC. Price Index Calculation at 1992 Base Price SAR Actual(Current) INFLATION INFLATION Year rate factor rate factor 1992 8.90% 1.000 8.90% 1.000 1993 5.60% 1.056 7.60% 1.076 1994 6.00% 1.119 9.00% 1.173 1995 7.00% 1.198 8.10% 1.268 1996 7.00% 1.282 8.45% 1.375 1997 7.00% 1.371 5.10% 1.445 1998 7.00% 1.467 9.70% 1.585 1999 7.00% 1.570 6.60% 1.690 2000 7.00% 1.680 6.50% 1.800  - 68 - Annex 2-b TAX COMPUTERIZATION PROJECT ICR Financial Evaluation BUREAU OF INTERNAL REVENUE Schedule of Collection Target VS. Actual (1993-1999) (In Billion Pesos) .I 1993 1994 1995 1996 1997 1998 1999 Total Number of computerized RDOs 4 5 9 9 Computerized RDOs Actual 27.581 43.749 156.499 157.231 Target 33.615 48.326 180.093 164.82 % Attainment 82% 91% 87% 95% Non Computerized RDOs Actual 145.931 187.444 211.462 228.912 270.792 181.138 183.927 Target 143.336 185.486 214.105 225.176 284.849 174.971 188.812 % Attainment 102% 101% 99% 102% 95% 104% 97% Grand total Actual 145.931 187.444 211.462 256.493 314.542 337.637 341.158 Target 143.336 185.486 214.105 258.791 333.175 355.064 353.632 % Attainment 102% 101% 99% 99% 94% 95% 96% Incremental Collections 12.027 41.513 24.018 45.031 58.048 23.095 3.522 Growth Rate 8.98% 28.45% 12.81 % 21.30% 22.63% 7.34%1 1.04%  - 69 - Annex 2-c TAX COMPUTERIZATION PROJECT ICR Financial Evaluation BUREAU OF CUSTOMS Schedule of Collection Target vs. Actual (1993-1999) (Billion Pesos) Increment Year Target Actual Attainment Amount Percent 1993 77.550 81.971 106% 9.101 12.49% 1994 84.515 81.610 97% (0.361) -0.44% 1995 93.345 97.601 105% 15.991 19.59% 1996 101.348 104.566 103% 6.965 7.14% 1997 110.801 94.800 86% (9.766) -9.34% 1998 78.279 76.005 97% (18.795) -19.83% 1999 83.637 86.497 103% 10.492 13.80%  - 70 - TAX COMPUTERIZATION PROJECT Annex 3 ICR Financial Evaluation Financial Rate of Return Calculations (Consolidated - BIR and BOC) (Billion Pesos) GDP Tax Effort Tax Effort Total Project Benefits(3 Years Delay) Real 1992 Without With the Project Incremental Net Year Growth(1) Prices(2) Project Project(3) Expenses Collection Benefit 1351.559 1993 1.39% 1370.313 15.53% 15.46% 0.130 0.000 -0.130 1994 5.40% 1444.242 15.61% 15.88% 0.374 0.000 -0.374 1995 4.11% 1503.603 15.69% 16.21% 0.610 0.000 -0.610 1996 5.06% 1579.611 15.76% 16.62% 0.706 13.65 12.942 1997 6.07% 1675.533 15.77% 16.91% 0.424 19.03 18.606 1998 0.41% 1682.430 15.78% 15.51% 0.378 0.00 -0.378 1999 5.13% 1768.784 15.79% 14.31% 0.573 0.00 -0.573 2000 5.00% 1857.220 15.80% 15.19% 0.156 0.00 -0.156 Total 3.351 32.678 29.327 Notes: FIRR 307.35% 1. (GDP CY-GDP PY) I GDP PY 2. GDP current year divided by Discount inflation factor during the year Rate Present Values 3. Tax collection to GDP ratio at 10% 2.21 21.14 18.93 current prices 12% 2.05 19.47 17.42 14% 1.91 17.97 16.06 16% 1.77 16.60 14.82 20% 1.55 14.23 12.67 Benefit Cost Ratio (Times) 10% 9.6 12% 9.5 14% 9.4 16% 9.4 20% 9.2 GDP Calculation GDP Year INFLATION 1992 Current rate(1) factor(2) Prices(3) Prices(4) 1992 8.90% 1.000 1351.559 1351.559 1993 7.60% 1.076 1370.313 1474.457 1994 9.00% 1.173 1444.242 1693.865 1995 8.10% 1.268 1503.603 1906.328 1996 8.45% 1.375 1579.611 2171.922 1997 5.10% 1.445 1675.533 2421.306 1998 9.70% 1.585 1682.430 2667.106 1999 6.60% 1.690 1768.784 2989.065 2000 6.5%(5) 1.800 1857.220 3343.000 Notes: 1 NEDA Statistics 2 Previous Year's inflation factor X (Current Year's Inflation Rate + 1) 3 GDP Current year divided by Inflation factor for the year 4 Major Statistical Indicators (BIR Statistics Division) 5 Estimate based on fist quarter average monthly inflation rates  - 71 - Annex 4 TAX COMPUTERIZATION PROJECT ICR Financial Evaluation Assumptions for the Project's Financial Rate of Return 1. All costs and expenses are stated at actual or current prices when they are incurred and have been restated at 1992 prices. 2. All incremental costs during the project period are included in the project costs. 3. Foreign exchange transactions are converted into PhP at the exchange rate prevailing on the transaction dates. Taxes and duties are excluded. 4. The project average useful life is estimated at 5 years. However, since different project components were not in operation by the fifth year , the project useful life is assumed to be completed beyond the year 2001 as initially estimated.  - 72 - Final Versiop - February 2, 2000 Republic of the Philippines Department of Finance Bureau of Internal Revenue National Office Building BIR Road, Diliman 1100 ureau o tema Kevenue BIR Implementation Completion Report * * S * * . * *0 BIR TaC Computerization Project World Bank Loan No. 3603 PH : *.