China - Rural Industrial Technology (SPARK) Project

Document of The 'World Bank FOR OFFICIAL USE ONLY Report No. 18126 IMPLEMENTATION COMPLET-ION REPORT CHINA RURAL INDUSTRIAL TECHNOLOGY (SPARK) PROJECT (LOAN 3274-CHA/CR]EDIT 2186-CHA/JGF TF27230-CHAO June 30, 1998 Private Sector Development 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 Currency = Renminbi . Cunrency Unit = Yuan (Y) 1988 US$1 = Y 3.71 1989 US$1 = Y 3.71 1990 US$I = Y 4.72 1991 US$1 = Y 5.22 1992 US$1 = Y 5.38 1993 US$1 = Y 5.45 1994 US$1 = Y 8.70 1995 US$1 = Y 8.31 1996 US$1 = Y 8.30 1997 US$1 = Y 8.30 FISCAL YEAR January I - December 31 WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS ABC - Agricultural Bank of China CIB - China lnvestment Bank CIEC - China International Economic Consultants EIRR - Economic Internal Rate of Return FIRR - Financial Internal Rate of Return ICR - Implementation Completion Report ISTIC - Institute of Scientific and Technical Information of China ISTCs - Institutes of Scientific and Technical Information JEDIC - Jilin Province Economic Development Investment Corporation JGF - Japanese Grant Facility JPTIC - Jilin Province Trust and Investment Corporation MIS - Management Information System MOF - Ministry of Finance MOST - Ministry of Science and Technology PFIs - Participating Financial Institutions SOEs - State-Owned Enterprises SPO - Spark Program Office SSTC - State Science and Technology Commission STB - Science and Technology Bureau TA - Technical Assistance TVEs - Town & Village Enterprises Vice President Jean Michel Severino, EAP Country Director Yukon Huang, EACCF Sector Manager Hoon Mok Chung, EASPS Task Manager. Helen Chan, Operations Officer, EASFS  FOR OFFICIAL USE ONLY CONTENTS Preface .............. iii Evaluation Summary .v PART I: PROJECT IMPLEMENTATION ASSESSMENT ........................1 A. Statement/Evaluation of Objectives .1 B. Achievement of Project Objectives .4 C. Major Factors Affecting the Project .13 D. Project Sustainability .17 E. Bank Group Performance .18 F. Borrower Performance .19 G. Assessment of Outcome ................... 19 H. Future Operation ................... 20 I. Key Lessons Leamed .20 PART II: STATISTICAL TABLES .23 Table 1: Summary of Assessments .23 Table 2: Related Bank Group Loans/Credits .24 Table 3: Project Timetable .25 Table 4: Loan/Credit/JGF Disbursement: Cumulative Estimate and Actual .26 Table 5: Key Indicators for Project Implementation .26 Table 6: Key Indicators For Project Operations .28 Table 7: Studies included in Project .28 Table 8a: Project Costs in Yuan Million .29 Table 8b: Project Costs in US$ million .30 Table 8c: Project Financing ................. 30 Table 9: Economic Costs and Benefits .31 Table 10: Status of Legal Covenants .32 Table 11: Compliance with Opeiational Manual Statements .32 Table 12: Bank Group Resources: Staff Inputs .33 Table 13: Bank Group Resources: Missions .33 Appendix A: ICR Mission's Aide Menioire. 34 Appendix B: Borrower's Contribution to the ICR .62 Appendix C: List of Subprojects Financed under the Loan and Credit . ..113 Appendix D: Agricultural Bank of China-Financial Statements for Years Ending December 31, 1994-96 ... .......116 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.  - ii - Appendix E: China Investment Bank-Comparative (Projected And Actuals) Financial Statements for Years Ending December 31, 1990-97 ........ 117 Appendix F: Jilin Province Trust And Investment Corporation-Comparative (Projected And Actuals) Financial Statements for Years Ending December 31, 1990-97 ............................................ 118  - iii - IMPLEMENTATION COMPLETION REPORT CHINA RURAL INDUSTRIAL TECHNOLOGY (SPARK) PROJECT (LOAN 3274-CHA /CREDIT 2186-CHA /JGF TF27230-CHA) PREFACE This is the Implementation Completion Report (ICR) of the Rural Industrial Technology (Spark) Project, for whichi Loan 3274-CHA in the amount of US$50 million and Credit 2186-CHA in the amount of SDR 45.1 million (US$64.3 million equivalent) were approved on December 4, 1990 and made effective on June 16, 1991. The project was cofinanced with a Japanese Grant Facility (JGF 27230-CHA) which was effective on the same date. The JGF was closed on December 31, 1996. Its last disbursement took place on April 3, 1997. The loan and credit were closed on September 30, 1997 and were fully disbursed. The last disbursement of the loan took place on December 22, 1997 and the last disbursement of the credit took place on December 31, 1997. The ICR was prepared by Helen Chan (Task Manager) in the Financial Sector Development Unit, East Asia and Pacific Region and Naiqin Lu, Operations Officer in the Resident Mission of China, and reviewed by Messrs. Hoon Mok Chung, Manager, Private Sector Development Unit, East Asia and Pacific Region and Yukon Huang, Director, China Country Unit. The boirrower provided comments that are included as an appendix to the ICR. Preparation of this ICR began during the Bank Group's final supervision mission in July/August 1997. The ICR is based on discussions and findings at the final supervision mission and discussions at the end of April 1998 in Beijing, and documentation in the project file. The borrower contributed to the preparation of the ICR by preparing relevant data and information, contributing views expressed in the mission's aide memoire, preparing its own evaluation of the project's execution and initial preparation, and commenting on the draft ICR.   v - CHINA RURAL INDUSTRLIL TECHNOLOGY (SPARK) PROJECT (LOAN 3274-CHA/CREDIT 2186-CHA /JGF TF27230-CHA) EVA]LUATION SUMMARY Background 1. With the emergence of a rapidly growing and dynamic rural nonstate enterprise sector in the early 1980s, and with the Chinese Government's determination to be more active in utilizing science and technology developed in China in the real sector, the Ministry of Science and Technology, (MOST)1 initiated a pilot Spark program in early 1985. The Spark program became a nationwide program in 1986. Its overall objective is to help transfer technological and managerial knowledge from the more advanced sectors to rural enterprises in support of continued growth and development in the nonstate rural enterprise sector, mostly town and village enterprises (TVEs) and to help increase output and employment. This project was the first Bank Group-supported operation in China specifically oriented to rural nonstate industry. Spark is not an acronym, it signifies that "one small spark can start a prairie fire," a reference to the anticipated catalytic effect of the program on rural enterprise development. Project Objectives 2. The project was designed to: (a) provide institutional support services to the rural enterprises through the provision of training, technical and management information, technical evaluation methodology, anld consultant services to the national Spark program, but some support services were targeted to the three project areas: Jiangsu and Jilin Provinces, and Chongming Island in the Shanghai municipality; and (b) provide financial support to China's ongoing Spark program in the three project areas. The project also emphasized the dissemination of technological and managerial knowledge acquired through project activities to other rural enterprises which are not direct project beneficiaries. 3. The project comprised four components: (a) an enterprise modernization component which provided commercial and precommercial credits to rural nonstate industrial enterprises; (b) a Spark training component to help upgrade training programs and Spark training centers; (c) a technology information component to help design a TVE-focused information system at the Institute of Scientific and Technical Information of China (ISTIC) and 16 local Institutes of Scientific and Technical Information (ISTIs); In the early 1998 government reorganization, the State Science and Technology Commission (SSTC) was renamed as the Ministry of Science and Technology.  - vi - and (d) an institutional development component for a Management Information System (MIS) development, technical evaluation training development for Spark program offices throughout China, and financial institutions training. 4. Component (a) was a line of credit operation through participating financial institutions (PFIs). Components (b), (c) and (d) were technical assistance activities which offered training, construction and equipping training centers, development of and equipment needed for information services. Principal target beneficiaries were the Spark Program Office (SPO) in MOST, Spark program offices in local Science and Technology Bureaus (STBs), ISTIC, local ISTIs, TVEs and county collectives. Implementation Experience and Results 5. Overall, the implementation experience and results of the project have been positive. Despite a very slow start, all project components were satisfactorily completed. About 98 percent of the project funds were disbursed within the original closing date of September 30, 1996. The closing date was extended for one year, mainly to complete disbursements for the precommercial subprojects. All project objectives were met, but subloan repayments remain low. Technical assistance components had very positive effects on rural nonstate enterprises development. Some subprojects and some technical assistance activities had substantive dissemination effects on nonproject beneficiaries. The above positive outcome notwithstanding, one disappointment has been the poor subproject performances and low subloan repayment status. This possibly reflects a countrywide tendency, highlighting the seriousness of financial and enterprise development issues. 6. Technical Assistance Components. About 10 percent of the project fund went to institutional support. Achievement of this objective was substantial. All project activities were completed, and knowledge acquired in the project has been applied to other Spark program activities far beyond the project scope during implementation and is expected to be sustainable. The design of the three technical assistance components was appropriate for sustainable development of the nonstate rural industrial sector. As a result of the project, rural nonstate enterprises have access to technical and marketing information, consultant services, and training opportunities that were either nonexistent or inadequate at the time of project preparation. Government institutions, such as the Spark Program Offices in Beijing and the three project areas, ISTIC, lower levels ISTIs, and training centers, are now much better equipped to serve the rural enterprise sector both in terms of range and quality of services. 7. Under Spark training, a total of 34 training courses were conducted for 1,500 trainees, four modem training centers were established (Nanjing, Jiangyin, Changchun, and Chongming) and modem equipment in computer, video production facilities, language and scientific laboratories, and infornation systems were procured for the four Spark training centers. The four training centers evolved into key training centers in the Spark program system. As the TVE sector grew rapidly in the past decade, training for  - vii - rural enterprises was very much demand-driven. The Spark training program had responded appropriately using teaching methodologies learned and training materials developed in the project, such as the instructional packages, teaching materials, curriculum, and audio/video productions. The beneficiaries, including rural enterprises, Spark training centers and technical colleges, found these materials very useful and used them frequently. 8. A computerized technical information system was set up in the ISTIC, interlinked with 16 local ISTIs in the three project areas, with thousands of technical databases for rural enterprises. These networked systems provide technical, economic, marketing, and sales channel information to TVEs and have been utilized frequently. Subsequently, in response to advances in and growing use of personal computers and popularity of the internet, ISTIC developed a nationwide network information resource system in 1997. It focuses primarily on science and technology information, but also includes economic, cultural and educational information. Some services are fee-based. This development was not originally anticipated at project appraisal. It reflected the adaptive approach of the Spark program and the catalytic role the project has played. The project also helped to produce broadcast-quality videos of "Spark Science and Technology" programs targeting TVEIs and farmers. They featured mature technology recently adopted by both TVEs and farmers nationwide. These videos are continuously being prepared by ISTIC and the programs are broadcasted twice a week on national, provincial, and local TV stations and are very well received by audiences leading to numerous requests for follow-up information on featured technologies. A consulting company was established at ISTIC to provide consultant services to rural enterprises. This project has thus greatly enhanced consultant service development in China. 9. A management information system (MIS) was set up for the national Spark program in MOST and the three project areas. The MIS provides detailed information on Spark subprojects, from their applications for funds to project completion. However, financial information of subprojects in terms of profitability, subloan repayments, and sustainability are inadequate. The MIS therefore needs improvements in order to serve the Spark program more effectively. 10. The project offered technical evaluation training to staff in national, provincial and local Spark program offices (SPOs). This training was to provide SPO staff with technology evaluation skills so as to approach Spark project evaluation in a systematic way, and to equip them with analytical techniques and sources of information so that they could offer quality help to rural enterprises. 11. The project also provided training for PFIs. In 1993, the Jiangsu ABC and CIB went to Philippines for project appraisal training at a Philippine financial institution. Shanghai ABC had local training in 1993 as well. JPTIC arranged training at its own cost. All PFIs attended a course on technical evaluation arranged in the project. In retrospect, training provided to PFIs was too little too late. If we expect the PFIs to perform satisfactorily, either they should have received sufficient training prior to project  - viii - implementation or they should have been established financial internediaries which had sufficient experience in sound project lending and good portfolio management. In addition, the macroeconomic framework has to be in place. 12. All technical assistance activities in Beijing were cofinanced with a Japanese Grant Facility (JGF). Due to substantial appreciation2 of Japanese Yen against US dollars from 1991 to 1995, there were additional resources in the JGF in terms of US dollars. Seven additional TA activities, largely extensions of the original TA activities, were undertaken in 1996. These activities included multimedia and CNC technology, pilot consultant services, case studies to extend project-oriented TA activities to Spark intensive areas and an overall assessment of the Spark program from 1986 to 1996. Most of these activities were trial applications of knowledge and techniques acquired in the project. Lessons learned from these applications would greatly benefit future applications and dissemination. 13. Enterprise Modernization Component. About 90 percent of the project funds were relent to 92 rural nonstate enterprises through- four PFIs in the commercial credit subcomponent and 16 precommercial subprojects through the two Jiangsu PFIs which acted as financial agents. All project funds have been fully utilized. 14. Of the 108 rural nonstate enterprises supported by the project, 83 were TVEs and 25 county collectives. A total of 86 were existing enterprises, while 22 were new establishments. Funds were committed over a period of six years from 1991 to 1996, much slower than appraisal estimates which expected all subprojects to be committed by December 1992. The subprojects were first prescreened by local STBs for their eligibility under the Spark program, then they were forwarded to the PFIs for project appraisal. The average subloan size was US$950,000, and 13,600 new jobs were directly created by these subprojects. 15. By project closing date of September 30, 1997, subproject results have been mixed. Among the commercial subprojects, about 38 percent reached operational capacity and operated profitably. Yet some are not repaying their loans to PFIs. About 39 percent were not profitable due to cost overruns during implementation, delays in operational startup, delays in obtaining sufficient working capital, and tight liquidity (due mainly to the typical triangular debts among Chinese enterprises). If some of these could be resolved, like availability of additional funding from project sponsors and guarantors, some of these enterprises would have an opportunity to turn around and be profitable in two to three years. The remaining 23 percent have serious problems, such as failure to obtain counterpart funds, prolonged delays in construction, market changes, poor product quality, technical incompetence and poor management. Out of these 23 percent, 10 percent had stopped production, 9 percent have serious financial difficulties and 4 percent 2 There was an additional USS570,000 more in the JGF account in 1995 due to the continuous Yen appreciation relative to the US dollar between 1991 and 1995 (from ' 142 in 1990 to 3f10S in 1995).  - ix- have been under construction since 1995, and the outlook for them does not look promising. Chances of full loan recoveiy are unlikely. 16. A major objective of the national Spark program was the diffusion of technical and managerial knowledge from successful projects to other nonproject beneficiaries. In the Bank Group project, this objective was realized through some successful subprojects. At the end of 1997, there were about 15 subprojects that were able to achieve this objective. Most successful subprojects became pillar industries in their respective "Spark intensive areas." They led to vertical and horizontal integration of related industries either in their own localities or extended to other provinces. One Jiangsu beneficiary enterprise joined the national Spark West Exploration Program in setting up plants in the Sichuan province applying the managerial and technological know-how nurtured in its Jiangsu plant. With technology mastered through a Spark subproject, a Jilin group company helped to design machinery in plants in Zhejiang province. A small Chongming beneficiary enterprise helped to set up four plants and has developed cooperative production partnerships with eight plants in Chongming and Shanghai. These are successful examples of Spark's diffusion activities. SPOs expressed that without the Bank Group project, these projects probably would not be implemented as availability of capital remained a constraint to rural eilterprises. 17. Out of a total of 92 commercial subprojects, the PFIs reevaluated financial and economic viability of nine subprojects after subproject completion. Subprojects were selected for calculation of ex-post EIRR and FIRR based on the following criteria: (a) a minimum of two subprojects from each of the PFIs; (b) enterprises from different sectors and different cities of the three project areas; (c) an even mix of successful and less successful subprojects; (d) half in repayment stage; and (e) subprojects of different project sizes. The nine subprojects accounted for about 10 percent by number of commercial subprojects and 13 percent of the loan/credit amount disbursed to the commercial subprojects. The ex-ante economic internal rates of return (EIRRs) ranged from 19 percent to 73 percent while the ex-post EIRR ranged from 1 percent to 84 percent. The ex-ante financial internal rates of return (FIRRs) ranged from 18 percent to 49 percent while the ex-post FIRR ranged from 8 percent to 31 percent. Actual results were generally lower than original estimates because of project delays, higher input prices and lower output prices than appraisal estimates. 18. The financial institutions were the central actors in determining project sustainability, as they appraised and supervised 90 percent of project funds to final beneficiaries. Generally, financial institutions were inadequate in their project appraisal and supervision. Their unsatisfactory performance was partly due to existing financial, enterprise, and legal systems; policies and regulations in China. The current systems and policies did not offer financial institutions much freedom nor incentives in exercising their decisions in key operating areas such as: bad debt provisions, subloan restructuring and rescheduling, legal proceedings against subborrowers in case of deliberate default, * legal obligations of guarantors, bankruptcy, or other actions that may be needed in case of default.  x - 19. Despite operational success of a substantial number of subprojects, subloan repayment has been low. As of December 31, 1997, principal subloan repayment was only 18 percent of total principal due and interest payment 31 percent of cumulative interest accrued. Jiangsu CIB's subloan repayment reached 65 percent of principal due, Jiangsu ABC 15 percent, JPTIC 20 percent, and Shanghai ABC 2 percent at the end of 1997. Many subborrowers were not able to repay on schedule mainly because of: (a) the original subloan repayment terms were very short, usually only a one year grace period; (b) nonviable projects resulted from inadequate market analysis; (c) cost overruns were frequently due to unrealistic cost estimates; and (d) local government interfered in internal enterprise management, such as requests to absorb bankrupt enterprises. Another issue is enterprise culture. Loan repayment was often not the top priority of borrowing enterprises. Additionally, loan supervision by financial institutions was often secondary to new lending. Given this background, if the PFIs reschedule repayment terms based on realistic subproject cash flows, half of the enterprises could repay their loans in the next two to three years, including some that were not operating profitably in 1997. 20. The precommercial subprojects are risk-sharing ventures that support technology development in enterprises preceding full-scale commercialization. As the concept was innovative, there was a lack of understanding of the delivery mechanism in the early stage of project implementation. Together with many other issues on other project components, this component was held in abeyance while SPO and Jiangsu STB explored different delivery alternatives. A draft contract, and roles and responsibilities of different parties were finally agreed in mid-1994 and commitments under this subcomponent started in August 1994, with 16 subprojects committed between 1994 and 1996. All subborrowers are expected to repay the subloans. As of December 1997, ten had completed implementation and six, approved in 1996, had started implementation. Jiangsu STB expected that about 50 percent could become commercial subprojects in the medium term. Lessons learned from one of the more promising precommercial subproject are that: (a) the sponsor has prior successful experience in similar fields; (b) the sponsor has strong technical support, not only from the research institute that provided the technology, but also from national expertise; and (c) market assessment is correct. Failing precommercial ventures are missing one or more of these elements. Summary of Findings, Future Operations, and Key Lessons Learned 21. The project's simultaneous focus on multiple target groups, different government entities, enterprises and financial institutions in various provinces and localities made the project very complex and difficult to manage and supervise. It was particularly difficult because: (a) all project beneficiaries (except CIB) were new Bank Group borrowers; (b) the proper financial and enterprise sector framework was not in place; and (c) expectations that financial institutions perform as bona fide lending institutions and that enterprises, including the less-regulated TVEs, behave truly as independent entities were unrealistic. Another major factor that made the project very difficult to implement was the three-year elapse time between project appraisal in June 1998 to project effectiveness in June 1991. Economic and financial environment on which the project was built on had  - xi - changed substantially during that periodl. Performance of the initial batch of subprojects was less satisfactory because of this substantial delay. 22. At the time the project was appraised, it was expected that it would be the first of a series of subsequent operations to support the rural nonstate industrial sector. However, as a result of the experiences gained from this project and other industrial projects and lines of credit operations in China, the Government and the Bank Group realized that the generic issues in the financial and enterprise sectors needed to be resolved before committing to further lending operations to enterprises through financial institutions. This means that subsequent operations have been significantly curtailed. 23. There are six main lessons to be learned from this operation: (a) Successful execution of technical assistance requires full commitment and participation of all stakeholders throughout the process. It was achieved in this project. It shows that our client agencies are very receptive to technical assistance if TA activities are well conceived. (b) National ownership is important. Spark is a national program and the Bank Group is only helping to strengthen it. The Ministry of Science and Technology takes prides in the Spark accomplishments. (c) Sound organizational setup and dynamic leadership for project implementation are essential to project success. This is particularly important for projects that have many implementing agencies in different project areas. (d) The Bank Group's project supervision missions should be staffed adequately with technical staff/consultants throughout implementation. Sufficient budget should be allocated for this purpose. This was unfortunately not the case. (e) Financial institutions should strengthen their expertise in subproject evaluation, including adequate selection criteria, market assessment, cost estimates, realistic implementation schedules, subborrowers' implementation and repayment capabilities, counterpart funding requirements and availability, and awareness of environment issues. However, even if these issues were fully addresseid by financial institutions, enterprises undertaking projects in the context of limited financial and enterprise reforms still face many constraints under the current national policy framework. The Bank Group should recognize these constraints in project preparation and try to seek solutions. If the Banik Group realizes that these issues could not be solved through individual investment project in the absence of macroeconomic structural adjustmnents, then they should be clearly spelled out in project documents in order to avoid unrealistic expectations.  - xii - (f) For lines of credit operations using financial intermediaries to be successful, some basic structural changes in the economy in the financial and enterprise sectors are needed: (i) stricter enforcement of laws, such as commercial banking law, company law, and bankruptcy law; (ii) enterprise behavior in terms of honoring loan repayment obligations and ensuring timely repayments; (iii) institutional behavior and performance standards of the financial institutions, such as subproject selection criteria, appraisal standards, portfolio management, adequacy of bad debt provision, etc.; and (iv) local governments' attitudes to financial institutions and enterprises, i.e. local governments treat banks as conduits of funds to enterprises, and at the same time, they are more concerned with enterprises' social obligations and taxes rather than causing enterprises to repay bank loans.  - I - CHINA RURAL INDUSTRLAL TECHNOLOGY (SPARK) PROJECT (LOAN 3274-CHA /CREDIT 2186-CHA/JGF TF27230-CHA) PART 1: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT/EVALUATION OF OBJECTIVES 1. Background. With the emergence of a rapidly growing and dynamic rural nonstate enterprise sector in early 1980s, and with the Chinese Government's determination to be more active in utilizing science and technology developed in China in the real sector, the Ministry of Science and Technology (MOST)1 initiated a pilot Spark program in early 1985. The Spark program became a nationwide program in 1986. Its overall objective was to help transfer technological and managerial knowledge from the more advanced industrial sector to rurl enterprises in support of continued growth and development in the nonstate rural enterprise sector, mostly town and village enterprises (TVEs) and to help increase output and employment. This project was the first Bank Group-supported operation in China specifically oriented to rural nonstate industry. Spark was not an acronym, it signifies that "one small spark can start a prairie fire," a reference to the anticipated catalytic effect of the program on rural enterprise development. 2. Project Objectives. The project was designed to provide: (a) institutional support services to the rural enterprises including: (i) transfer effective training tools to the Spark training program, (ii) transfer to the Institute of Scientific and Technical Information of China (ISTIC) and local Institute Scientific and Technical Information (ISTCs) of a system of new approaches to build up relevant databases and delivery information services to rural enterprises, (iii) development of consultant services for town and village enterprises (TVEs), (iv) development of tools for technology evaluation, (v) design ajid introduction of an effective and responsive management information system for the national Spark program, and (vi) strengthening of the Spark Program offices in the MOST system and the ISTIC system by providing necessary tools; and (b) financial support to China's ongoing Spark program in three diverse areas of China: Jiangsu province., Jilin province, and Chongming Island in the Shanghai municipality. Jiangsu has the most dynamic TVE sector in China and is increasingly technically sophisticated and export-oriented. Jilin is an In the early 1998 govemnment reorganization, the State Science and Technology Commission (SSTC) was renamed as the Ministry of Science and Technology.  -2 - important agricultural center, and its economy is dominated by state-owned enterprises (SOEs) in heavy industry. Chongming is an island off the coast of Shanghai, which, despite its proximity to Shanghai, remains an underdeveloped agricultural island. The project also pioneered the concept of risk-sharing precommercial credits in the Jiangsu province. The project provided credits to eligible rural enterprises through four participating financial institutions (PFIs). Results obtained in both objectives were expected to be replicable or disseminated to other rural enterprises all over the country. 3. Objective (a) was fulfilled satisfactorily, and the outcome far exceeded appraisal estimates. The many innovative features introduced in the project were not only successfully implemented and put into application, but were also used in activities which were not anticipated at project appraisal. This demonstrates the fact that technical assistance, if well implemented, provides more value added to our clients than investment financing. Objective (b) was considered achieved even though some enterprises experienced financial and operational difficulties and most have repayment problems. 4. Project Description. The project comprised four components: (a) enterprise modernization component, its two subcomponents were: (i) commercial credit subcomponent: provided credits to rural nonstate enterprises through four PFIs, the Shanghai and Jiangsu branches of the Agricultural Bank of China (ABC), the Jiangsu branch of the China Investment Bank (CIB), and the Jilin Province Trust and Investment Corporation (JPTIC), and (ii) precommercial credit subcomponent: provided credits for risk-sharing precommercial activities in the Jiangsu province through the two Jiangsu PFIs; (b) Spark training component: supported technical assistance and training programs for key managers and trainers at the national and lower-level Spark training centers; (c) . technology information (TI) component: supported the design of a TVE- focused information system at ISTIC and 16 Institutes of Scientific and Technical Information (ISTIs) in the three project areas, together with pilot programs in diffusion and information consultancy; and (d) institutional development component, its three subcomponents were: (i) management information system (MIS) subcomponent, to design and implement a small, user-friendly national Spark Program MIS in  - 3 - Jiangsu and Jilin, and at MOST, with the intention of ultimately creating a shared, interlinked MIS for the entire Spark program, (ii) technology evaluation (TE) subcomponent-to design and deliver training programs on the evaluation of industrial subprojects incorporating innovative technology for the Spark program offices all over China, and (iii) financial institution training-to provide training to project staff in the PFIs for preparation amd management of TVE projects. 5. Component (a) was a line of credit operation through PFIs. Components (b), (c) and (d) were technical assistance activities which offered training, construction and equipment for training centers, development of and equipment needed for infornation services. Principal target beneficiaries were the Spark Program Office (SPO) in MOST, Spark program offices in local Science and Technology Bureaus (STBs), ISTIC, local ISTIs, TVEs and county collectives. 6. Main Features of the Bank loan, IDA credit and the Japanese Grant Facility (JGF). The People's Republic of China was the borrower of the Bank loan and IDA credit and the beneficiary of the Japanese Grant Facility. The Bank loan has 20 years of maturity with five years of grace and carries the Bank's standard variable interest rate. The IDA credit has 35 years of maturity with 10 year grace and has the standard IDA terms. All proceeds from the Bank loan and the IDA credit were relent to Jiangsu and Jilin provinces and the Shanghai municipality, of which: (a) 85 percent of the IDA credit and 100 percent of the Bank loan were for the commercial credit subcomponent. Bank proceeds were relent to ABC, CB and JPTIC for 15 years including 5 years of grace for the commercial credit component. The PFIs were allowed to roll over the funds; (b) 5 percent of the credit was allocated to precommercial credit activities in the Jiangsu province. The Jianigsu provincial government bore the credit risk and the Jiangsu PFIs acted as financial agents, receiving fees for their services; and (c) 10 percent of the credit was for technical assistance activities in components (b), (c), and (d). The JGF provided support to activities at the national Spark program in Beijing. The Spark Program Office (SPO) in MOST was in charge of all technical activities in Beijing and coordinated all project activities in the three project areas. 7. Onlending Arrangements of the Enterprise Modernization Component. Onlending arrangements of the two subcomponents were:  -4 - (a) Commercial Credit Subcomponent. The PFIs onlent the Bank Group proceeds to rural nonstate enterprises. The subloans were normally between 3 and 7 years, and under no circumstances to exceed 12 years. In this subcomponent, the PFIs bore the credit risks. There were foreign currency and local currency subloans: (i) For foreign currency subloans, the PFIs relent Bank Group proceeds to subborrowers at the Bank Group's variable rate plus a minimum of 2 percent. Subborrowers were responsible for the foreign exchange risk between the dollar and the Renminbi. (ii) For local currency subloans, the PFIs relent Bank Group proceeds to subborrowers at an interest rate of not less than the prevailing rate on fixed assets loans for technical renovation projects of similar maturity set by the People's Bank of China. At the start of project, the Bank Group reviewed all subproject appraisal reports submitted by each PFI. The free limit for each PFI began when the Bank Group was satisfied with the quality of appraisal reports. (b) Precommercial Credit Subcomponent. The two Jiangsu PFIs managed Bank Group funds for the Jiangsu provincial government and received an agent fee for their services. The Jiangsu government bore the credit risk. The interest rates on the precommercial subprojects were higher than prevailing commercial rates and were different for each subproject depending on their profitability rates after subproject completion. 8. Eligibility Criteria of Subprojects. The subprojects financed with the Bank loan and IDA credit were to meet the following eligibility criteria: (a) industrial projects, including agroprocessing, (b) TVEs, county collectives or joint ventures controlled by TVEs; (c) STBs justification that proposed projects have a technological content beyond the standard investment, expansion, or modernization; (d) economic internal rates of return (EIRRs) of at least 12 percent for subloans exceeding US$300,000 and financial internal rates of returns (FIRRs) of at least 12 percent for all subloans; and (e) subborrowers to finance at least 20 percent of the subproject investment costs. B. ACHIEVEMENT OF PROJECT OBJECTIVES Institutional Support to the National Spark Program. 9. About 10 percent of the project funds went to technical assistance activities for institutional support. Achievement of this objective was substantial. All project activities were completed, and knowledge gained during the project has been applied to other Spark program activities far beyond the project scope during implementation and are expected to be sustainable. The design of the three technical assistance components was appropriate for sustainable development of the nonstate rural industrial sector. As a result  - 5 - of the project, rural nonstate enterprises have access to technical and marketing information, consultant services, and training opportunities that were either nonexistent or inadequate at project preparation. Government institutions, such as the SPOs in the national Spark program, ISTIC, lower levels ISTIs, and training centers are now much better equipped to serve the rural enterprise sector both in terms of range and quality of services. From 1992 to 1997, a total of 54 courses/workshops were conducted in China for 2,100 trainees, along with 10 overseas training/intermships for 80 trainees. 10. The Spark training component started in April 1992 with a three-year, 41 person-month contract with ORT, a British management training institute. The objective was to strengthen and modernize the Spark training program to meet TVEs' development needs more effectively. Major emphasis was placed on training demand assessment, training program evaluation and curriculum development, participatory training methodology, design of training program for microcomputer applications, multimedia application in training, trainers' training, and design and production of multimedia instructional packages for adult training. During the implementation period, 34 workshops and training seminars were conducted by foreign and local experts in the four training centers and 4 overseas study tours were undertaken. Participants totaled 277 and all were managerial, training, andL technical staff from more than 20 provinces in the Spark training system. Their participation resulted from better planning in training programs and training development in Spark training centers. The SPO prepared a Study on the Long-Term Training Program as required in the project at the end of 1993. Since the Spark training component was delayed for more than a year, the study was revised at the end of 1994 to reassess contributions of the training component and set priorities for further short- and long-term improvement of Spark training. It has also helped to guide the Spark training program development over the next few years. At the same time, as the TVE sector grew rapidly in the past decade, training for rural enterprises was very much demand-driven. The Spark training program had responded appropriately using teaching methodologies learned and training materials developed in the project, such as technical instructional packages, manuals, curriculum, and audio/video productions. The beneficiaries, including rural enterprises, Spark training centers and technical colleges, foumd these materials very useful and used them frequently. 11. Multimedia Instructional Packages. Five instructional packages using multimedia technology were prepared from 1994 to 1997-"CNC machine tool" and "human resources development in Tv'Es" packages in Jiangsu, "modified starch" in Jilin, "household electronics" in Chongming, and "DPC applications in cotton production" in Beijing. All instructional packages, except the "household electronic appliances" instructional package, were found to be very useful for users. Continuous upgrading has been underway. They were used for adult education in the Spark training centers, in technical colleges, and by industrial enterprises. As TVEs mushroomed in the 1990s, the Jiangsu SPO understood that human resources development in the TVE sector was essential to TVEs' healthy growth in the future, it published a book on "Human Resources Management for Small and Medium Enterprises" in 1996, based on the  - 6 - "human resources development in TVEs" instructional package, so that the general public could access the knowledge easier. The book is sold in bookstores and is very popular. 12. Training Centers. The IDA credit provided 50 percent of the construction cost and provided essential equipment for four training centers. It helped to equip different laboratories, such as microcomputer applications, language, video production, instructional production, video production laboratories. Other than the 34 workshops and courses offered under the Spark training component, all training courses in these centers were financed either by the provincial/municipal government or by training fees charged to the trainees. Of the four training centers targeted to be financed under the credit, only the Nanjing training center did not utilize its allocated funds for building construction, as it found alternative financing. However, the training center was equipped with equipment from the credit. The training it offered is intensive, emphasizing computer applications, ISO 9000, technical assessment, financial management, and auditing. The Chongming Training Center on Chongming Island was the first to be completed among the three training centers whose construction was financed under the credit. It started training at the end of 1992. Despite its geographic disadvantage (no road linked it to mainland China, thus only sea transportation is available), it evolved into a major regional training center among the Spark training facilities. It provides many types of adult education with emphasis on computer application, finance and accounting, English language, mechanical automation, public relations, documentation management, air conditioner installation techniques, automatic instruments, construction quality, legal regulations, etc. Many participants were from Shanghai and other provinces. By 1997, it reached nearly 100 percent capacity utilization and was able to reach a cost recovery rate (direct training cost) of 40 percent. It aims at full cost recovery in the next few years. This is a significant improvement over other Spark training centers which rely principally on government financing of their training programs. Construction of the new Jiangyin training center in Jiangsu was completed only in late 1996, but it had utilized its older building to carry out many training activities since 1993, principally in language, accounting, and hotel services. The 1997 mission advised it to place more emphasis on training such as marketing, project risk analysis and project management. Construction of the new Changchun training center in Jilin was completed in early 1996. Training activities since 1993 were conducted in its old building. It emphasizes finance and auditing, Spark project management, computer applications, and enterprise management. Utilization rate was high, but cost recovery was low. Training programs took many forms and varied in duration. Most participants were technical, financial, and managerial staff from rural enterprises. Post training evaluations of the training programs was very positive. 13. The technology information component was designed to: (a) set up a technology inforrnation system, including databases and computer networks--the information organization and retrieval subcomponent; (b) develop an extension mechanism to transfer information more effectively to TVEs through audiovisual, publications, training, and development of a fee-based, service-oriented consultant services for TVEs--the information diffusion subcomponent; and (c) strengthen information transfer process--the  - 7 - information design and structure subcomponent. A contract was signed with the China International Economic Consultants (CIEC) in March 1993. A total of 14 person-months of technical experts and advisors worked with ISTIC counterpart group. Five training courses were conducted by foreign consultants on technical information systems, information analysis and synthesis, teichnical information management, and development of fee-based consultant services for senior managerial and technical staff at ISTIC and local ISTIs. The design and implementation plan of the technology information system was completed in October 1993, and equipment was procured for ISTIC and local ISTIs to help information transfer. Subsequently, this led to: (a) establishment of computerized technical information systems in Jiangsu, Jilin and Chongming ISTIs and 16 local ISTIs. This is a major objective of the component. Several thousand databases covering tens of thousands of rural enterprises were developed and were linked to the network. These networked systems provide technical, economic, marketing, and sales channel information to TVEs and have been utilized frequently. Their current emphasis is on expancling information sources, establishing regional networks, and further development of information and consulting services; (b) due to advances in and growing use of personal computers and popularity of the internet, ISTIC de veloped the Chinainfo, a nationwide network information resource system, in 1997. It focuses primarily on science and technology information, but also includes economic, cultural and educational information. This center has 26 local networks in different provinces and municipalities across the country. It also provides domestic and international on-line retrieval and verification services. Retrieval requests exceeded 1,000 in 1997. Some services offered are fee-based. Its income has been used to upgrade their facilities and information content. Chinainfo, a legally independent entity, has been financially self-sufficient since 1997. This was not originally anticipated at project appraisal. Its establishment reflects the adaptive approach of the Spark program; (c) broadcast-quality video production of the "Spark Science and Technology" programs targeting TVEs and farmers. They feature mature technologies recently adopted both by farmers and rural enterprises nationwide, development of Spark intensive areas and pillar industries, and directions of the Spark program. These programs are currently broadcasted twice a week on national, provincial and local televisions. These programs are very well received by audiences, and there have been many requests to the program asking for additional information about technology introduced in the TV program. This is an excellent tool to diffuse knowledge acquired and technologies mastered in Spark projects. Its popularity, coverage and frequency of use has gone well beyond original project scope;  -8 - (d) production of many high quality audio and video tapes of newly established and matured technologies that could be adopted by farms and rural enterprises, such as poultry and fish raising, fruit and vegetable cultivation, etc. for farmers; and food processing, machine tools, packaging technologies, etc. for rural industries. These tapes are provided to farmers and rural entrepreneurs and are particularly useful in remote areas; (e) establishment of consulting services at ISTIC. A follow-up consulting contract with a foreign consultant firm encouraged further development of domestic consulting services in China. Two senior staff from SPO and ISTIC served internships with the foreign consulting firm to learn hands-on consulting services in the United States. Starting in 1996, ISTIC started to offer technology exchange conferences 30-40 times a year to disseminate technical information to TVEs. These are free and have been attended by over 8,000 TVEs so far. ISTIC also offered market studies and provided management and strategy consulting to TVEs. These are fee-based; and (f) publication of a book by the Spark Program Office on "Guideline to Consulting Services for TVEs" in 1997. This book was published due to the growing demand to develop quality consulting services for TVEs, particularly in the more remote rural areas. 14. Under the Institutional Development Component, the MIS subcomponent started in May 1992 and was completed in August 1994. It was intended to develop Spark program management-oriented databases in MOST, Jiangsu, and Jilin STBs and would become the basis for a more comprehensive MIS for the Spark program at a later stage. Spark project databases were constructed, and a Spark program MIS was developed in MOST and the three project areas. The national MIS was upgraded in 1997. The system provides Spark project applications nationwide, approvals, sectoral and regional distribution, financing, implementation, and the results of Spark subprojects (not only limited to World Bank-financed projects). However, financial information of subprojects in terms of profitability and subloan repayments are inadequate. The MIS therefore need improvement in order to serve the Spark program more effectively. 15. A second subcomponent was the technology evaluation subcomponent which started implementation in August 1993 and was completed in September 1995. This subcomponent focused on training provincial and county SPO staff. This training program was intended to help SPO staff acquire technology evaluation skills, allowing them to approach Spark project evaluations in a systematic way, and equip them with analytical techniques and sources of information so that they could offer quality help to rural enterprises. This is necessary, since in some remote rural areas the local SPO may be the only source of assistance for rural enterprises. Three training courses were conducted for 80 SPO staff in 1993 and 1994, and in September 1996, two training manuals were prepared: the Industrial Project Technical Evaluation for Small and  -9- Medium Enterprises, one for trainers and one for trainees. They are currently used by SPOs in many Spark locations. 16. A third subcomponent was the financial intermediary training subcomponent which provided project appraisal and management to the Jiangsu and Shanghai PFIs. Both Jiangsu ABC and CIB went to Philippines for training in 1992 and 1993 respectively in a course specifically arranged for them by a Philippines financial institution. Shanghai ABC had training in Shanghai in 1993. All PFIs attended the training courses offered in the technology evaluation course in 1993. In retrospect, the training program was too little too laile. If we expect the PFIs to perform satisfactorily, either they should have received sufficient training prior to project implementation; or they should have been well-establ:ished financial intermediaries which had prior experience in project lending and portfolio management. In addition, the macroeconomic framework has to be in place. 17. The Japanese Grant Facility (JGF). All technical assistance activities in Beijing were cofmanced with a Japanese Grant Facility. Due to substantial appreciation2 of Japanese Yen against US dollars from 1991 to 1995, there were additional resources in terms of US dollars in the JGF for the Spark project in 1995. Seven additional TA activities, extensions of the original TA, were undertaken in 1995 and 1996: (a) ORT was awarded another training contract for further study of multimedia and CNC technology. "The DPC (a chemical) application in cotton production" instructional package was the outcome. This teaching material was well received and has been used by cotton farmers, technical colleges, and Spark training centers. (b) A pilot project on consulting services was undertaken. Under the guidance of a foreign consulting firm, four local consulting firms won contracts to provide technical, management, and marketing consultant services to selected rural enterprises in the three project areas. The local consulting firms were selected through national competition, and they had to look for enterprises which were willing to participate and willing to share some of the cost. The main objective of this activity was to help local consultant firms to conduct consulting services to better serve TVEs. Through this pilot project, it could be concluded that consulting services for TVEs in China is still in its infancy stage. (c) Three case studies in Chongming, Jiangsu, and Hunan Spark intensive areas were undertaken. The emphasis of the Spark training, TI, and TE components was largely oriented towards individual projects. These case studies went a 2 There was approximately $570,000 more in the JGF account in terms of US dollars in 1995 due to the continuous Yen appreciation relative to the US dollar between 1991 and 1995 (from 142 in 1990 to 1f05 in 1995).  - 10- step further. They took the knowledge acquired and experiences learned from the three TA components and put them into the context of the entire Spark intensive areas development with the objective of expediting industrialization and development of rural areas which have a high concentration of TVEs. This is important for the Spark program as the program is changing its focus from project-oriented assistance to Spark intensive areas and Spark small areas development. (d) An overall assessment of the Spark program was undertaken in 1996. The book "Evaluation of the Spark program" provided candid assessment and development of the Spark program from 1986 to 1996. The book was well prepared and was a useful tool for further program development. (e) The MIS was upgraded in 1996 to improve capability of the system. These were all small TA consultant contracts, except for the MIS which was a technical upgrade, from US$12,000 to US$100,000. They drew on experiences learned and techniques developed in the project for dissemination to the rural enterprise sector. 18. Originally, 50 percent of JGF was allocated to consultant services and 50 percent to equipment purchase. At project completion, 60 percent of the grant was utilized for consultant services and 40 percent to equipment purchase. 19. Experiences with Consultant Contracts. The Spark project had a total of 15 consultant contracts. The SPO in Beijing was in charge of selection and signing contracts with all the consultant firms on behalf of MOST. The contracts ranged from US$982,000 (the ORT contract) to US$12,000. The two larger ones were with foreign consultant firms. The rest were with Chinese consultant firms, some had foreign consultants as their core team. Over 20 foreign consultants were involved in the Spark project. The project entities generally felt that project coordination of the consultant project was key to the project's success. They felt that foreign consultant firms were better organized than Chinese consultant firms. Their consultants were generally highly qualified, offered advanced ideas and different visions, had better presentation and teaching skills and materials, prepared good reports, and were very dedicated to their work. Foreign consultants recruited by Chinese consultant firms had similar work attitudes and organizational skills. However, the project entities believed that if the foreign consultants had been more knowledgeable about China and the Chinese rural industrial sector, their contributions could have been far greater. Yet, the overall experience was positive. The project entities considered the Chinese consultant firms adequate, but could learn a lot from foreign consultant firms, particularly documentation and dedication. Financial Support to China's Spark Program 20. Fund Utilization. About 90 percent of the project funds were relent to 92 rural nonstate enterprises through four PFIs in the commercial credit subcomponent and 16  - Il - precommercial subprojects through the two Jiangsu PFIs in the precommercial subcomponents. All project funds have been fully utilized. 21. Subproject Characteristics. Out of these 108 subprojects, 83 of the subprojects were TVEs and 25 county collectives. A total of 86 were existing enterprises and 22 were new establishments. The Bank Group approved the subprojects committed over a period of six years from 1991 to 1996 with 60 percent approved between 1994 and 1996, much slower than appraisal estimates which, expected all subproject funds to be committed by December 1992. Subloan commitments were late partly because the project had a difficult and slow start, and partly because the first few subproject appraisal reports from PFIs reviewed by the Bank Group needed substantive clarifications and revisions. About 55 percent of the subproject appraisal reports were reviewed by the Bank Group. The subprojects were first prescreened by local STBs for their eligibility in the Spark program, then they were forwarded to the PFIs for project appraisal. As a result, these 108 subprojects were selected from several hundred prospective applicants. The largest subloan was about US$2.8 million, anid the smallest were the precommercial subprojects of about US$200,000 each, with an average subloan size of US$950,000. Total financing cost amounted to Y 1,902 million3, of which 39 percent was financed with Bank Group loan/credit, 35 percent with loans from local banks, and 26 percent self financing. Over 60 percent of the total project costs was for foreign currency expenditures. From 1993 to 1996, 13,600 new jobs were directly created by these subprojects. 22. Commercial Subprojects lriplementation. Subproject results are mixed. At project completion, 35 subprojects (out of 92) of the commercial subprojects have reached operational capacity and are oiperating profitably. Yet some are not repaying their loans to PFIs. 36 subprojects were not profitable due to cost overruns during implementation, delays in operational startup, delays in obtaining sufficient working capital, and tight liquidity (due mainly to the typical triangular debts among Chinese enterprises). If some of these could be resolved, like availability of additional funding from project sponsors and guarantors, some of these enterprises would have an opportunity to turn around and be profitable in two to three years. 21 subprojects have serious problems, such as failure in obtaining counterpart funds, prolonged delay in construction period, market changes, poor product quality, technical incompetence, and poor management. Ten had stopped production already. Eight subprojects have serious financial difficulties and chances of turning around to be financially viable is unlikely. Three are still under construction since 1995, and chances of full loan recovery in the near future are not promising. 23. A major objective of the national Spark program is the diffusion of technical and managerial knowledge from successful projects to other enterprises. In the Bank Group Spark project, this objective was realized through some successful subprojects. At the end of 1997, there were about 15 subprojects that were able to achieve this objective. There 3 US$232 million at the current exchange of US$I to Y8.2.  - 12 - could be additional demonstration effects when more subprojects become operationally successful. As most subprojects started operations only in 1995 and 1996, it is likely that more diffusion cases could be introduced in the next few years. Most successful subprojects became pillar industries in their respective "Spark intensive areas." They led to vertical and horizontal integration of related industries either in their own localities or extended to other provinces. One Jiangsu beneficiary enterprise joined the national Spark West Exploration Program in setting up plants in Sichuan province applying the managerial and technological know-how nurtured in its Jiangsu plant. With technology mastered through the Spark subproject, a Jilin group company helped to design machinery in plants in Zhejiang province. A small Chongming beneficiary enterprise helped to set up four plants using technology mastered and has developed cooperative production partnerships with eight plants in Chongming and Shanghai. These are successful examples of Spark diffusion activities. The SPOs expressed that without the Bank loan, these projects probably would not be implemented as availability of capital remained a constraint to rural enterprises. Annex 5 of Appendix A shows some representative cases of subproject diffusion. 24. Economic and Financial Internal Rates of Return (EIRR and FIRR) of Commercial Subprojects. Out of a total of 92 commercial subprojects, the PFIs reevaluated financial and economic viability of nine subprojects after subproject completion. Subprojects were selected for calculations of ex-post EIRR and FIRR based on the following criteria: (a) a minimum of two subprojects from each of the PFIs; (b) enterprises from different sectors and different cities of the three project areas; (c) an even mix of successful and less successful subprojects; (d) half in repayment stage; and (e) subprojects of different sizes. The nine subprojects accounted for about 10 percent by number of commercial subprojects and 13 percent of loan/credit amount disbursed to the commercial subprojects. The ex-ante EIRRs ranged from 19 percent to 73 percent while the ex-post EIRR ranged from 1 percent to 84 percent. The ex-ante FIRRs ranged from 18 percent to 49 percent while the ex-post FIRR ranged from 8 percent to 31 percent. Actual results were generally lower than original estimates because of project delays, higher input prices and lower output prices than appraisal estimates. 25. Precommercial Subprojects Implementation. These subprojects are risk- sharing ventures that support technology development in enterprises preceding full-scale commercialization. As the concept was innovative, there was a lack of understanding of the delivery mechanism in the early stage of project implementation. Together with the many issues on the other components, this subcomponent was held in abeyance while SPO and Jiangsu STB explored different delivery alternatives. A draft contract and roles and responsibilities of different parties were finally agreed in mid-1994, and commitments under this subcomponent started in August 1994, with 16 subprojects committed between 1994 and 1996. As of December 1997, ten have completed implementation and six, approved in 1996, have started implementation. Jiangsu STB expected that about 50 percent could become commercial subprojects in the medium term. Experiences learned from one of the more promising precommercial subproject are that: (i) the sponsor has prior successful experience in similar fields; (ii) the sponsor has  - 13- strong technical support, not only from the research institute that provided the technology, but also from national experts; and (iii) market assessment is correct. Failing precommercial ventures are missing one or more of these elements. C. MAJOR FACTORS AFFECTING THE PROJECT 26. Loan Effectiveness and Impllementation Schedule. The project was approved by the Bank Group's board on December 4, 1990. The Loan, Credit, Provincial Project, Financial Intermediaries Project and JGF Agreements were signed on January 17, 1991 and became effective on June 14, 1991. The JGF was scheduled to be closed on March 31, 1996 which was extended to December 31, 1996. The last disbursement was on December 22, 1996. The project completion and closing for the loan and credit were scheduled for March 31 and September 30, 1996, respectively. About 98 percent of the project funds were disbursed within the original closing date of September 30, 1996. The closing date was extended for one year to September 30, 1997, mainly in order to complete disbursements of the precommercial subprojects. The last disbursement of the loan was on December 31, 1997. 27. Factors Affecting the Project. The project's simultaneous focus on multiple target groups-different government entities, enterprises and financial institutions in various provinces and localities-made the project very complex and difficult to manage and supervise. It was particularly difficult since: (a) all project beneficiaries (except CIB) were new Bank Group borrowers; (b) the proper financial and enterprise sector framework was not in place; and (c) the assumption that financial institutions perform as bona fide lending institutions and enterprises, including the less-regulated TVEs, behave truly as independent entities was unrealistic. Consequently, there were three major factors affecting implementation of the project: (a) a delayed start in project implementation; (b) low subloan repayments; and (c) financial institutions unable to perform their duties as a bank due to lack of macroeconomic framework. 28. A Delayed Start in Project Implementation. The project had a very difficult and slow start for two major reasons: (a) a three-year elapse time from project appraisal in June 1988 to project effectiveness in Jume 1991. Economic and financial environment on which the project was built on had chamged substantially during that period. The initial batch of subproject was less satisfactory because of this substantial delay; and (b) the project's simultaneous focus on multiple target groups, different government entities, enterprises and financial institutions in various provinces and localities made the project very complex and difficult to manage and supervise. Except for CIB4, all other project beneficiaries were new Bank Group borrowers, thus procedures and requirements had to be explained and clarified at the start of the project. In addition, due to the diversity of project components, it took over a year to clarify with project entities on project activities and approaches to start. With the diligence and commitment of the Spark Program 4 The Bank had not dealt with the Jiangsu and Shanghai ABC branches before this project.  - 14- Offices, project activities started to pick up in 1992. But it had already resulted in delays of all components by about a year. Eventually, project activities came back on track and implementation progressed satisfactorily. The enterprise modernization component was the most affected. Commercial subprojects were fully committed only in 1996, while appraisal estimates expected them to be fully commitment by 1992. The precommercial component did not start until mid-1994. 29. Subloan Repayments. Of the 92 commercial subprojects, three have fully repaid. Despite relative operational success of a substantial number of subprojects, principal subloan repayment was only 18 percent of total principal due, and total interest payment was 31 percent of total cumulative interest accrued as of December 31, 1997. Major reasons for the low repayments were: (a) Unreasonably Short Repayment Terms. Subloans usually had three year repayment terms, including a one year grace period. All Spark subprojects were investment projects and some had to import equipment. Thus it was unrealistic to expect a subproject to be completed in one year and repayment to start in the second year. In addition, most subprojects started commercial operations only in 1995 and 1996, and 1997 was their first or second year of operations. Even those enterprises which could repay expressed the heavy burden of loan repayment in their first or second year of operations; (b) Inadequate Market Forecasts. As technologies introduced in the Spark subprojects were usually innovative in the locality or maybe even in the country, insufficient market analysis led to many not so successful subprojects, affecting sponsors' ability to repay; (c) Unrealistic Cost Estimates. Cost estimates were usually too low, resulting in cost overruns. This affected both the sponsors' ability to raise additional counterpart funding and their ability to repay; (d) Financial Institutions' Delayed Remedial Actions and Subloan Rescheduling. Loan supervision was often secondary to new lending. Late payments stayed on books for years without remedial actions. Rescheduling was often legally not allowed until repayments were seriously late for a few years; (e) Honoring Financial Obligations by Borrowing Enterprises. Loan repayment was often not the top priority of borrowing enterprises. New project investment, construction, and other social obligations took precedence over loan repayment; and (f) Local Government Interference. Some successful Spark subprojects were asked to absorb financially bankrupt enterprises by local governments. This has had serious consequences on Spark subloan repayments. The sponsors had to support these enterprises financially, divert their managerial and  - 15 - technical staff to help them, and ultimately failed to meet their own loan obligations. During Bank Group desk reviews of subproject appraisal reports of above free limit subloans, issues (a) to (c) and the issue of financial viability of prospective beneficiaries had always been the issues. It was these issues that often delayed Bank Group approval of the subprojects. However, after a sequence of queues and correspondence, the PFIs had always been able to revise the subproject appraisal reports to the Bank Group's satisfaction. In retrospect, the Bank Group should probably be stricter in its review of subproject appraisal reports. However, desk review does not constitute Bank Group appraisal of the subprojects and the Bank Group ultimately relies on the PFIs' appraisals. 30. Remedial Actions and Outlook of Subloan Repayments. The poor repayment performance of these subprojects reflects more generic issues in China's financial and enterprise systems that need to be addressed at the macro level. The recent financial and enterprise reforms would help to improve matters in the longer term. The Bank Group had requested the PFIs and local SPOs to prepare remedial actions for subprojects which were not repaying. These remedial actions were reflected in the rescheduling/ restructuring plans in the Borrower's contribution to the ICR. Some subprojects were rescheduled, some received technical assistance to change technology which was technically unsound, some strengthened their sales and marketing strategies, some obtained additional financial assistance from guarantors, local governments and other banks, and some diversified to other production activities. Even with this rescheduling, the Bank Group felt that repayment is imlikely to pick up in 1998. This low expectation is due to the general economic conditions in China and the East Asia region which are slowing down China's economy, and. also because most of the subprojects had only begun operations in 1995 and 1996. Blut in the medium term of 3 to 5 years, about 60 percent will probably succeed and could repay their loans. This is very much contingent on the success of financial and enterprise reforms currently underway in China. 31. Roles and Performances of Financial Institutions. The PFIs were the central actors in determining project sustainabtility, as they appraised and supervised 90 percent of project funds to final beneficiaries. 'However, the project provided only 0.2 percent of the budget for training the two ABC branches and CIB Jiangsu. All PFIs attended one training course on technology evaluation. This is considered totally inadequate in relations to the role they played in the project. Nonetheless, the performance of the PFIs could be improved if they had paid more attention to: (a) realistic assumptions for financial projections of enterprises; (b) attainable capacity utilization of the subprojects; (c) sufficient emphasis on enterprises' management and technical abilities; and (d) adequate market analysis and counterpart funding requirements of the subprojects. Additionally, PFIs should also pay more attention to portfolio management and the importance of realistic subloan repaynment terms. Their unsatisfactory performance was partly due to existing financial, enterpriise and legal systems, and policies and regulations in China. The current systems and policies did not offer financial institutions much freedom nor incentives in exercising their decisions in key operational areas, such as bad  - 16- debt provision, subloan restructuring and rescheduling, legal proceedings against subborrowers in case of deliberate default, legal obligations of guarantors, bankruptcy, or other actions that may be needed in case of default. SUBLOAN DISTRIBUTION AND SUBLOAN REPAYMENT STATUS AMONG FINANCIAL INSTITUTIONS AS OF DECEMBER 31, 1997 No. of Amounts of Percent Principal Interest payment Financial commercia subloans distribution repayment as % as % of institutions 1 (US$ of subloan of total principal cumulative subprojects million) amount due interest accrued ABC Jiangsu 30 39 39 15 17 ABC 16 19 19 2 0 Shanghai CIB Jiangsu 17 22 21 65 67 JPTIC 29 21 21 20 22 Total 92 101 100 18 31 32. ABC delegated all subproject appraisal and supervision responsibilities to its Jiangsu and Shanghai branches. The subprojects in Jiangsu ABC were in various cities in southern Jiangsu, and the subprojects in Shanghai ABC were on Chongming Island. Both branches were inadequate in their portfolio management and paid very little attention to subloan collection. Its loan collection of the Spark subprojects at the end of 1997 was close to zero. 33. During project implementation, the Bank Group was unable to obtain meaningful financial information, such as profitability, arrears, and collection data from the ABC branches. Thus, the Bank Group was not able to assess financial performances of the ABC branches. ABC did not provide the Bank Group with its 1997 financial statements and 1994-96 financial data were extracted from their published annual reports only. Thus the Bank Group was not able to confirm accuracy of these data. 34. In 1995, ABC was reclassified from a specialized bank to a commercial bank. ABC and the Bank Group are currently preparing an ABC Commercialization Project aiming to accelerate its transformation into a profit-oriented, modem commercial bank in both operations and management. In light of the Spark project experience, both China and the Bank Group should review carefully the Bank Group's assistance strategy to ABC. 35. CIB delegated all subproject appraisal and supervision to its Jiangsu branch, which performed adequately. At the end of 1997, CIB's current ratio remained very low  -17- at less than 0.5 percent-the lowest since 1990, its total debts were 25 percent of equity -the highest since 1990, and net income was only 0.4 percent of total assets-a slight improvement from previous years. All these financial indicators implied that CIB's financial position was extremely weak. 36. The Jilin Provincial Trust and Investment Corporation (JPTIC) was merged with the Jilin Province Economic Development Investment Corporation (JEDIC) in March 1991. JPTIC had assumed the rights and obligations of JEDIC and is the financial intermnediary channeling the World Bank Spark project funds to Jilin Spark enterprises. The new operating license was approved by the People's Bank of China in November 1991. Essentially there were no changes in its policy statement and operations. 37. Its current staff totals 115, of which about 25 percent are financial and economic staff. JPTIC has a Spark Project Department whose function includes appraisal, approval and supervision of all Spark subprojects. The Spark Department has been staffed sufficiently during the Spark project implementation period. JPTIC's activities include deposit taking (local and foreign currencies), trust and investment (in industrial enterprises and real estate), leasing, securities and brokerage services and underwriting, credit information services, and consulting services to enterprises. JPTIC remains very closely affiliated with the Jilin provincial government and is still supporting provincial development projects. It is the princilpal window in Jilin province dealing with foreign currency operations. 38. JPTIC has been marginally profitable in the last few years, yet its financial position weakened in 1997, due to the general economic conditions in China. At the end of December 1997, its current ratio was 1.3, declining from 1996, but better than its early years from 1990-95. Its total debt was 40 percent of total equity and had been at this level for many years. In 1997, its net income was only 1.2 percent of total assets. These financial ratios indicated that JPTIC vas financially weak and is not sustainable in the long run if among others its loan collection does not improve. D. PROJE CT SUSTAINABILITY 39. It is likely that project objectives are sustainable for all technical assistance components, but uncertain for thte enterprise modernization component. The achievements generated in all technical assistance components have already exceeded appraisal expectations and were demonstrated by the extensive use of various project products at the national and lower levels of the Spark prograrn and by rural enterprises, technical colleges, and the Spark training centers. The four training centers which the project helped financed evolved into Icey training centers in the Spark training system. The courses they offered are in high demand and are always completely filled. Project results had reached far beyond the thr-e project areas. The MOST and ISTIC have also determined to disseminate as much as possible project products to rural enterprises in the interior provinces.  - 18- 40. During project implementation, there was a gradual change in the Spark program from provisions of preferential policies and funding to rural enterprises to provisions of technology, information, and management services to rural enterprises. At the same time, the focus of the Spark program has moved from project-oriented financial and technical assistance to Spark intensive areas and Spark small areas development. This project has played a significant catalytic role in this change. 41. The Spark project operated in an environment that inhibits financial institutions from playing their role as banks. Their emphasis was and still is on lending operation, and project supervision is not a priority. They could rarely take legal actions against those enterprises that did not repay their loans, nor could they collect from their guarantors. They were not allowed to provide sufficient bad debt provisions, nor could they write off their aging outstanding loans. Enterprises were also accustomed to a culture that loan repayment is secondary to social obligations and new investment. At the same time, there has been mounting pressure recently towards mergers/acquisitions that may not be advantageous to successful rural enterprises. Thus, unless there are substantial and rapid financial and enterprise reforms in China, it would be difficult to expect subloan repayments to improve. Thus, as far as the enterprise modernization component is concerned, the outcome is uncertain. E. BANK GROUP PERFORMANCE 42. The project preparation and appraisal missions did a comprehensive review of technical assistance requirements to improve the national Spark program for the benefit of rural enterprises. The Bank Group contributed enormous resources in preparation for the technical assistance components with a well-defined program that properly addressed the most critical areas to be attended to in the first project. However, expectations of the Bank Group on financial institutions were very high for them to deliver the commercial credits subcomponent satisfactorily despite the rudimentary nature of financial and economic reforms in China in the late 1980s. Less than 0.2 percent of project funds were allocated to training of PFI staff in project appraisal and supervision. Nonetheless, even a much larger training program for the PFIs would not have improved subloan repayment status and subproject performances given the macroeconomic framework. 43. After project effectiveness, the Bank Group provided continual advice to project entities on Bank Group policies and procedures, reviews of procurement document, particularly on procurement of consultant services, and disbursement requirements. The Bank Group also provided advice on preparation of subproject appraisal reports and portfolio management. Bank Group staff followed closely the progress of all project components and provided support to minimize delays. Nonetheless, Bank Group supervision activities on the project could have been more effective had additional staff resources been allocated. As the borrowers had always welcomed the Bank Group's advice and comments and had the Bank Group allocated more budget for technical support for supervision activities, the Bank Group could have contributed more to the project's development.  - 19- F. BORROWER PERFORMANCE 44. The SPO in MOST and provincial/municipal STBs in the three project areas demonstrated full commitment to the project. The Spark Program Office, in charge of the entire project, did an excellent job in managing implementation of all the technical assistance components and coordinating work among all participating agencies, including ISTIC, local ISTIs, local STBs, financial institutions and training centers. Each province/municipality also formed a World Bank Spark Project office which managed and oversaw implementation of their respective allocations. From 1992 to 1997, the SPO organized seven Spark Project Implementation Workshops in the three project areas for all project entities to review work progress, discuss work programs, evaluate achievements, and determine immediate and longer term goals. These workshops greatly facilitated project implementation andl kept all participating agencies aware of project development. The SPO also published project progress reports to be circulated among all project entities on implementation of all components. These reports, sometimes published as frequently as every quarter, provided informative news on progress and performances of different agencies. Even though there were more than thirty project entities (excluding the Spark subprojects) in the three project areas, Bank Group missions noted that the project had been very well managed, and there was genuine cooperation among all project entities. Even though there were many managerial changes in MOST and SPO over the six implementation years, due to this extremely well-organized setup, project issues have generally been resolved quickly in the past six years. G. ASSESSMENT OF OUTCOME 45. Overall, project results are mixed. Technical assistance components were highly satisfactory for several reasons: (a) the Government completed all technical assistance components with great efficiency; (b) these activities provided essential support to the TVE sector which had exhibited extraordinary growth in the past decade; and (c) the national Spark program was strengthened as a result. Whereas, results of the enterprise modernization component was unsatisfactory. As most subborrowers did not repay their loans, the PFIs and ultimately the provincial/municipal finance bureaus would have to bear the financial cost. Despite some Bank Group subprojects demonstrated diffusion activities of technology and managemLent know-how to nonproject TVE beneficiaries nationwide, these benefits in terms of income and employment generated in other enterprises from the diffusion mechanism, may or may not recover the financial cost that provincial/municipal governments have to bear eventually. 46. With the support of local STBs, many of the Spark subprojects became pillar industries of their respective "Spark intensive areas." They provided not only direct and indirect employment opportunities, but led to growth of vertical and horizontal integration of industries in the localities. 47. The technical assistance components had extensive direct and indirect effects on the overall provision of technical suppcrt services by different levels of STBs to TVEs.  - 20 - The Chongming Training Center, despite its geographic disadvantage, has evolved into an intensive high quality training center in the Shanghai area and its utilization rate has reached 100 percent for more than two years. H. FUTURE OPERATION 48. The project has laid a good foundation for the national Spark program to grow and expand. In the short and medium term, the Spark program should concentrate on dissemination of various project activities, such as continual development of consultant services for rural enterprises, further improvement and promotion of instructional packages, utilization/promotion of the two technology evaluation manuals, extensive utilization of all training programs and skills learned in the Spark training program, cultivating productive use of the internet, improved utilization of the MIS, and many other activities that deserve further cultivation and dissemination, to the interior provinces in particular. In the longer term, the Spark program should reevaluate: (a) its role as government agency to foster the development of a rural enterprise sector which is the least regulated and the most robust in Chinese industries; and (b) its strategy in the development of rural enterprises in the interior provinces. 49. In regards to the subprojects operating profitably, the Government should avoid burdening them with asset acquisitions/mergers which may be financially unattractive to them. As to those subprojects with operational and financial difficulties, certain actions need to be taken by the local government, the local STBs, and the financial institutions to help the subborrowers turn around to be operationally competitive, financially viable, and environmentally sound over time. X. KEY LESSONS LEARNED 50. At the time the project was appraised, it was expected that it would be the first of a series of subsequent operations to support the rural nonstate industrial sector. However, as a result of the experiences gained from this project and other industrial projects and lines of credit operations in China, the Government and the Bank Group realized that the generic issues in the financial and enterprise sectors needed to be resolved before committing to further lending operations to enterprises through financial institutions. This means that subsequent operations have been significantly curtailed. 51. The following major lessons were learned under the project: (a) Successful execution of technical assistance requires full commitment and participation of all stakeholders throughout the process. It was achieved in this project. It shows that our client agencies are very receptive to technical assistance which has proved to be very useful to the Spark program. (b) National ownership is important. Spark is a national program and the Bank Group is only helping to strengthen it. The Ministry of Science and Technology takes prides in the Spark accomplishments.  - 21 - (c) Sound organization setup and dynamic leadership for implementing the project are essential to project success. This is especially essential for projects that have many implementing agencies in different project areas. (d) Provision of technical assistance needs to be driven by demand, rather than supply. Measures and incentives to promote demand rather than promote supply would made technical assistance more effective. (e) Long-term (over one year) consultant contract should have the flexibility of adjustment with the changing environment in the country. (f) Rapid dissemination of project products should be carried out by MOST immediately to avoid being outdated, and that could benefit the rural enterprise sector tremendously. At the same time, those instruments and products developed should be updated and revised continuously to keep their usefulness. (g) The Bank Group's project supervision missions should be staffed adequately with technical staff/consultants throughout-implementation. Sufficient budget should be allocated for this purpose. This was unfortunately not the case. (h) Financial institutions should strengthen their expertise in subproject evaluation including adequate selection criteria, market assessment, cost estimates, realistic implementation schedules, subborrowers' implementation and repayment capabilities., counterpart funding requirement and availability, and awareness of environnmental issues. However, even if these issues were fully addressed by financial institutions, enterprises undertaking projects in the context of limited finanicial and enterprise reforms would still face many constraints under the current national policy framework. The Bank Group should recognize these constraints in project preparation and try to seek solutions. If the Bank Group realizes that these issues could not be solved with one individual investment project in the absence of macroeconomic structural adjustments, then they should be clearly spelled out in project documents in order to avoid unrealistic expectations. (i) For lines of credit operations using financial intermediaries to be successful, some basic structural changes in the financial and enterprise sectors are needed: (i) stricter enforcement of laws, such as commercial banking law, company law, and bankruptcy law; (ii) enterprise behavior in terms of honoring loan repayment obligations and ensuring timely repayment; (iii) institutional behavior and performance standards of the financial institutions, such as subproject selection criteria, appraisal standards, portfolio managemenit, adequacy of bad debt provision, etc.; and  - 22 - (iv) local governments' attitudes to financial institutions and enterprises, i.e. local governments treat banks as conduits of funds to enterprises, and at the same time, they are more concerned with enterprises' social obligations and taxes rather than causing enterprises to repay bank loans.  - 23 - PART II: STATISTICAL TABLES TABLE 1: SUMMARY OF ASSESSMENTS A. Achievement of Objectives Substantial Partial Negligible Not Applicable x Macroeconomic policies X Sector policies X Financial objectives Institutional development X Physical objectives X Poverty reduction X Gender issues X Other social objectives X Environmental objectives X Public sector management X Private sector development X B. Project Sustainabiity Likely Unlikely Uncertain Technical Assistance Components X Enterprise Modenization Component X CG. Bank Group Performance Highly Satisfactory Satisfactory Deficient Identification X Preparation assistance X Appraisal X Supervision X D. Borrower Performance Highly Satisfactory Satisfactory Deficient Preparation X Implementation X Covenant compliance *X Operation (if applicable) E. Assessment of Outcome Highly Satisfactory Unsatisfactory Highly Satisfactory Unsatisfictory Technical assistance Components /a X Enterprise Modemization Component X /a It is necessary to assess the outcome separamely for the two distinct categories which were implemented by different agencies and had very different outcome.  - 24 - TABLE 2: RELATED BANK GROUP LOANS/CREDITS Loan/Credit Title Purpose Year of Status Approval Ln. 2226-CHAI To assist CIB in meeting the need for foreign exchange for industrial financing, 1983 Completed Cr. 1313-CHA improving investment efficiency, and organizing and improving CIB's capacity First Industrial for project design, selection, appraisal and technology transfer. Credit Project ($71.5 million) Ln 2434-CHAI To continue the institutional building efforts of the first project, including 1984 Completed Cr. 1491-CHA expanding branches, assist in appraisal of larger projects and meet CIB's resource Second Industrial requirement. Credit Project (S183.2 million) Ln 2658-CHAI To assist CIB in fnancing productive facilities and resources in China to 1986 Completed Cr. 1663-CHA contribute to the country's economic and social development; to provide CIB with Third Industrial funds needed to develop its operations and carry out its Charter and Statement of Credit Project Policy Strategy. ($103.6 million) Ln 2783-CHA/ To assist the government in promoting and implementing technology upgrading 1987 Completed Cr 1763-CHA and modernization in industry, to build up sound institutions and procedures for Fourth Industrial project appraisal and investment financing and to improve intermediation Credit Project practices; specifically, to continue institutional assistance to CIB and provide a ($300 million) general line of credit for industry. Ln. 2784-CHA To assist the Shanghai Municipality to achieve product rationalization and product 1987 Completed Shanghai Machine upgrading; modernization and rehabilitation of manufacturing facilities; expansion Tool Project ($100 of the design and engineering capabilities; and improvements in management milUion) systems for the machine tool sector within Shanghai. Ln. 2838-CHA To improve the production efficiency of existing medium-size plants through 1987 Completed Fertilizer plant renovations and upgrading technologies; expand phosphate production to Rationalization help achieve a more balanced nutrient input; and improve institutional capacity Project ($97.4 through the introduction of more modem systems and techniques in financil and inillion) operational management Ln. 2943-CHA To assist the govermment expand phafmaceutical production by using modem 1988 Completed Phannaceutical technology; improve the quality of the products and introduce Good Project Manufacturing Practices (GMP) in selected enterprise and training programs; and ($127 million) improve the quality, maintenance, marieting and management in the phamaceutical industry. Ln. 3075-CHA Continuation of activities under the Fourth Industrial Credit Projeet 1989 Completed Fiffh Indusri Credit Project (S300 miUion) Ln. 3022-CHA To assist the Tianjin Municipal Government in the implementation of the 1989 Completed Ihanjin Light development program and strategy that it had prepared for the three light industry Industry Project subsectors of the project (S154 million)  -25 - Loan/Credit Title Purpose Year of Status Approval lia. 3582-CHA To assist the local authorities in Southern Jiangsu to strengthen their 1993 to be Southern Jiangsu environmental planning and majiagement capabilities, and to cany out priority completed Environmental initiatives in the regional envirornmental strategy and action plans in an efficient 06130/2001 Protection Project and effective manner. ($250 million) Ln. 3788-CHA Support Shenyang Municipal Government's reform program through the joint 1994 to be Shenyang Industrial SMG/Bank monitoring of implementation of a Reform Action Program and the completed Project provision of technical assistance and training to underpin key aspects of SMG's 06/30/2001 ($175 million) reorganization. The project also r,upports two investment components: industrial restructuring and corporatization, and environmental protection. Ln. 3572-CHA The main project objectives are to: (a) expedite policy and enterprise reforms at 1994 to be Tianjin Industrial the regional level; (b) modemize and restructure the subsector based on economic completed Development and commercial viability; (c) strcngthen institutional infrastructure; (d) improve 06/30/2000 Project the organizational structure and intemal enterprise management systems; and (e) (S150 million) diversify and strengthen the financial intermediation process. Ln. 4045-CHA To assist Chongqing Municipality to: (a) achieve a significant reduction in 1996 to be Chongqing pollution and restructure productive facilities from its most polluting industries, completed Industrial Reform iron and steel; (b) establish a strategy and prepare a long-term plan to achieve a 12/31/2002 Project major reduction in pollution for the whole sector, and (c) initiate a pilot effort to ($170 million) assist industrial enterprises in oth[er industrial sectors to restructure their productive facilities, control polluition and transform themselves into modem corporations. TABLE 3: PROJECT TIMETABLE Steps in project cycle Date planned Date actual Identification September 21, 1987 September21, 1987 Preappraisal March 10, 1988 March 10, 1988 Appraisal June 17, 1988 June 17, 1988 Negotiations February 26, 1990 February 26, 1990 Board presentation December 4, 1990 December 4, 1990 Signing January 17, 1991 January 17, 1991 Effectiveness April 17, 1991 June 14, 1991 Project completion March31, 1996 September30, 1997 Loan closing September 30, 1996 September 30, 1997  - 26 - TABLE 4: LOAN/CREDIT/JGF DISBURSEMENT: CUMULATIVE ESTIMATE AND ACTUAL (US$ million) FY91 FY92 FY93 FY94 FY95 FY96 FY97 FY98 Loan/Credit Appraisal estimate 6.00 12.10 43.50 75.90 98.70 114.30 Actual 4.01 22.50 75.71 94.31 109.64 112.20 114.17/a Actual as % of appraisal estimate 33.10 51.70 99.80 95.60 95.90 n.a n.a Date of final disbursement 12/31/97 JGF Appraisal estimate 0.40 1.00 1.35 1.70 Actual 0.20 0.39 0.64 1.56 1.84 2.27/b Actual as % of appraisal estimate 50.00 39.00 47.40 91.80 n.a. n.a. Date of final disbursement 04/03/97 /a Final disbursement amount was slightly smaller than the appraisal amount due to fluctuations in the US dollar to Special Drawing Rights rate. /b Final disbursement amount was substantially larger than appraisal amount due to Japanese Yen appreciation during implementation years from 1991 to 1996. TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION Key implementation indicators in SAR Estimated date of Actual date of completion completion Enterprise Modernization Component Commercial credit subcomponent Subproject commitments Dec 1992 Jun 1996 Subproject disbursements Jun 1995 Dec 1997 Precommercial credit subcomponent Subproject commitments Dec 1992 Jun 1996 Subproject disbursements Jun 1994 Dec 1997 Spark Training Component Contract signature for technical assistance Dec 1990 Jan 1992 Arrival of contractor Jun 1991 Apr 1992 Implement technical assistance Jun 1994 Apr 1995 Training center construction Ground breaking Jun 1991 Jun 1991 Physical completion Jun 1994 June 1996 Equipment procurement commences Dec 1991 Jan 1992 Completion of equipment procurement Jun 1994 Jun 1995 Spark training study Team convenes Dec 1993 Jan 1994 Submission of draft report Jun 1994 Dec 1994  - 27 - Key implementation indicators in SAR Estimated date of Actual date of completion completion Technology Information Component Contract signature for technical assistance Dec 1990 Apr 1993 Arrival of advisor Jun 1991 May 1993 Assignment database/computer specialists Dec 1991 Jul 1993 Assignment of training team Dec 1991 Aug 1993 Equipment procurement commences Dec 1990 May 1993 Completion of equipment procurement Jul 1994 Database construction Jun 1993 Jun 1994 ISTIC central data base Jun*1992 Jun 1994 Local information consultancy services Staff training commences Jun 1992 Aug 1993 Staff training completed Jun 1993 Feb 1996 Institutional Development Component MIS subcomponent Contract signature Dec 1990 Apr 1992 Working group convenes Jun 1991 Jun 1992 Design phase commences Jun 1991 Jun 1992 Design phase completed Dec 1991 Dec 1992 System establishment commences Dec 1991 Dec 1992 System establishment completed Jun 1992 Jun 1993 System operations manual finalized Dec 1992 Jun 1994 Operational phase commences Dec 1992 Jun 1994 Technology evaluation subcomponent Contract signature Dec 1990 Jul 1993 Course design commences Jun 1991 Jul 1993 Course 1 and 2 delivered Dec 1991 Sep 1994 & Aug 1995 Manual finalized Jun 1992 Nov 1996 Financial intermediary training Training commences Dec 1990 Aug 1992 Training completed Jun 1994 Jul 1993  - 28 - TABLE 6: KEY INDICATORS FOR PROJECT OPERATIONS The key indicators for project operations were in terms of (a) full utilization of training facilities financed under the project, application of training programs and instructional packages developed in the Spark training component, (b) commercial utilization of data bases developed, frequent public and national broadcasting of the spark program and commercial dissemination of information and manuals to be used by Spark program offices throughout the country, (e) operations and financial performance of subprojects financed under the line of credit, and (f) diffusion of technological and management knowledge of the subprojects and all project components. TABLE 7: STUDIES INCLUDED IN PROJECT Study Purpose as defined at Status Impact of Study Appraisal/Redefined Study on Long-Term To assess the contributions made by Completed The study was completed as scheduled Training Program the project's training component and in December 1993. As training the priorities for firther short-tern and component was delayed for more than long-term improvement and a year, the study was revised in development of Spark training. December, 1994. Based on the training component, the Study identified six areas for further improvements which guided Spark. training development Between 1995-1997, SSTC arranged 1.323 classes for 76,160 trainees for management personnel in the Spark program nationwide. Some specific impact include: (a) to support the Spark Western Exploration Program by offering triining to SPO management personnel and entrepreneurs in the interior provinces, (b) to provide tining on evaluation methods and ISO 9000 quality control management for personnel in the Spark intensive areas to help their development, and (c) to provide computer network taining for SPO staff nationwide to help technical and information management of the Spark program.  - 29 - TABLE 8A: PROJECT COSTS IN YUAN MILLION Appraisal Estimate Actual Cost Component Local Foreign Total Local Foreign Total Enterprise Modernization Commercial Credit 682.0 252.5 934.5 1,197.9 696.2 1,894.2 Precommercial Credit 23.1 8.5 31.6 82.4 31.8 114.2 Subtotal 705.1 261.0 966.1 1,280.3 728.0 2,008.4 Spark Training Technology Information and Institutional Strengthening Civil Works 17.5 2.4 19.9 29.6 4.0 33.6 Equipment/Fumiture 18.9 18.4 37.3 17.6 42.4 60.0 Technical Assistance/Training 3.8 8.0 11.8 11.1 18.2 29.4 Program Support 5.2 - 5.2 14.7 0.0 14.7 Subtotal 45.3 28.3 74.2 73.0 64.6 137.7 Total Base Cost 750.4 57.1 1,040.3 1,353.3 792.6 2,145.8 Physical & Price Contingencies 4.7 .3.3 8.0 TOTAL PROJECT COST 755.1 60.4 1,0483 1,3533 792.6 2,145.8  - 30 - TABLE 8B: PROJECT COSTS IN US$ MILLION Appraisal Estimate Actual Cost Component Local Foreign Total Local Foreign Total Enterprise Modernization Commercial Credit 144.5 53.5 198.0 149.7 87.0 236.8 Precommercial Credit 4.9 1.8 6.7 10.3 4.0 14.3 Subtotal 149.4 55.3 204.7 160.0 91.0 251.0 Spark Training Technology Information and Institutional Strengthening Civil Works 3.7 0.5 4.2 3.7 0.5 4.2 Equipment/Furniture 4.0 3.9 7.9 2.2 5.3 7.5 Technical Assistance/Training 0.8 1.7 2.5 1.4 2.3 3.7 Program Support 1.1 - 1.1 1.8 0.0 1.8 Subtotal 9.6 6.0 15.7 9.1 8.1 17.2 Total Base Cost 159.0 61.4 220.4 169.2 99.1 268.3 Physical & Price Contingencies 1.0 0.7 1.7 0.0 0.0 0.0 TOTAL PROJECT COST 160.0 62.1 222.1 169.2 99.1 268.3 TABLE 8C: PROJECT FINANCING (US$ million) Appraisal Estimate Actual Financing Component Local Foreign Total Local Foreign Total IBRD/IDA 53.9 60.4 114.3 17.4 96.8 114.2 Central Govenunent 1.0 - 1.0 1.1 - 1.1 Provincial Government 8.3 - 8.3 8.3 - 8.4 Financial Intermediaries 35.4 - 35.4 82.6 - 82.7 Enterprises 61.4 - 61.4 59.6 - 59.6 JGF - 1.7 1.7 0.0 2.3 2.3 Total Project Financing 160.0 62.1 222.1 169.2 99.1 268.2  -31 - TABLE 9: ECONOMIC COSTS AND BENEFITS The PFIs reevaluated financial and economic viability of nine subprojects after subproject completion. Both EIRR1 and FIRR were required to be above 12 percent, according to the project document. Based on a sample of nine subprojects, the ex-post EIRR and FIRR were all lower than originally estimated due mainly to increasing input prices and decreasing output prices. Subprojects were selected for calculation of ex-post EIRR and FIRR based on the following criteria: (a) a minimum of two subprojects from each of the PFIs; (b) enterprises from different sectors and different cities of the three project areas; (c) an even mix of successful and less successful subprojects; (d) half in repayment stage; and (e) subprojects of different project sizes. The PFIs reevaluated financial and economic viability of nine subprojects after subproject comapletion. The nine subprojects accounted for about 10 percent by number of commercial subprojects and 13 percent of the loan/credit amount disbursed to the commercial subprojects. FIRR (%) EIRR (