Burundi - Coffee Sector Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 8347-BU STAFF APPRAISAL REPORT BURUNDI COFFEE SECTOR PROJECT HARCH 14, 1990 Agriculture Operations Division South-Central and Indian Ocean Department Africa Region This document has a restricted distribution and may he used by recipients only in the performance of be I  CURRENCY EQUIVALENTS Currency Unit - Burundi Franc (FBu) us $1.00 FBu 173 FBu 100 US $0.578 WEIGHTS AND MEASURES Metric British/US Equivalents I meter = 3.28 feet 1 cubic meter 35.3 cubic feet 1 hectare = 100 ares 2.47 acres 1 kilometer 0.64 miles 1 square kilometer X 0.386 square mile 1 liter = 0.26 U.S. gallon 0.22 imperial gallon 1 metric tonne 2,204 pounds GLOSSARY OF ABREVIATIONS AND ACRONYMS ASSP = Agricultural Sector Services rroject BCC - Burundi Coffee Company BRB - Banque do la Republique du Burundi CCCE Caisse Centrale de Cooperation Economique CIRAD a Centre de cooperation internationale pour la recherche agricole et le developpement (French coordinating authority for tropical agricultural research) COOPEC - Cooperative d'epargne et de credit (Savings and Credit Cooperative) FAC Fonds d'aide et de cooperation (French Assistance) ICA International Coffee Agreement ICO = International Coffee Organization ISABU = Institut des sciences agronomiques du Burundi MOAL = Ministry of Agriculture and Livestock OCIBU Office des cultures industrielles du Burundi RDC Regional Development Company SODECO Societe de deparchage et de conditionnement SOGESTAL Societe de gestion de stations de lavage GOVERNMENT OF BURUNDI FISCAL YEAR January 1 to December 31  FOR OMCIuL USE ONLY BURUNDI COFFEE SECTOR PROJECT STAFF APPRAISAL REPORT 1/ Table of Contents Credit and Project Summary . . ............... i-iii I. GENERAL ECONOMIC AND AGRICULTURAL SETTING . . 1 A. Background .. ...............................................1 B. Structural Issues . . ....................................... 2 C. Agricultural Sector Issues .... 3 D. IDA's Involvement in the Agricultural Sector ............... 3 II. THE COFFEE SUB-SECTOR ......................................... 4 A. Brief History .............................................. 4 B. Structure of the Industry .. 5 C. P-oduction ................................................. 5 D. Processing ................................................. 6 E. Marketing, Logistics and the World Market .................. 9 F. The Coffee Payment System ................................. 11 G. Research .................................................. 13 H. Institutional Development ................................. 13 I. Private Sector Participation .............................. 14 III. THE PROJECT .................................................. 15 A. Coffee Policy and Strategy . . .............................. 15 B. Background of the Project . . ............................... 16 C. Project Objectives and Description .... 16 D. Detailed Features . . .................................. 17 1- Institutional Reforms . . .............................. 17 2- Investments . . .................................. 27 E. Project Costs . . ........ 30 F. Financing . . ..............................................33 G. Procurement . . ........................................... 34 H. Disbursements . . ........................................... 36 I. Accounts and Audits . . ..................................... 38 J. Environmental Aspects . . ................................... 39 IV. PROJECT IMPLEMENTATION ....................................... 39 A. Organization and Management . . . 39 B. Implementation of Specific Components ..................... 39 C. Work Programs, Monitoring and Mid-Term Review ............. 42 l/This report is based on findings of an IDA mission which visited Burundi in November 1989, comprising Messrs Charles Bolduc, Jean-Francois Barres, Walter Schwermer, Allan Yates (IDA), and Jan Van Hilten (Consultant). The mission was joined by Mr. Y. Grouitch of CCCE. 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 witheut World Bank authorization.  V. FINANCIAL ANALYSIS .............................. 42 VI. JUSTIFICATION, BENEFITS AND RI S^S .................. ..oo*043 A. Justification ............ .t..o.9.......so........ *ooo.. 43 B. Benefits .......... .............. eo.. 43 C. Economic Analysis^ ... t ........ .***** **** *... *. *.44 D. Risks ................so......... .o.........o 9 ..... 45 VII. ASSURANCES AND RECOMMENDATIONS ......... *.......... ...... .0....^46 TABLES IN MAIN TEXT Table 1. Price Contingencies ....... s........... ....... . o. * ... . . 31 Table 2. Summary of Project Costs ............. t ..... .. ....... . 32 Table 3. Financing Plan.......... .... ... ..... too ........ ... 33 Table 4. Procurement Arrangements ..... ..... ... ....... ... . 36 Table 5. Disbursement Schedule .. ........................ . 37 ANNEXES AND MAPS L  i BURUNDI COFFEE SECTOR PROJECT Credit and Proiect Summary Borrowert Republic of Burundi Beneficiariess The coffee industry; Burundi coffee farmers; private investors; Banque de la Republique du Burundi (BRB); participating financial intermediaries (PFI); the Government (the Ministries of Agriculture and Livestock, of Commerce and of Public Works). kmounts SDR 21.3 million (US$ 28.0 million equivalent). Terms: Standard IDA, with 40 years maturity. On-lending terms: Credit funds to finance coffee washing stations to be built by private investors will be passed on by the Borrower to BRB at 1/2 of 1% below the rate of Treasury Certificates for 30 years, including a grace period of 10 years. BRB will onlend the funds to PFIs at a rate equal to that of Treasury Certificates. Interest rates charged by the PFIs to private investors will be market determined. Proiect Description: The fundamental objective of the project will be to improve the earning ability of Burundi coffee, the country's principal export crop. To achieve this, the project will help implement: (a) a package of policy and institutional reforms to improve the efficiency and productivity of the industry; and (b) investments in coffee washing stations, research and trairtng leading to higher value exports. An important principle behind the institutional reforms will be the opening of the industry to private investors and the ability to attract private capital. The major institutional measures to be implemented through the Project will include: (a) the creation of a regulatory authority (Office du Cafe); (b) the establishment of 5 autonomous entities (SOGESTALs) to take over the management of the Government owned coffee washing stations from the Regional Development Companies, and one autonomous entity (SODECO) to take over operations of the country's two hulling factories from OCIBU; these entities, including individual washing stations, will be open to private ownership; (c) the establishment of an Assets Management Facility to concentrate the ownership of the Government-owned physical assets in the industry; (d) the replacement  ii of the current administered pricing system by market-based prices for intermediaries, quality incentives for producers and, for the Government, the establishment of a direct export tax to replace the current levy on the marging and (e) the establishment of an auction facility for private exporters to compete with the state-owned BCC. The major investments to be financed by the project will include: (a) the construction of up to 100 additional coffee washing stations by both the public and private sectors, the mix depending on pz.rvate investors' response to improved incentives; (b) support to coffee research with emphasis on solving market related problems (e.g. off-flavors, technological issues); and (c) an industry-wide training program. Benefits and R'isks: Sector performance will benefit from increaees in efficiency, competition and transpatence at all levels resulting from institutional restructuring, market based prices and greater private investor participation Quantifiable benefits will be generated from: (a) quantifiable improvemente in processing coefficients resulting from increased efficiencies; (b) the incremental value of coffee sales resulting from the shift from washed to fully washed coffee and from avoiding a likely discount of the price of semi-washed coffee associated with inadequate quality control. Other *nefits will result from the lowering of producticn costs at farm level resulting from the suppression of hand pulping (an activity often carried out by women, thus freeing their time), and the employment generated by the additional washing stations. The possibility of resumption of the ICA quotas has not been counted as a benefit. The economic rate of return (EER) of the whole project is 22 parcent, with a net present value of about US$19 million. The risks will include: (a) delays by Government in ensuring the effective implementation of the proposed reforms; (b) a lack of interest from private investors to enter the industry; (c) slowness on the part of private exporters in participating in the coffee auctions.  ift Estimated Proiect Costa Local Foreign Total ---- US$ million ----- Institutional Restructuring 4.6 2.8 7.4 Washing Stations 14.3 10.7 25.0 Research 2.0 2.5 4.5 Training 0.2 0.6 0.8 Total Basel'ne Costs 21.1 16.6 37.7 Physical Contingencies 1.7 1.4 3.1 Price Contingencies 3.2 2.3 5.5 Total Project Costs 11 26.0 20.3 46.3 Financing Plan Local Foreign Total - _____ -US$ million ----- Government 9.0 0.3 9.3 IDA 12.9 15.1 28.0 r CCCE 2.9 3.2 6.1 FAC 0.3 1.2 1.5 CIRAD 0.1 0.3 0.4 p Belgian Assistance 0.8 0.2 1.0 N Estimated IDA Disbursements IDA Fiscal Year 1991 1992 1993 1994 1995 1996 1997 Annual 2.1 4.6 6.3 6.0 5.1 2.5 1.4 Cumulative 2.1 6.7 13.0 19.0 24.1 26.6 28.0 Economic Rate of Return: 22Z Maps: IBRD 22033, 22247 ll/Includes direct taxes of about US$ 5.2 million.  1 BURUNDI COFFEE SECTOR PROJECT 1/ Staff Appraisal Report. I. General Economic and Agricultural Setting. A. Background. 1.01 Burundi is a small, landlocked country in East-Central Africa. It:z GNP per capita, estimated at about US$240 (1987) ranks among the lowest on the continent. With a population of about 5 million, Burundi has the second highest population density in Africa estimated at 170 persons per square kilometer. Population growth is high at 3% per annum projected over the 1985-2000 period. Despite high demographic pressure on an already constrained land area, migration to the urban centers has been limited; urban population accounts for only 5% of total population. 1.02 Agriculture is Burundi's most important economic activity accounting for aboue 50% of GDP. In addition, agriculture generates about 88% of the country's export earnings. More than 90% of the population live in the rural areas and depend directly on igriculture for their livelihood. The countrv's climate is variable but moderate and the soils, even if among the poorest in Africa in terms of acidity and nutrient content, permit year-round cropping and support one of the most intensive and productive agricultural systems on the African continent. Burundi is one of the few African countries self-sufficient in food (at least for the most popular staples). Most if not all the country's food production comes from small private farm holdings averaging about 0.7 ha. in size. 1.03 High altitude arabica coffee is Burundi's major export crop. Average annual exportable production over the past three years has hovered around 33,000 tonnes. Some robusta coffee is also grown at lower altitudes; production averages around 1,400 tonnes p.a. Total area under coffee by 1991 should amount to about 115,000 ha representing about 4% of the total area of the country. Arabica coffee is grown on some 500,000 smallholdings supporting about 50% of the farm population of the country. Export grade robusta is grown mostly on industrial plantations. A typical arabica smallholding may contain from 80 to 100 coffee trees (0.05 to 0.07 ha). Yields are not substantial and vary widely from 300 to 1000 kg of 'clean", or export grade, coffee per hectare equivalent, depending on the potential of the area. Few chemical inputs are used with the notable exception of insecticides. For small farmers, coffee returns represent from 40 to 60% of their monetary revenues, most other revenues coming from the sale of banana 'beer' and from off-farm labour. For the Government, coffee generates some 75 to 85% of the country's foreign exchange and 1/ In French: Projet Filiere Cafe.  2 represents a widely variable but sometimes significant source of budgetary revenue ranging from zero (1983) to about 42% of total revenue (1986). B. Structural Issues. 1.04 Beset with serious economic and finar ial difficulties In the early 1980's, the Government undertook in 198v with the assistdnce of the IMF and IDA, a major adjustment program the objectives of whic'i were to restore financial equilibria in the short term and to initiate medium- and long-term structural changes in the economy with more reliance on market forces and a greater outward orientation. This on-going progtsm, started under SAC I (Credit 1705-BU) ard strengthened under SAC II (1982) (Credit 1919-BU), addresses a broad range of structural issues through measures aiming at: (a) improving macro-economic management; (b) stimulating exports; (c) developing the private sector; (d) reforming public enterprises; (e) improving public expenditure management; (f) increasing the efficiency of the financial sector; (g) liberalizing the labor market; and (h) alleviating poverty. On the whole, the Government's performance at implementing Its adjustment program has been satisfactory despite difficulties in such areas as public enterprise management and balancing public expenditures with revenues. Much of the recent dialogue between IDA/IMF and the Government has centered on these problems. Of particular importance has been the impact of the recent fall in coffee prices which have accentuated the Governmant's dependence on these volatile revenues as generated under the present payment system. 1.05 The financial covenants of SAC I and SAC II combined with those of tne Apex Project (Credit 1889-BU) have resulted in major on-going reforms in the financial sector, the most important of which include: (a) a complete liberalization of interest rates, achieved in August 1989; (b) the establishment of monthly auctions of Treasury auctions, since early 1988; and (c) a flexible exchange rate policy leading to several devaluations of the Burundi franc, the latest of which was 15% in November 1989. To strengthen its adjustment strategy, the Government has recently undertaken a number of sectoral operations in agriculture 2/ and transport 3/ which address the specific issues of these sectors. A third operation will be the Coffer qector Project which combines important policy, institutional and investment measures in the country's most important agro-industrial sector. One of the innovations of this operation will be the promotion of private investor participation and the reduction of Government's direct involvement in a sector which up to now has been strictly in the public domain. The measures of this operation will also aim at liberalizing internal prices and improving the reliability of Government's revenues from the sector, enhancing the market value of the country's most important export commodity, and creating employment opportunities as well as strengthening local management and construction capacity. Furthermore, the proposed project will also serve as a useful testing ground in such areas 2/ Agricultural Services Sector Project (ASSP) (Credit 2024-BU) 3/ Transport Sector Project, negotiated in January 1990.  3 as public enterprise reform and privatization which could be replicated in the form of broad policy measures as part of a prospecLive SAC III operation. C. Agricultural Sector Iss3es. 1.06 The preparation of the ASSP provided the opportunity to focus on sector-spezific issues the most important of which include: (a) the low level of market development especially for fooderops; (b) the ineffectiveness of agricultural services, particular3y extension; (c) the lack of direct relevance of agricultural research to farmers' needs; (d) the lack of institutional agricultural credit; (e) the lack of sector-wide programming; (f) the fragmentatior. of budgetary channels; and (g) the inefficient utilization and monitoring of financial resources. To address these issues the Government has defined a strategy which conforms with its broader adjustment objectives and includes sector specific measures to: (a) improve productivity levels through increased utilization of inputs, particularly fertilizers; (b) adopt pricing policies in line with production and export objectives; (c) strengthen the coffee sector while promoting the development of alternative agricultural exports; (d) redefine the role of public institutions and restructure the Regional Development Companies (RDCs); (e) strengthen its programming, budgeting and monitoring capability; (f) restructure its research facilities in closer confol-ity with the needs of the agricultural sector; and (g) address the shortcomings of the cooperative movement and expand the coverage of the Savings and Credit Cooperatives (COOPEC). An Important package of such measures has been included in the recently approved ASSP particularly as regards the improvement of fooderop productivity, the reform of central and regional institutions, the strenghtening of financial management, and the promotion of cooperatives. Measures to address the issues which concern specifically the coffee sector are the object of this Coffee Sector Project. D. IDA's involvement in the Agricultural Sector. 1,07 The Bank Group is one of the major sources of development assistance to Burundi. To date, IDA has financed nine agricultural projects, of which six have been completed. 1.08 Implementation of the Fisheries Development Project (Cr. 626-BU) was hampered by weak project management coupled with complex project design; its results were disappointing. The First and Second Coffee Improvement Projects (otherwise known as Ngozi I and II, Cr. 147-BU and Cr. 593-BU, respectively) succeeded in promoting the production of arabica coffee but otherwise had limited results in the area of food crops. Two successor projects, the Ngozi III (Cr. 1192-BU) and Kirimiro (Cr. 1165-BU) were just recently completed. While physical implementation of these projects was generally satisfactory, the Regional Development Companies (RDCs), responsible for their execution, were faced with a number of financial and managerial problems (see para. 2.08) the lessons of which served as the basis for the teorientation of IDA's strategy in the sector as reflected in the design of the recently approved Agricultural Services Sector Project (Cr. 2024-BU). The First Forestry Project (Cr. 918-BU) was  4 completed in 1986; result3 were generally satisfactory. Implementation of the Second Forestry Project (Cr. 1620-BU) started in 1988; the project aims at further developing basic forsstry services, promoting tree planting by farmers and developing plantations to supply fuelwood, poles and timber. The Muyinga Agricultural Development Project (Cr. 1857-BU), which started implementation in 1988, aims at strengthening agricultural iervices and cooneratives in Muyinga Province. 1.09 The coffee operations undertaken under the IDA financed Ngozi (3 phases) and Kirimizo projects pioneered the production of fully washed arabica coffee in Burundi through the introduction and expanded use of wet processing technology (washing stations). Other donors, such as the Kuwait Pund, which co-financed the Ngozi III project, and the French Caisse Centrale (CCCE 4/) have also contributed to this development. The Caisse Centrale has also financed washing stations in the Makamba-Mumirwa areas of the country as well as twio large hulling factories in Bujumbura and Gitega. Also, the Government, from its own resources, has financed a number of washing stations its the Kirundo area. The success of this experience is reflected by the farmars' positive response in delivering their production to the washing stations, and by the improving performance and image of Burundi coffee on world markets. II. THE COFFEE SECTOR. Current Situation. Issues ard Developments. A. Brief History. 2.01 Arabica coffee was introduced in Burundi at the beginning of the century. Recognizing its potential as a smallholder crop, colonial authorities encouraged its development and, at one time in the 1930'8, even made it compulsory for farmers to plant coffee. In 1945, OCIRU 5/ was established tc manage the development of production and the processing of the crop; it was replaced, at independence, by the present OCIBU 6/. Whereas arabica coffee has always been a amallholder crop, robusta coffee, which was introduced in 1934 by Belgian interests, has been grown mostly on irrigated estates totalling about 700 ha. Export marketing was carried out largely by private exporters until 1978 when it became a Government monopoly under a nationalized BCC 7/. The stabilization fund was established in 1S48. 4/ Caisse Centrale de Cooperation Economique. 5/ Office de3 Cultures Indigenes du Ruanda-Urundi 6/ Office des Cultures Industrielles du Burundi. 7/ Burundi Coffee Company.  5 B. Structure of the Industry. 2.02 The arabica coffee industry is structured along the lines of its product flow and consists of two related but distinct product paths, one for "semi-washed" and one for "fully washed" coffee 8/. Each path comprises a number of successive steps. For "semi washed" coffee, these includes (a) the farmers who grow the coffee, do-pulp the harvested coffee cherries (by hand or by manual de-pulpers), and sell the resulting parchment coffee to market traders; (b) the traders who collact, deliver and resell the parchment coffee to OCIBU hulling factories; (c) the OCIBU factories which hull the parchment coffee into "clean' exportable coffee and sell the latter to BCC; and (d) BCC which has a do facto monopoly in selling the coffee tc international buyers and delivering the product to the port of Dar es Salaam for ultimate shipment. The "fully washed" path is similar in most respects to that for 'semi washed" but differs in that: (a) farmers deliver and sell the coffee cherries immediately ar r harvesting to coffee wash!ng stations operated by the RDCs and by OCIBU; and (b) the washing stations convert the cherries into parchment and deliver and sell the latter to the OCIBU hulling factories. Thi6 latter process eliminates the need for the farmer to de-pulp the cherry and dry the parchment coffee and the intervention of the trader middleman. The two products generally have different end uses: "semi washed" coffees are mostly used as "bulk" while "fully washed" are generally used as "blenders" to enhance the body an-4 flavor of the bulk product. As a result, 'fully wAshed" arabica :sually sells at a premium over 'semi washed" arabica or other types of coffees 91. C. Production. 2.03 Reflecting the importance it attaches to coffee, the Government has, over the years, undertaken measures to increase production through the expansion, intensification and replanting of coffee areas, strengthening and diversification. of research, improvement of cult.vatica -qthods and, where possible, increased use of fertilisers and insecticides. As a result, coffee production has shown a positive trend increasing from an annual average of about 20,000 tonnes in 1976-80 to abort 31,000 tonnes in 1984-88. During the period 1986 to 1988, the Government subsidized production and distribution of coffee plants to increase coffee tree 81 "Semi washed" and "fully washed" are trade expressions used almost exclusively to describe coffees originating from Burundi, Rwanda and Kivu (Zaire). They are seldom used for other origins. For instance, almost all Kenya coffees are technically "fully washed", yet the trade refers to them simply as "washed". 9/ Based on price data for the higher quality classes of Kenya coffees (Standards 1 to 5, all "fully washed"), during the period 1974-88, the premium obtained for these coffees (FOB Mombasa) averaged about 19% over the standard indices (GPW - Guatemalan Priie Wash - or ICO Other Milds Indicator) on an equivalent basis. By coivarison, the prices for the "semi-washed" coffees of Burundi and Rwanda, over the same period and also FOB Mombasa or Dar-ss-Salaam, hovered by +/- 1% about these indices.  6 population from 102 million in 1986 to about 165 miiI'.on, an effort which should result by 1992 in a production ranging from 45,000 to 50,000 tonnes. Since 1988, however, Government has deliberately rest:ained its expansion policy focusing mostly on intensification and the replacement of older trees. In a recent letter to IDA, the Minister of Agriculture confirmed this policy stating that area expansion has been halted in marginal areas and is just about comp.eted in high and medium potential areas. Future interventions will focus mostly in improving the productivity of existing plantations in high potential areas. 2.04 The use of inputs has been limited mostly to insecticides to combat antestia, a sucking pest. Because these insecticides can only be effective if sprayed systematically over the whole of the coffee area, their annual procurement and distribution was arranged by OCIBU and their application carried out by spraying teams also organized by OCIBU. These tasks are being taken over by the Ministry of Agriculture. The cost of insecticides is and will continue to be covered by a levy on coffee export revenues 10/. Few if any fertilizers have been applied in the past but farmers in the Kirimiro area have recently shown a growing interast in their use. There are no subsidies on coffee fertilizers. 2.05 In terms of world coffee production and exports, Burundi's proportion is small representing in 1988 about 0.652 of total world production and about 0.722 of total exports. In the African context, Burundi's importance is equally small representing about 3X of production and exports. D. Processing. 2.06 Washine Stations. In the processing of 'fully-washed' coffee, de-pulping, washing and drying are carried out at the washing station. In the production of high quality coffee, processing by washing stations offers a number of advantages, notably: (a) that of requiring from farmers that they deliver only freshly harvested coffee cherries which are of adequate ripeness and generally free of disease; (b) since quality can be assessed at this stage, a price incentive can be introduced to motivate farmers towards quality; and (c) strict adherence to the washing and drying process. These types of quality controls and incentives are not possible in the case of "semi-washed" coffee. The disadvantage has been a gradual reduction of the role which the middleman/trader (i.e. the private sector) played in the handling of "semi-washed" coffee. This, however, does not mean that private traders have disappeared altogether since most are market merchants trading in other commodities. In fac'-- they compete with washing stations by offering services (such as credit using "semi- 10/ This levy and others (such as those proposed to fund coffee research, access road maintenance and the costs of the Office du Cafe) are industry levies which would be collected at time of export to finance sector specific activities. The amount of the levies would be determined on the basis of need and would be set by industry representatives on the Board of the Office du Cafe. These levies should not be confused with the export tax which would be destined to the Government's budget. u  7 washed' coffee as collateral) which washing stations have not been in a position to offer as yet. Ill 2.07 Recognizing the potential for higher returns from the inherently high quality of the country's coffee, the Government has, over the past 17 years, with the assistance of the Bank and other lenders, invested in the construction of 88 washing stations. For the coffee year 1988-89, some 9,900 tonnes of fully washed coffee were produced, representing 32% of total production, almost tripling the amount of similar coffee produced in 1985-86. The Government's objective is to produce enough fully washed coffee to meet its foreseeable quotas in the event of the re-establishment of quotas under the International Coffee Agreement (para 2.19). To this end, and assuming a reasonable quota increase over the next five years, the country will need a total of 196 stations, leaving 108 stations to be constructed of which 8 are scheduled for constzuction in early 1990. The balance of 100 stations are those which uill be constructed in the course of the Coffee Sector Project. 2.08 All washing stations currently operating in Burundi are owned by the Gcverment and operated by the RDCs (Buyenzi and Kirimiro) and OCIBU. The financial and managerial problems of the RDCs were the object of a specific study in 1985 and were extensively reviewed in the course of the preparation of the ASSP. Conceived initially as agro-industrial activities, the objectives of this basic concept were never fulfilled due to: (a) an inadequate remuneration for the RDCs' fully washed coffee under the echelle mobil:. a situation which required the RDCs to depend on subsidies to cover the losses from these activities; and (b) the scope of their service-type activities (extension, research and input production) the cost of which always exceeded the commercial earning power of the RDCs. The RDCs were viewed by the Governmerit as regional branches of the Administration: their style of management and the conditions of service of their staff have always been those of the Government. Coupled to this was the general inability of the RDCs to establish adequate accounting and budgeting practices commensurate with their activities. 2.09 The Government has recognized these issues and agreed, under the ASSP, to restructure the RDCs by returning the service-type activities to the Ministry of Agriculture and by spinning off the operations of the washing stations to autonomous entities created for this purpose. As regards the washing stations, a study prepared during the preparation of the Coffee Sector Project recommended that 5 autonomous companies, or SOGESTALs 12/, be created to manage the stations currently operated by the RDCs and OCIBU, and those to be built for the Government in the future. Groups of washing stations, varying in number between 23 and 45 per group, 111 Moreover, with the opening of washing station ownership to the private sector, these traders are potential investors in washing stations. 12/ SOciete de GEStion de STAtions de Lavage.  8 will be assigned to these SOGESTALs 131. The Government has accepted the study's recommendations. The establishment of the SOGESTALs will constitute an important element of the proposed Coffee Sector Project's institutional restructuring component. 2.10 Hulling Factories. In addition to its investments in washing stations, the Government has undertaken, with the assistance of the CCCE, sizeable investments in the construction of two hulling factories, one in Bujumbura, the other in Gitega. These large factories hull the "parchment coffee' delivered to them by washing stations and by private traders, sort the resulting "clean' coffee into export grades and bag it for shipment. Under an on-going project, also financed by CCCE, considerable improvements are being carried out at the Gitega factory to increase the sorting and wharehousing capacity of this factory. When these improvements are completed (1990), most of the "fully washed' parchment coffee delivered by washing stations will be processed at the Gitega factory while the "semi- washed" parchment will be delivered to the Bujumbura factory. These arrangements will ensure that Burundi's fully washed coffee is processed under conditions which protect its inherent quality while providing the technical facilities to tailor the quality grades more closely to the needs of the market. 2.11 The factories' combined nominal annual capacity amounts to about 40,000 tons of clean coffee. However, technical staff have indicated that this capacity could readily be increased to about 50,000 tons through internal task and scheduling arrangements. This capacity is sufficient to handle projected crop increases to about 1992-93. Beyond that time, additional hulling capacity would likely be required if the production increase estimates are realized. By virtue of operating the only hulling facilitie3 in the country, these Government-owned factories have a de facto monopoly in this area. However, the Government has agreed to private investors building and owning future additional hulling capacity. Moreover, the possibility that the existing factories could be sold to private investors is also under active consideration. 2.12 The two factories are currently operated by OCIBU. In the course of preparution of the Coffee Sector Project the Government agreed, as part of the conversion of OCIBU into an Office du Cafe (para 2.28), to transfer the operating management of these factories to an autonomous company to be named SODECO 14/. As for the earlier mentioned SOGESTALs, the establishment of SODECO will constitute an important element of the institutional restructuring component of the oroject. 13/ Indicative numbers per SOGESTAL (the numbers would vary depending on the number of washing stations owned and operated by private entrepreneurs): SOGESTAL Ngozi: 45 stations; SOGESTAL Kirimiro: 42 stations; SOGESTAL Muyinga: 23 stations; SOGESTAL Kirimiro: 45 stations; and SOGESTAL Mumirwa: 41 stations. 14/ SOciete de DEparchage et de COnditionnement.  9 E. Marketing. Logistics and the World Market. 2.13 Marketing. The solling of Burundi coffee for export is currently the de facto monopoly of the Burundi Coffee Company (BCC). Major concerns have been expressed in the past about BCC's lack of proficiency at me:keting the country's coffee. Issues had been raised concerning the inexperience and lack of market knowledge on the part of BCC staff, the lack of agressive marketing, inadequate communications, lack of proper facilities to evaluate coffee, etc. Underlying these issues is the general laxity of the institution a good part of which results from the absence of competition for the agency and the lack of motivation of its key personnel. It is run very much as a branch of Government operating at hours when an important part of the wor'd arabica market is closed. Staff remuneration is also based on Government scales, although at a slightly higher rate. 2.14 During the preparation of this proposed project, IDA funded technical assistance was provided to BCC as a result of which the agency has, periodically, succeeded in obtaining prices for its coffee which are more in line with other competing world coffees of similar grades and quality levels. To illustrate, whereas in the past BCC has only managed to sell 'fully washed' coffee at relatively low premiums over its "semi- washed, in 1988-89 this premium averaged close to 30Z over the better qualities of "semi-washed" coffee. Also with IDA support, satellite-linked communication equipment was recently installed and office organization was improved. 2.15 BCC coffee sales are carried out by a sales committee composed of the Managing Director and the heads of the Logistics and Administration Departments. The Government has agreed that the all-important sales function should henceforth be handled by a specialized Marketing Department which will be created for this purpose. The by-laws of the company will be ammended to provide the new Department with the necessary autonomy to formulate marketing strategy, execute sales and promote the product. Moreover, staff conditions will be improved in order to attract qualified candidates from the private sector. A Marketing Director will be recruited following the establishment of these conditions of service. 2.16 With the aim of increasing the efficiency of coffee export operations and providing competition to BCC, the Government agreed, in the course of project preparation, to the establishment of an auction system where coffee will be sold to private exporters in parallel with and as an alternative to the existing BCC channel. This private auction facility will be introduced gradually and, for the first two years will be allocated increasing proportions of the export crop 15/. The advantages of the proposed approach include: (a) the avoidance of a situation where marketing operations could be seriously disrupted if the marketing responsibility were suddenly and totally transfered from the public to the private sector; (b) the transparence of transactions to set prices on the basis of which foreign exchange and export tax earnings will be calculated; (c) the maintenance of BCC as a safeguard against the recurrence of past abuses 15/ For 1991: a minimum of 12,000 tonnes, or 30% if the crop exceeds 40,000 tonnes. For 1991-92: 50X of the crop.  10 while motivating the company to improve its results since these will compared with those of the parallel facility; (d) greater market penetration and greater competition for Burundi coffee; (e) the gradual build-up of competence and experience of exporters (nationals or foreigners) in doing business in Burundi; (f) the possibility that operators within the industry (such as SOGESTALs and SODECO) will become more closely connected to the market by selling directly through the auction; and (g) the need for private exporters to ship their coffee to Indian Ocean ports will likely stimulate a greater utilization of private transport facilities and more alternative (i.e. less expensive) corridors than is presently the case. 2.17 The above measures, as regards both BCC and the private auction facility will be implemented by the project as part of the institutional strengthening compomert. 2.18 Logistics. In addition to its selling functions, BCC's responsibilities include the delivery of coffee to Indian Ocean ports. Many issues remain surrounding the shipment of coffee through Tanzania but solutions are currently being sought for most of them. Where the Tanzania Railway (TRC) is concerned, a multi donor (including IDA) project is being prepared to improve the rail line, to replace or renew the equipment and to improve the company's management. Complementary to this project, the EEC is appraising a project to finance a number of locomotives and wagons to be dedicated to the operations of hlock trains for transit traffic, including Burundi traffic. Another project to improve porthandling facilities at Kigoma and Dar es Salaam is under preparation. Finally, an EEC funded project would rehabilitate some of the roads in north-western Tanzania, a situation which would significantly facilitate the shipment of Burundi coffee by road to Dar es Salaam (via Muyinga). This route was recently tried by BCC and found competitive with the railway (despite the current poor condition of the road). In fact, because of much shorter transport time and a greater reliability of timely arrival at Dar, BCC intends to ship as much as possible of its fully washed coffee over this road over the coming year. The Mombasa route has yet to be tried but all current estimations indicate that, because of the number of countries to cross, the distance and the volume to be carried, costs would be considerably higher than the Tanzania alternatives. Much of BCC's efforts at resolving its transport problems are the result of CCCE funded technical assistance. This assistance will be maintained under the Coffee Sector Project. 2.19 The World Market. On July 3, 1989, following the breakdown of discussions on the International Coffee Agreement (ICA), the then prevailing quotas were suspended resulting in a fall in world prices of about 30-35% from earlier levels. The major issues surrounding the suspension were the insistence of consumer countries (and some producing countries) on: (a) easier access to the types of coffees, especially higher quality coffees, which consumers want; and (b) ensuring price equity between the quota and non-quota markets. In September 1989, ICA members agreed to renew the coffee agreement without the all-important economic provisions (quotas and price mechanism) which will remain suspended for the next two years. The resulting outlook is that, unless quotas are re- established, world coffee prices will remain low at least until current surplus stocks are liquidated and marginal production is discouraged.  11 2.20 Despite these circumstances, Burundi is in a good position to benefit from the present world coffee situation. First, it produces the type of highland arabica coffee for which there is a growing demand. Second, as reflected by past and proposed investments, the Government's policy is to increase the proportion of fully washed coffee the country exports in order to fetch the relatively higher prices offered. Third, as reflected in this proposed project, there is a clear determination to increase the competitivity of Burundi coffee by improving the efficiency of the industry at both production and marketing levels. These arguments are corroborated by the conclusions of a World Bank Staff Working Paper 16/ which places Burundi among the countries which have the capacity to export more under a non-quota scenario than under a quota scenario, and thus may be able to compensate for some of the lost revenue resulting from the current low prices. 2.21 Burundi's position, in the event of a resumption the ICA quotas, would also be favorable since, in addition to benefiting from the likely price increases, and assuming that consuming countries succeed in their quest for higher quality coffees, there is a fair probability that Burundi would be included among those countries to wthich a significant quota increase would be allocated. In the past, Burundi's annual quotas have generally followed the country's increasing average production trend, and the country has thus been able to dispose of almost all its exportable production on the quota market. Whether this situation would prevail under a new quota agreement, or whether Burundi would be classed as a 'basic quota" producer, cannot be foreseen at this time. F. The Coffee Payment System. 2.22 Currently, the payment which each participant from production to marketing receives for its contribution, is prescribed by a payment schedule which is refered to as the "6chelle mobile". This schedule is set each year reflecting projections on production, export sales and the estimated budgetary requirements of the intermediaries. The coimmission which establishes the schedule is chaired by OCIBU which is also responsible for administering its application. The recommendations of the commission are subject to Government approval. 2.23 The system operates a 'stabilization fund" to which surplus coffee revenues 17/ are deposited in years when export proceeds exceed the remuneration of the industry, as prescribed by the echelle mobile, or resources are drawn from when sales proceeds cannot meet echelle mobile payments. In contrast to similar systems in Africa, the Burundi stabilization fund has, on the whole, worked reasonably well. The major reason for this is that the Government will only share in coffee receipts 16/ Akiyama, T. and P. N. Varangis (1989), Impact of the International Coffee Agreement's Export Quota System on the World's Coffee Market, PPR Working Papers WPS 148 171 Up to a maximum of FBu 1 billion.  12 at the margin, i.e. after all participants have been paid and the stabilization fund has been replenished up to its statutory 1 billion FBu. Thus, under the present arrangement, the Government assumes most of the risks and consequences of world coffee price fluctuations. 2.24 In common with other administered pricii.g systems, the Burundi system has important disadvantages: (a) the absence of motivation and incentives to improve efficiency and quality as 'costs" and "prices" are pre-set and administratively controlled; (b) the variability and unreliability of Government coffee revenue; and (c) the favored position of OCIBU, as arbiter and beneficiary of the system, which has resulted, at times, in comfortable margins for OCIBU and inadequate appropriations to the other members of the industry. The method of attributing remunerations on the basis of estimated budgetary requirements encourages intermediaries to inflate their costs and, in effect, constitutes a disincentive to become more efficient, all the more so since actual results of one year are not used to 'adjust" the next year's echelle mobile. Finally, the lack of transparent accounting at RDC and OCIBU levels have made it extremely difficult to determine the true costs of processing. 2.25 The Government has recognized the need for a reform of the payment system and that, for the industry to become more efficient, intermediaries should be exposed to market signals. To motivate producers toward quality, quality premiums should be introduced in producer pricing. In addition, and in order to attenuate the fluctuations of its own coffee revenues, the Government has indicated its clear preference to replace the current levy at the margin by an export tax regime. On the other hand, the Government has emphasized that one aspect of its present payment system, notably the fixed price to producers, will not to be abandoned in favor of allowing producer prices to fluctuate in line with those of the market. The Government's principal concern in this regard is that liberalizing prices to farmers would have a serious destabilizing effect on the fragile socio-political fabric of the country. 2.26 The studies from which the reform recommendations would be drawn will be carried out in 1990 under the project. These tasks include: (a) an inventory and evaluation of existing assets (currently under way) and audits of OCIBU and the RDCs; (b) the design and implementation of financial and cost accounting systems for the SOGESTALs and SODECO; (c) the review of the questions related to the export tax, the producer price and the price support mechanism by a joint IMF/IDA technical assistance mission; and (d) the formulation of the operating modalities of the new system, taking into account the findings and recommendations of the earlier mentioned assignments. 18/ The new system will be put in place in time for the 1991-92 coffee season. 18/ Details of these tasks are given in Section III.  13 G. Research. 2.27 Coffee research is the responsibility of ISABU's 19/ Industrial Crops Department. Although small in terms of resources and staff, Burundi's coffee research has remained relatively active over the years when compared to that in other countries of the region. To date, funding has been provided mostly through projects financed by IDA, CCCE, FAC and Belgian bilateral assistance. Research programs have reflected the priorities of these projects and did not always give sufficient attention to the needs of the industry as a whole. This is evidenced by the current focus on agronomic issues at the expense of processing, marketing and quality control. In the future, stronger emphasis will need to be given tot (a) the need to resolve the nagging "potato flavour" problem; (b) research in processing technology; (c) the strengthening of the extension/research linkage on coffee matters; (d) the need to evaluate the economic implications of coffee research findings; (e) better program coordination and planning; and (f) the funding of coffee research. H. Institutional Development. 2.28 OCIBU/Office du Cafe. Ever since its inception OCIBU has played a paramount role in the development of the country's coffee industry. Over time, however, the agency has cumulated a wide range of functions and activities. Some of these responsibilities, such as managing the echelle mobile, are of a regulatory nature, while others, such as operating washing stations and the hulling factories, give the agency a role at the operational level which has often resulted in conflict of interest situations. The cumulation of tasks coupled w&vb inadequate accounting and management information systems have made OCIBU unAieldly to manage. In addition, the expansion of coffee production, the greater mechanisation of processing and the requirements for more sophisticated marketing have created the need for more specialization and for a clearer distinction of the responsibilities of the industry's participants. The Government is conscious that the Industry should be more autonomous and self regulating (with private sector participation} and therefore has agreed that OCIBU will be broken up and that its role will be redefined. 2.29 The major changes will include a spin-off of OCIBU's industrial and production related activities and responsibilities to entities (some of which autonomous) either existing or to be establisheis for instance, funding for parchment collection has been reverted to the banking system; coffee extension and crop protection will be absorbed by the Ministry of Agriculture; the hulling operation, which is OCIBU's major industrial activity will be transfered to SODECO; and its washing station operations will be transfered to a SOGESTAL. Some of the priority tasks before achieving OCIBU's break-up include a full inventory and evaluation of its assets and an audit of its accounts. The charter of OCIBU will be amended to convert the institution into a regulatory authority (Office du Cafe) whose major role will be to oversee the industry as a whole. The present statutes already confer, in a general way, this role to OCIBU. The amended 19/ Institut des Sciences Agronomiques du Burundi.  14 statutes will emphasize and clarify the regulatory role of the Office and will refrain it from intervening in the day-to-day operations of the industry. The implementation of these changes will constitute one of the major elements of the institutional restructuring component of the Coffse Sector Project. I. Private Sector Participation. 2.30 Except for those activities connected with production (small farms), intermediate handling of semi-washed parchment coffee (by market traders), and some transport, all other arabica coffee related industrial activities are currently fully owned, controlled and operated by the Governaent or its agencies. Taken as a whole, the performance of this public sector industry has been reasonably good over the past decade. However, the built-in inefficiencies of its institutions, which operate both as monopolies and as branches of the administration, and the inadequacy of the payment system indicate that performance can be improved through a greater degree of competition in the industry. In line with its adjustment program, the Government recognizes this possibility and has agreed to restructure the industry and open it to private investment. However, this agreement to phase out its involvement in favor of the private sector, has not been made without hesitation. In considering the private sector participation, the Government's concerns were that: (a) access by private investors to industry assets and activities should be gradual and their participation should be beneficial both to the industry, in terms of improved efficiency, and to themselves, in terms of return on capital; (b) the measures to allow private investors access to the industry should include safeguards to prevent the repetition of past abuses (such as foreign exchange fraud, embezzlement, under-invoicing); (c) reforms of the coffee payment system should be designed to enable both private investors and the Government to draw an equitable return from their investments in the industry; in addition, the Government should be able to draw some revenue from coffee exports; and (d) industry-wide institutional reforms should provide for the Government to retain a regulatory role over the industry. 2.31 The Government's hesitations mus1: be seen in the Burundi context where private entrepreneurship, particular.y by Burundi nationals, is a relatively new development. Traditionally. Barundi have been farmers and herders, and private entrepreneurs, most o;' whom were traders, were often foreigners. Educated Barundi have generally opted for Government service. The situation has evolved somewhat in recent years as a result of the gradual liberalization of the economy and the increase in the numbers of educated nationals who have had to make a living in the private sector. Enterprises, mostly of small or medium size, are being created, a few of which are In the export sector. However, reforms are still needed to extirpate the remaining rules and obstacles which limit the cape of private investment, already constrained by the isolation of the country (and related transport difficulties) and the small size of the local market. Progress has been made under the IDA-financed SALs, but more reforms are needed to complete the liberalization process.  15 2.32 The Government's strategy to open the coffee industry to the private sector starts with the break-up of the de facto monopoly of BCC on exports. It remains constrained by the politically justifiable decision to maintain a floor producer price. The rationale is that quality premiums obtained by private exporters of fully washed coffee should allow them to offer more competitive prices to washing stations, which in turn would compete for higher quality coffee by offering premiums to farmers over the Government guaranteed floor price. This introduction of economic incentives will increase the return on investment in coffee stations and other processing facilities and make them attractive to private investors, which they are not under the present admuinistered pricing system. 2.33 Private investors will have access to all levels and aspects of the industry; they are expected to start at the marketing level and move to processing. They will also be represented on the Board of the Office du Cafe. While it is difficult at this time to foresee the rate at which they will invest in the various entities of the industry, their involvement is bound to be progressive and will depend on their perception of the viability and risks connected with the investments as well as on clarity of the role of the Government. The project has, however, been designed to permit private investment from the start. III. THE PROJECT. A. Coffee Policy and Strateay. 3.01 Given its importance in the Burundi economy, coffee plays a leading role in the Government's overall policy as a generator of foreign exchange, farm income and Government revenue. Since coffee is grown by close to half of the country's smallholdings, and the income derived from it represents some 40-60X of the monetary revenues of these small farmers, the Government is understandably sensitive to matters which may affect this income, particularly in the context of Burundi's special social fabric. The objective of the Government's policy as regards coffee is to maximize earnings from the sector by exploiting the country's comparative advantage as a relatively low cost producer of high quality coffee. The major elements of the strategy to achieve this objective comprise: (a) investing in the technology and the facilities which ensure this quality; (b) improving the efficiency at processing level through greater specialization, private investor involvement, competition and market-based prices; (c) introducing qn%ality incentives at producer level; (d) improving the efficiency at marketing level by abolishing the BCC monopoly and introducing competition by means of an auction facility for private exporters; and (e) rationalizing the regulatory function through the establishment of the Office du Cafe, whose membership would include representatives of all participants in the industry as well as Government, and by so doing reducing the latter'* direct involvement at the operational level of the industry. 3.02 Rationale. for Bank Group Support. The scope and content of the Government's coffee policy and strategy conform with its broader economic and structural adjustment objectives (para 1.04). They also conform with  16 IDA's strategy to assist the country in its macroeconomic and sectoral adjustment efforts including that of supporting increases in the share of fully washed coffee in the country's total coffee output, since such investments are demonstrably viable and are justified maco-economically. The Coffee Sector Project will help implement the country's coffee strategy and, by giving special emphasis to opening an otherwise public sector monopoly to competition and private investment, will break new grounds in Burundi. The experience thus gained will serve as basis for the design of similar interventions encompassing the whole of the economy in future structural operations. B. Background of the Proiect. 3.03 Proiect Origin and Concept. The project concept originated from the wish of the Government to accelerate its washing station construction program and the need for a sector and country-wide project to address the sector-related conditionalities of the structural adjustment program. As project preparation advanced, both Government and prospective financiers became convinced that substantial reforms were needed to enable the industry to improve its efficiency in the face of increasing production and a tighter and more competitive market. Also, it became clear that, in order to achieve this, the Government's role in the industry would have to be reduced and redefined, private sector participation would have to increase and competition would have to be introduced. 3.04 Project preparation was carried out by a small team of Burundi based staff with the assistance of consultants who undertoo,k a number of studies the most important of which dealt with the issues related to washing station management and the establishment of a private marketing facility. Important collaboration was also obtained from the Government's Committee on the Coffee Industry, from ISABU and from ECC. Funding for preparation was provided from IDA Credit 1456-BU (Third Technical Assistance Project) and from the Project Preparation Facility (P600-BU). The pre-appraisal of the project was carried out in June-July 1989, and appraisal in November 1989. C. Proiect Obiectives and Description. 3.05 Project Objectives. The fundamental objective of the project will be to improve the earning ability of Burundi coffee, the country's principal export crop. To achieve this, the project will implement: (a) a package of institutional and financial reforms to improve the efficiency of the industry and to open it up to private investors and attract private capital; and (b) investments in coffee washing stations, coffee research and training so as to increase the share of wet processed coffee, leading to higher value exports. 3-06 Prolect Description. The project, which will be implemented over a period of five years, will support institutional and policy reforms which will result in a more rational and more efficient organization of the industry. Major actions will include: (a) the creation of a regulatory authority (Office du Cafe) to replace the present OCIBU; (b) the  establishment of 5 autonomous companies (SOGESTALs) specialized in the management of coffee washing stations and one autonomous company (SODECO) to operate the country's two hulling factories; these companies as well as individual washing stations, will be open to private ownership; (c) the establishment of an Assets Management Facility to centralize the administration of Government-owned physical assets in the industry; (d) the replacement of the echelle mobile by an incentive-based payment system for producers and intermediaries and, for the Government, the establishment of a direct export tax to replace the current levy on the margin; and (e) the establishment of a parallel export marketing facility for private exporters to compete with the state-owned BCC. The major investments to be financed by the Project will includes (a) the phased construction of about 100 additional coffee washing stations, to be built by both the public and the private sector; (b) support to coffee research with emphasis on solving market related problems (e.g. off-flavors, technological issues); and (c) an industry-wide training program. 3.07 Detailed work and studies currently taking place as part of project preparation will continue throughout calendar 1990 under the project. The new entities to be established as part of the institutional restructuring will be scheduled to start operations at the beginning of 1991. The reformed payment/pricing system will be introduced at the same time. D. Detailed Features 201. 1- Institutional Reforms. 3.08 Office du Cafa. As a regulatory and coordinating authority, the principal functions of the Office du Cafe will includet (a) formulating industry policy and strategy on matters of production, processing, marketing, research and training; (b) representing Burundi in international discussions and negotiations relating to coffee; (c) monitoring national and international production trends; (d) monitoring international market trends; (e) managing the auction system; (f) supervising the activities of the industry; (g) establishing quality standards and classification systems; (h) arbitrating in matters of technical conflict within the industry; (i) monitoring the financial results and performance of the industry; (j) supervising the operation of the payment system as regards the application of the producer floor price and the hulling fees of SODECO; (k) coordinating industry-wide training; (1) promoting private sector end cooperative participation in the industry; and (k) assuming other non-industrial type activities in the coffee sector attributed by Government, including overseeing the implementation of projects and initiatives with industry-wide implications. 20/All costs referred to in this section are baceline costs.  18 3.09 The Office du Cafe, which will begin operations in 1991, will be governed by a Board of Directors composed of representatives from the major participants in the industry, including producers, traders, SOGESTALs and private washing station owner-operators, SODECO, private exporters and BCC, RDCs, the banking sector and Government. In addition to its broad sector responsibilities, the Board will also oversee the operations of the Offici. through its review and approval of budgets, financial results and audits and through the formulation of the internal operating rules of the Office. 3.10 Day-to-day operations of the Office will be managed by a General Manager who will also sit on the Board. The organization of the Office will comprise: (a) an economics department responsible for formulating strategy and policy proposals, coordinating the collection of statistical information, supervising the payment system and the financial results of the industry, and coordinating effo=ts to promote private sector and cooperative participation; (b) a technical department responsible for monitoring the industry's production, processing and research activities, formulating quality standards and classification norms; (c) a marketing department responsible for monitoring production and market trends and for managing the private auction facility; and (d) a financeladministrative unit for the oper-ting needs of the office. The unit would also be responsible for colLecting the export tax on behalf of the Government as well as other industry levies to fund coffee research, coffee road maintenance, insecticides and Office operations. (e) the Office's liquoring facility; and (f) a training unit. 3.11 The project will support the Office through funding for: (a) the ! preparation work leading to its establishment; and (b) technical assistance and the incremental cost of establishment. Regular Office operations will be financed from a levy on ccffee export revenues replacing the levy currently financing OCIBU. This levy is estimated to amount to about FBu 1,800 (USS 10.40) per tonne of exported coffee at current levels of exports. This represents about .05% of export revenues, also at current levels. The cost of operating the Office is not expected to exceed that of operating the current non-industrial activities of OCIBU. 3.12 Assurances regarding the Office du Cafe, its responsibilities, constitution, internal organization and financing have beer. included in a side letter on coffee strategy which was obtained from Government at negortations. The letter also provides assurances relative to the industry levies to finance research, insecticides, access road mainterance and the operation of the Office du Cafa. 3.13 The SOGESTALs. The SOGESTALs will have the critical responsibility of ensuring the proper technical management of the washing stations which they will lease from the Assets Management Facility. In terms of structure and equipment, washing stations are relatively simple facilities$ staffing consists of 8 permanent staff on average plus an additional 50-60 temporary workers at peak processing time. The quality of the output depends largely on the effectiveness of the technical  19 management. At the station, as coffee goes through the stages of pulping, fermentation, washing and drying, it is suojected to subtle chemical transformations which have a significant effect on the liquoring quality of the end product. The mishandling of any of these stages can have harmful effects on the inherent quality of the coffee to the point of destroying it altogether. In Kenya, the reduction of the quality of that country's smallholder coffee has been attributed, for a significant pert, to the poor management of cooperative washing stations which service smallholders. On the other hand, Burundi has, over the years, developed a management system for its stations which has had impressive results. The system combines intensive supervision 21/ with a thorough reporting system which allows the detectio. of deviations from quality standards. The system will be adopted by the SOGESTALs and will be improved to include information enabling the F measurement of the cost effectiveness of the process. Systematic follow-up of processing and cost coefficients combined with corrective action will constitute the basic tools by which SOGESTALs would be able to control their costs as well as the quality of their output to generate better returns. 3.14 SOGESTALs will own the coff_.e they will buy from coffee farmers, process it and sell it for their own account. From the beginning, they daill be able to either sell the parchment coffee to SODECO (as under the present system where parchment coffee is sold by RDCs to OCIBU) or, after having it processed by SODECO for a fee, sell the clean coffee directly through BCC or the auction. Working capital to finance coffee purchases will be provided by the banking system. 3.15 The initial capital endowment of the SOGESTALs will come from existing operating assets to be transferred from RDCs and OCIBU. These assets include vehicules, tools and some buildings. The project will provide additional resources for equipment and materials to complete the establishment needs of the SOGESTALs as well as the start-up working capital. This working capital represents the equivalent of about 9 months of SOGESTAL and washing station salaries and other operating costs to allow them to operate until coffee revenues start flowing in. The initial manAgement and operating staff of the SOGESTALs will be drawn from the staff of RDCs and OCIBU who currently operate the washing stations under these agencies. Additional personnel will be recruited and trained during the project to staff the new washing stations. Training costs will be financed by the project. Support will also be provided in the form of technical assistance in management and finance. 3.16 Given the nature of the assets which will constitute their starting capital, the SOGESTALs will initially be owned largely by the Government. However, their legal status will allow for their progressive passage from public to privately owned enterprises and the ownership of their capital shares will be open, from the beginning, to private investors. Active promotion in this regard will be carried out during the project as the SOGESTALs develop and -.re in a position to demonstrate their 21/ By a professional Itechnologue' and his assistants who will visit from 4 to 8 stations a day during the 4-5 months whan stations are processing coffee.  20 viability. The SOGESTALs will be operated as autonomous entities on a private enterprise basis and the Government will ensure that they are given all the autonomy needed to operate as efficiently as they would under conditions of private ownership. Four of the SOGESTALs (Ngozi, Kayanza, Kirimiro and Mumirwa), whose areas already include adequate numbers of washing stations to ensure viability, will start operations in 1991. The fifth SOGESTAL (Muyinga) would start operations in 1992. 3.17 Assurances regarding the SOGESTALs, their constitution, their operating autonomy, including the ownership of coffee, have been included in a side letter on coffee strategy which was obtained from Government at negotiations. 3.18 SODECO. The status of this company will be idantical to that of the SOGESTAEs in that, while initially owned by the Government, the ownership capital of the company will be open, from the beginning, to private investor participation and its legal. status will gradually be changed from a public to a private enterprise as private ownership increases. Government will ensure that SODECO is given all the autonomy reeded to operate as efficiently as it would under conditions of private ownership. As for the SOGESTALs, SODECO will lease the factories from the Government' Assets Management Facility. The initial capital assets of SODECO will come from the existing operating assets to be transferred from OCIBU. These assets will include vehicles, tools and some buildings. Funding to finance coffee purchases will be provided by the banking system. Start-up working capital will be financed by CCCE. The management and operating staff of SODECO will be drawn from the existing staff of OCIBU who currently operate the hulling factories. Other project support will be provided in the form of technical assistance in management and finance and to operate the company's sophisticated sorting equipment. 3.19 SODECO, which will start operations in 1991, will either buy parchment coffee for its own account or will process it for a fee for SOGESTALs, for private washing station owner-operators or for individual traders who would eventually elect to sell their coffee directly through the export marketing facilities. Coffee which SODECO will buy and process for its own account, will be sold for export through BCC or the auction, also for its own account. 3.20 Given that for some time SODECO will operate the only hulling facilities in the country, and therefore be in a position of monopoly, its hulling fees will be subject to the review and approval of the Office du Cafe and the methodology for reviewing and measuring the adequacy of the fees will be spelled out in the recommendations of the payment system study - (para 3.34). l 3.21 Assurances regarding SODECO, its constitution, operating autonomy, including the ownership of coffee, and the establishment of its hulling fees, have been included in a side letter on coffee strategy which was obtained from Government at negotiations.  21 3.22 Assets Management Facility 22/. The hulling factories to be leased to SODECO and most of the washing stations to be leased to the SOGESTALs or other parties have been financed by the Government from loans obtained from various donors including IDA, CCCE and Kuwait Fund. In leasing the assets, the objective of the Government will be to ensure that leasing revenues will at least cover the amortization of the debt incurred to finance them, including principal and interest. With a view to grouping the records of its past and future investments in the industry's pxaysical assets as well as maintaining proper accounts of its leasing revenues and debt servicing, the Government will establish an Assets Management Facility (AMF). The main tasks of the AMF will be tr negotiate the leases and maintain an assets register and a set of accounts to record revenues and debt service transfers to Treasury. 3.23 The powers of the AMF will include the capacity to enter, on behalf of the Goverrment, into leasing agreements with the SOGESTALs, SODECO and other parties. The modalities, terms and conditions of the leases will be spelled ouat in the course of 1990 with the assistance of legal and financial specialists recruited in the course of project preppnation and in the early stages of project implementation. In addit.on to the standard features of the leasing agreements, the leasing contracts - will -clude arrangements for the maintenance and repairs of the assets, a responsibility to be assumed mostly by the lessees. This will be an important aspect of the leasing arrangements given that the assets are scattered around the country and the difficulties for the AMF to act quickly when urgent repairs or replacements are needed particularly during the months when the leased facilities would be operating. While the underlying need for these arrangements will be to ensure the efficient maintenance and repair of the assets, there is the deliber&te intent on the part of the Government to ensure that the organization of the AMF will remain small and not be a financial burden on the country's coffee industry. 3.24 In addition to its leasing powers, the AMF will have the authority to sell, on behalf of the Government, the assetb to private investors. The leasing arrangements with SODECO, SOGESTALs and others will also include clauses to allow the lessees to eventually gain ownership of assets which they will lease. The definition of the modalities for these arrangements will be the task of the financial and legal consultants referred to earlier. The consultants will also have the task of recommending the mcst appropriate status for the Facility. Proposals have already been made in this regard one of which included contracting out the operation to a private firm and another to entrust it to the Office du Cafe. The facility will be established in 1990 in time for SODECO and the SOGESTALs to start their own operations. 3.25 Assurances regarding the establishment of the AMF, its leasing powers of the AMF, its authority to sell the assets have been included in a side letter on coffee strategy which was obtained from Government at negotiations. 22/Service du Patrimoine.  22 3.26 The Private Auction Facility. The Government has agreed to the establishment of a new auction facility where coffee will be auctioned off to private exporters. The facility will be open to all exporters, foreign and national. It will start operations in 1991 and will auction, during 1991 and 1992, a pre-established proportion of the crop which will increase to 50% by 1992 23/, with the remainder being sold by BCC. Preparation to set up the auction will be led by the project's marketing advisor assisted by a local consultant, and a Preparation Committee has been established to oversee their work. The Committee is composed of potential private exporters, representatives of the banking system, industry agencies and Government. Preparation tasks will include: (a) formulating recommendations for establishing the auction; (b) drawing up its operating rules; (c) identifying constraints in export regulations, procedures and practices, and recommending remedial measures; and (d) organizing a promotion campaign to attract Participants. The terms of reference for the preparation committee and the consultants have been approved by both the Government and IDA. A condition of effectiveness of the IDA Credit Is that the recommendations of the Preparation Committee be submitted to the ioint review, comments and agreement of the Government, IDA and CCCE. Assurances were obtained at negotiations to the effect that, on the basis of this agreement, the Government will establish the private auction facility by December 31, 1990. 3.27 The Office du Cafe will be responsible for overseeing and monitoring the operation of the auction 24/. In 1992, after two years of operations, the Government, IDA and CCCE will carry out an in-depth review of the performance of the auction. The performance of BCC, over the same period, will also be evaluated and compared with that of the auction. Based on the results of the review, consideration will be given at that time to either: (a) market the whole crop through the auction system; (b) continue sharing the crop between the auction and BCC; or (c) abolish the auction. An agreement between the Government, IDA and CCCE on the recommendations of the review and the commitment of the Government to implement them will be obtained by December 31, 1992. Assurances to this effect were obtained at negotiations. 3.28 Assurances regarding (i) the establishment of the auction, (ii) the joint agreements relating to its rules of operations and entry and the measures to remove operating constraints, and (iii) the review and agreements relating to its performance, have been included in a side letter on coffee strategy which was cbtained from Government at negotiations. 3.29 BCC. A number of measures, already approved by Government, still have to be implemented to provide BCC with the organization and staff it needs to perform and compete effectively. These include: (a) the 23/For 1991t 12,000 tonnes, or 30% if the crop exceeds 40,000 tonnes. For 1992: 50% of the crop. 241In a manner similar to that of the Coffee Board of Kenya with regard to its fully-owned subsidiary, Kenya Coffee Auctions Ltd. r7i  23 establishment of a Marketing Department; (b) the establishment of competitive conditions of service for its staff; and (c) the recruitment of a Marketing Manager. The Government has also agreed that measures relative to staff conditions of service will be implemented before undertaking the recruitment of a Marketing Manager in order to attract the best candidates possible for this post. Recruitment will be on the basis of open competition. The implementation of staff conditions of service and the appointment of the Marketing Manager on the basis of competition are conditions of effectiveness of the IDA Credit. 3.30 With the establishment of the private auction facility, BCC will no longer have the monopoly of marketing the total coffee crop although it will retain a specific share of the crop at least over the first two years of operation of the auction. It was agreed during appraisal that as long as BCC retains a share of the crop, it will not participate at the auctions of the privat.e auction facility to avoid the perception, on the part of private exporters, that BCC onjoys advantages and possibilities incompatible with open and fair competition. It was also agreed that, as and when the auctions become the only marketing channel for Burundi coffee, BCC will become an exporter like all others participating in the auction. 3.31 Assurances regarding BCC's role in relation to the private r auction facility have been included in a side letter on coffee strategy which was obtained from Government at negotiations. 3.32 The Payment/Pricing System. The replacement of the current pricing system will constitute one of the most important reforms to be implemented under the project. The new system will: (a) convert the current fixed producer price into a floor price; this will still guarantee Goverranent intervention to support the producer price if it falls below the floor price level, bat will enable producers to be paid higher prices if quality differences, market conditions and increased competition allow for it; (b) replace the current Government levy on the margin (para 2.23) by a progressive coffee export tax which will attenuate inter-annual swings in fiscal revenue and provide, in conjunction with exchange rate adjustments more flexibility in internal prices to absorb world market price variations while maintaining a floor price; and (c) allow greater flexibility to industry participants to set intermediate prices. 3.33 Before the new pricing system can be implemented, a number of important questions need to be resolved. First, the level(s) of the export tax needs to be determined: this has important macro-economic implications as such taxes have usually been an important part of fiscal revenue. Also, as the new floor price system will still include a potential Government obligation on price support, it will have to be determined whether: (a) part of the new export tax revenue will have to be earmarked as reserve against such liabilities; (b) both export tax revenue and producer price support expend'tures will be an integral part of the Goverrment budget; or (c) the price support reserve will be collected as a separate levy from the export tax. Finally, to be determined also will be the extent to which the Government should intervene through exchange rate revaluations to compensate for extreme market price movements, in light of the impact of such revaluations have on other sectors of the economy. These questions  24 will be reviewed by a joint Bank/IMF technical assistance mission which will advise Government at the latest by September 1990. The mission has already been requested by Government. 3.34 A second study will need to be carried out on: (a) the modalities which will determine SODECO's purchase and selling prices and/or hulling fees, given the company's monopoly position; (b) the modalities for establishing the commission which BCC would charge for its services; (c) the mechanism for paying quality premiums to producers 25/; and (d) the issues related to crop and working capital financing by the banking system (see next paragraph). The appointment of the consultants to undertake this study is a condition of effectiveness of the IDA Credit. Assurances were obtained at negotiations that the study's recommendations will be submitted to a joint GovertmentlIDA review and implementation by October 30, 1990 and that the new pricing/payment system would be implemented at the latest by February 28 1991 in time for the 1991 coffee season. 3.35 Under the current pricing/paymvent system, OCIBU, in view of its central position in the industry as the only purclaser of parchment coffee, is responsible for negotiating arrangements with the commercial banks to iinance the industry's coffee purchasing operations. In turn, commercial banks are re-financed by the Banque de la Republique du Burundi (BRB), the country's central bank, at a fixed 5% rate, to which financial institutions may add a maximum 1Z margin. This margin has been found adequate by the banks in light of the low risk associated with OCIBU's monopoly position. This arrangement will end, however, with the dismantling of OCIBU and of the administered pricing system, the result of which will be that SOGESTALs, SODECO and other intermediaries will henceforth have to deal directly with the banks. Moreover, the Minister of Finance has recently indicated that the preferential 5% rate will be abolished starting with the 1990-91 coffee season and that coffee operations will be subjected to BRB's normal refinancing rate which is presently set at 8.5%. This measure would be in harmony with the recent dialogue between the Government, the Fund and IDA on further liberalizing interest rates. Although the matter has yet to be decided, it is highly likely that the bank margin will also be subject to increase reflecting the higher level of risks connected with the new situation, thus bringing the effective rate for coffee transactions closer to market conditions. 3.36 One of the keystones of any quality based pricing system is the methodology for classifying products by quality grades. The classification system currently in use in Burundi has changed little from colonial times and aeeds to be adapted to conform with current processing techniques and marketing requirements. Recommendations have already been made for a new system but decisions have yet to be made to move ahead with it. A first step in this direction will to test and fine tune the proposed system during the 1990 coffee season. The second step will be to draw and promulgate legislation to establish the new system in time for the 1991 coffee season. The marketing consultant of the Office du Cafe will assist the industry in implementing this program. 25/ The technical criteria to measure quality would be determined with the assistance of ISABU (para 3.54).  25 3.37 Assurances regarding the studies, recommendations and agreements relative to the pricing and classifications systems, as well as confirmation regarding the interest rate for the financing of coffee purchases have been included in a side letter on coffee strategy which was obtained from Government at negotiations. 3.38 Privatization. In the context of the Government's macro- economic objectives, the project will provide opportunities to stimulate private sector development. However, given the history of public sector control over the coffee industry and the present weakness of the private sector in Burundi, it is anticipated that the entry of private exporters and investors in the industry will start slowly and will depend on their respective perception of0 (a) the ease of access at the various levels of the industry; (b) their ability to earn a fair return on investments in a eontext of liberalized prices; (c) their ability to finance their acquisitions and operations; (d) the degree of operational flexibility resulting from the removal of administrative constraints and the reduction of Government intervention; and (e) their expectations regarding the future of the coffee industry. 3.39 As a result of the reforms to be implemented under the project, private entrepreneurs will have access to the industry as: (a) buyers and exporters of coffee through the auction; (b) shareholders in the SOGESTALs and SODECO; (c) investors in washing stations that they will either construct themselves or buy or lease from the Assets Management Facility. Private investors will also be able to own and construct hulling factories. Initially (end 1990, early 1991), the private auction facility will likely be the most attractive form of entry to private entrepreneurs in view of the relatively low cost of establishment and the ability to trade the coffee on the world market. Of high relevance to retain the interest of exporters would be the regularity of the auctions and the fact that coffee is actually sold in the course of these auctions. Entry at the processing level would occur later (starting possibly by the end of 1991) as entrepreneurs become convinced that the policy and institutional reforms are taking hold. Of particular importance to the latter will be the reform of the pricing system which will aim at giving washing stations, SOGESTALs and SODECO a greater degree of freedom at controlling and recovering their costs and seeking the highest market prices possible, within the constraints of a floor producer price. 3.40 Due to the present lack of organizations to promote private sector interests in the industry, the responsibility for this task will initially fall on the Office du Cafe through a special committee created for this purpose. This Committee will be composed of members of the Office and representatives of the Chamber of Commexce and other private sector interests. The role of the Committee will be to identify constraints on private investments and to devise measures to overcome them. The Committee will become operational at the latest by June 30, 1991 under terms of reference and work plan acceptable to IDA. Assurances to this effect were obtained at negotiations.  26 3.41 To finance their equity, working capital requirements and acquisitions of existing installations, private entrepreneurs will draw from their own resources as well as those of the banking system. A World Bank mission, which reviewed the banking sector at the time of appraisal, reported no shortages of liquidity in the system. It is expected that foreign private exporters and their local representatives will pre-finance their establishment costs from the resources of their parent organizations. To finance new constructions at the processing level, private investors will have access to existing lines of credit in the banking system, the most important of which, at the moment, is that provided under the IDA- funded Small Enterprise Apex Project (Credit 1889-BU). 3.42 This Apex project, which became effective in March 1989, is designed to provide term financing to small to medium scale enterprises (SMEs), through local commercial banks and other financial institutions. The project is part of the Government's strategy to develop the sector and ts oupported by IDA's own country assistance strategy in both the financial and SHE sectors (para 1.03). The project is managed through the Banque de la Republique du Burundi (BRB), the local central baik, which lends Credit funds to participating financial institutions (PFIs) 26/ which, in turn, on-lend to the SMEs. Technical assistance to appraise SME projects is I provided oy the local Chamber of Commerce (CofC). Proceeds of the Credit are made available to PFIs at a reference rate equivalent to the average yield of 3-month Treasury certificates. In view of the liberalized interest rate policies, including alignment of coffee rates to the general regime (para 3.35) and the continued flexible exchange rate policy applied by the Government (para 1.05), the foreign exchange risk can be assumed by the Government which is adequately compensated under the relending rate of IDA Credit proceeds to PFIs. 3.43 Implementation of the Apex project has been slow and the project's USS 7.2 million equivalent available for small-scale investments have yet to be committed one year after effectiveness. However, the administrative and interpretational problems which initially hampered implementation have now been worked out 27/ and a recent supervision mission (November 1989) expressed confidence that commitments should start in the near future. The assessment of the mission is that the pent up demand for the resources of the credit is considerable and that once lending is started these could be utilized quickly, with a high probability that most of the line of credit could be committed by end 1991. The objectives, institutional arrangements, terms and conditions of the Apex project are, for the most part, ideally suited for the type of financing needed by private investors for the construction of new coffee washing 26/ Essentially 3 commercial and 2 development banks. See Annex 5 for conditions of eligibility. 27/ The problems centered on the role which the CofC had assumed for itself notably by insisting that its approval was needed for BRB to on-lend to PFIs. This was contrary to the original intent whereby CofC's role was essentially one of providing technical assistance in the appraisal of projects submitted to PFIs. This misunderstanding has resulted in an overly cumbersome approval process, literally paralysing the project.  27 stations. Since these investments are demonstrably viLble both financially and economically, they would constitute attractive investments for private entrepreneurs as well as worth while additions to the loan portfolios of the lending institutions. However, given the time needed to implement the industry reforms and particularly that of the pricing system to create the appropriate investment conditions, private entrepreneurs will not likely venture into washing station investments until the second half of 1991, at the earliest, by which time most of the resources of the current Apex credit will have been conitted. 3.44 To ensure that private entrepreneurs have access to financing to construct their own washing stations, the IDA Credit resources for the washing station construction program of this Coffee Sector Project will be made available from the beginning to private entrepreneurs as well as to the public sector. This approach is consistent with and enhances IDA's strategy to support Government's efforts to stimulate private sector involvement in the coffee industry. Private lavestor access to the funds will be through en-lending arrangements and conditions aimilar to those of the on-going Apex project although procurement conditions will be those of this Coffee Sector Project (para 3.69). The modalities of this arrangement are given at paras 3.47 and 3.67 of this report. Public sector financing will be done through stand3rd procedures. 3.45 Assurances with regard to private investor participation at various levels of the industry and the financing of the construction of project washing stations by private entrepreneurs have been included in a side letter on coffee strategy which was obtained from Government at negotiations. 2 - Investments. 3.46 Washing Stations. Given the economic justification to produce fully-washed coffee, particularly in the current context of a world market without ICA quotas, it is urgent for the country to proceed with its washing station construction program. Some 100 coffee washing stations will be built under the project and will constitute the project9s major investment, increasing Burundt.'s production capacity of fully washed coffee by some 20,500 tonnes to approximately 37,000 tonnes at full development. At an estimated unit cost of FBu 42.9 million (US$ 248,000), including civil works, equipment and access roads, the total cost of this program is estimated at FBu 4,292.6 million (US$ 24.8 million). The stations will be constructed at sites which will be selected on the basis of coffee production potential, physical suitability and economic viability. A condition for funding will be that each station be justified on the basis of a feasibility study carried out on the basis of procedures and criteria acceptable to IDA. Assurances to this effect were obtained at negotiations.  28 3.47 The annual building program 28/ is based on the existing construction capacity of the country and assumes an increase in pace as experience would be gained in building the facilities. Given that the extent of private entrepreneur interest in such investments cannot be assessed reliably before the reforms are in place, it was not possible at appraisal to determine the number of program stations that will be taken up by private entrepreneurs. Thus, no specific allocation has been made between the private sector and the public sector. However, IDA Credit financing has been designed to provide sufficient flexibility to ensure that private entrepreneurs will have access from the beginning to funding as and when needed. It will also ensure that, whatever happens in this regard, the balance of the construction program can still be met by the public sector. 3.48 Washing stations vill be constructed by contractors selected through local competitive bidding procedures which will not exclude international bidders (para 3.69). These requirements will apply equally to public as well as to privatAly constructed stations. 3.49 The design of washing stations to be constructed under the project will include connections for future retrofitting of recirculation systems and siting of related equipment. These measures are needed in the event that discharges from washing stations into adjacent water streams exceed pre-determined levels of naturally degradable pollution. ISABU will establish and maintain a monitoring system on pollution of water streams in the vicinity of all washing stations and, in case of excessive pollution, will issue recommendations to remedy the situation. Implementation of ISABU's recomendations will be the responsibility of the owners of washing stations. Compliance will be enforced by Government through the Ministry responsible for environment matters. 3.50 Assurances with regard to the washing station construction program, private sector involvement and anti-pollution measures have been included in a side letter on coffee strategy which was obtained from Government at negotiations. 3.51 Access Roads. Roads leading to washing stations will be constructed by the contractors responsible for station construction. The project will provide for an estimated average of 3 kilometers of access road per washing station at an estimated cost of FBu 500,000 (US$ 2,900) per kilometer. The conditions attached to the funding of the roads (para 3.63) require that labor intensive technology be utilized for their construction. Through an agreement between the Ministry of Agriculture and the Ministry of Public Works, the Direction Generale des Routes (DGR) of the latter ministry will supervise the contractors' road construction operations. Maintenance of the roads will be the responsibility of the washing stations to which they give access; funding for maintenance would be ensured through a levy on coffee export receipts (paras 2.04, 3.10 and 3.12). 281 14 in 1990; 20 in 1991; 22 each for 1992,1993 and 1994.  29 3.52 Research. The project will support a comprehensive coffee research program ir.cludingt (a) soil conservation; (b) washing station technology; (c) off-flavor research (potato flavor); (d) coffee husbandry; (e) coffee agronomy; (f) soil and leaf analysis; (g) entomological research; (h) coffee berry disease (CBD) and rust research; (i) varietal research; (j) coffee seed production; (k) on-farm trials; and (1) coffee- related socio-economic studies. Project resources will also be provided to support the development of a documentation facility, the organization of conferences and symposia, improved communications and relations with international research estabaishments, and training. Research related training will be coordinated by ISABU separately from other project training. 3.53 It will be important to continue improving the liaison between research and the coffee industry and strengthening the relationship between research and agricultural extension. A coffee research coordinator in post before appraisal has been promoted and a replacement has not. yet been appointed. Given its importance, the filling of this position will be a condition of Credit effectiveness. Moreover, in view of the need to strengthen the liaison between research and extension and of improving the research response to farmers' requirements, this particular activity will be the object of a specific annex to the annual work programs which will comment on the results of the past year and propose improvements and new initiatives for the coming year. 3.54 Another important feature of the project will be the introduction of a levy on coffee export sales to finance coffee research operating costs. Up to now, these costs have been covered largely by donor funding; this situation will continue for the first two years of the project but will be gradually p'nased out so that, by the end of the project, all coffee research operating costs will be funded from the proceeds of this levy. It is expected that, with funding coming from coffee returns, the industry will require that its concerns be given close attention in future research programs; this will be a healthy development (para. 3.12). 3.55 Under the project, ISABU will also develop practical measurement standards to evaluate the quality of coffee delivered by farmers. This will be a significant achievement as it will enable the establishment of incentive payments to farmers based on the quality of their coffee. 3.56 Assurances with regard to the measures to strengthen research- extension linkages and quality measurement standards have been included in a side letter on coffee strategy which was obtained from Gcvernment at negotiations. 3.57 Training. Although the restructured industry will be in a position to offer attractive conditions of service, finding technically competent staff is likely to be difficult. The industry is already short of competent staff in such areas as marketing, liquoring and financial management. While existing washing stations are adequately staffed, more staff will be needed for the stations to be constructed under the project.  30 To meet these needs and to ensure that staff are given the particular skills required by the industry, the project will provide resources to train existing and future staff in: (a) coffee liquoring, for a total of 30 staff in neighboring and overseas countries; (b) coffee marketing, for 6 staff ia overseas institutions and with international coffee exporters/importers; (c) production, processing and wharehousing technology, for 27 staff in neighboring and overseas countries; (d) administration and financial management, for 8 staff in overseas institutions; and (e) washing station operation and management, for a total of 650 staff through a locally operated program. 3.58 The Office du Csfe will have the responsibility for the organization and coordination of the program 291. To assist the Office in this function, a training consultant with specialization in human resource management, will be recruited at the beginning of the project. In the course of periodic missions, this consultant wills (a) design the annual training activities; (b) organize and coordinate the implementation of the program; and Ic) assist the industry entities in defining their staff profiles and needs as well as their training requirements. Assurances were obtained at negotiations that the training specialist be appointed by December 31, 1990. E. Pro1ect Costs. 3.59 Project costs have been estimated at Fnu 8,010 million or US$ 46.3 million of which FBu 3,507 million or US$ 20.3 million represent foreign exchange costs. Base cost estimates reflect prices at end December 1989. Taxes have been estimated at FBu 893.8 million or US$ 5.2 million, about 112 of total project costs. Equipment imported for the project and expenditures related to foreign technical assistance would be exempt from Import duties. The local transaction tax of 15% would apply. 3.60 Physical contingencies amounting to 101 have been applied to selected categories to reflect uncertainties regarding detailed quantities and possible design modifications. Price contingencies have been calculated for local and foreign exchange costs on the basis of projected inflation rates as shown in Table 1: 29/ Except for research related training which would be coordlnated by ISABU (para 3.51)  31 Table 1 Price ContinRencies Project Year 1 2 3 4 5 Calendar Year 1990 1991 1992 1993 1994 -________-_---- percent ------------------ Local 7.0 6.0 5.0 5.0 5.0 Foreign 7.2 4.4 4.4 4.4 4.4 3.61 Project costs are summarized in Table 2.  -32- Table 2 ~1 FeEm SECTOR PROECT COST SUJY (FlU '000) (USt '000) 2 total - 2 Foreisg Base Local Fowign Total Local Poprign Total Exchange Costs A. INSTITUTIONAL RESTRMURINS OFFICE WI CAFE 1524001.5 148P488,5 300M490.0 878.6 85P.3 1736.9 49 5 SOESTALS 630*827.0 245193.0 976.010.0 3,646.4 1,417.2 5063.6 28 13 SODECO 12,456.0 524324.0 64,790.0 n.o 302.a 374.5 81 1 PCC 9704.0 41*316.0 51420.0 56.1 238.8 294.9 81 1 Sub-Total INSTITUTIONAL RESTRUCTtIRIN6 890,93.5 W87e311.5 1292.300.0 64653.1 2,916*. 79469.9 38 ^ 3. IISNI1' STATIONS 2,4684263.4 1U856t 990 4t3259062.4 14,267.4 10,732.9 254000.4 43 66 C. RESEARCH CWtEE RESEARCH COORDINATION 199074,2 13M883.8 32995890 110.3 80.3 190.5 42 1 SOIL CONSERVATION RESEARCH 1*405.0 500.0 190.S0 8.1 2.9 11.0 26 0 VASNINs STATION TECNO4ISY 42,883.9 76,206.2 119,090.0 247.9 440.5 688.4 64 2 OFF FLAVOUR STUDIES 37,812.4 82,962.9 120775.3 218.6 479.6 698.1 69 2 HUSBANDRY IWROW. TN BUT., KIR. AND ERAS. 319074.5 18491.5 49,566.0 179,6 106.9 286.5 37 I AVRONWM (RECONRCNE ELARSIE) 71,370.2 84.919.6 156,289.8 412.5 490.9 903.4 54 2 SOIL AND LEAF NALYSIS 28,392.0 46,668.0 759560.0 167.0 269.8 436.9 62 1 ENTOMOLOGY 131048.0 10,636.0 23*684.0 75.4 61.5 136.9 45 0 CID AND RUST RESEARRH 274616.0 13,184.0 40h800.0 159.6 76.2 235.9 32 1 VARIEr.4L RESEARCH 23h490.0 10,584.4 34#074.4 135.9 61.2 197.0 31 1 COFfEE SEED PRODUCTION 4,011.5 2,298.5 6,310.0 23.2 13.3 36.5 36 0 SOCIO-ECNMDIC STUDIES 11,518.0 7,616.0 19,134.0 66.6 44,0 110.6 40 0 MONITORING AND EVALUATION 49470.0 900.0 59370.0 25.8 5.2 31.0 17 0 DICUMENTATION FACILITY 4,688.4 4g731.6 9,20.0 27.1 27.4 54.5 SO 0 ON FARR TRIALS 16*929.0 79626.0 24P5550 97.9 44.1 141.9 31 0 RESEARCH TRAINING - 49,080.0 49,080.0 - 283.7 283.7 100 1 LIAISON AND CONSULTANCIES 5,308.0 2442.0 7*750.0 30.7 14.1 44.8 32 0 Sub-Total RESEARCH 343,591.1 4329730.4 776,321.5 1,986.1 2M5OW.3 4,487.4 56 12 D. TRAINING 294250.0 96,320.0 125,570.0 169.1 W6.8 725.8 77 2 Tothl BSELINE COSTS 36464093.0 2.873.160.9 6,519,253.9 214075.7 16607.9 37683.5 44 1O0 Phssical Contirngeies 3009061#3 234,504,1 534,565.4 1.734.5 1,355.5 39090.0 44 8 Price Contin*ncies 5574590.2 398t964.4 956,574.6 3P223.1 2,306.3 5,529.3 42 15 ,..tal PROJECT COSTS 4.503.744.5 3P5069649.4 8,010,393.9 264033.2 20,269.6 46,302.9 44 123 Martch 5. 1990 16:25  33 F. Financina 3.62 The financing plan of the project is summarized in Table 3: Table 3 Financing Pian Million FBu Million US$ % of total IDA 4,851.4 28.0 62 Government of Burundi 1,614.6 9.3 20 CCCE 1,047.9 6.1 13 French Assistance 251.4 1.5 3 CIRAD 73.7 0.4 1 Belgian Assistance 171.4 1.0 1 Total 8,010.4 46.3 100 3.63 The proposed IDA Credit of SDR 21.3 million (US$ 2!.8 million equivalent), including a PPF of US$ 650,000 will finance about 60% of total project costs, or 682 of project costs net of taxes and dutles. An additional US $ 9.0 million equivalent of external financing will be provided by Belgian Government assistance and by three French agencies, namely: (i) the CCCE; (ii) the French assistance agency Fonds d'aide et de cooperation (FAC): and (iii) the Centre de cooperation internationale pour la recherche agricole et le developpement (CIRAD). The Belgian assistance grant will finance the construction of access roads to washing stations. The agreement between the Belgian and Burundi governments relative to this financing is expected to be finalized by June 1990 and will be a condition of effectiveness of the IDA Credit. The CCCE loan will finance the construction of 20 project washing stations and some operating costs. The agreement between CCCE and the Burundi Government is expected to be finalized at the latest by December 31, 1990 and assurances to this effect weru obtained at negotiations. The FAC and CIRAD grants will finance the investment, technical assistance and some operating costs of three research sub-components 30/. The agreement between the Burundi Government and these agencies was concluded on September 27, 1989. 3.64 The Government will finance the equivalent of 20.2Z of total project costs, or 10.1% of total project costs net of taxes and duties. It is expected that the Government's contribution will be drawn largely from existing levies on coffee export revenues. 3.65 Coffee research operating costs have in recent years been totally funded by external financing notably through the Ngozi III and Kirimiro projects. For the first two years, this proposed project will continue this practice. However, starting in year 3, project funding of 3O/Processing Technology research, Off-Flavour research and Agronomy research.  34 these expenditures will gradually decrease 31/ and be replaced by funding from a research levy on coffee export revenues. By the end of the project all research operating costs will be funded from this levy 321. 3.66 The project will be implemented on the basis of annual work programs which will be subject to the prior review and approval of Government and the financiers. An agreenent in this regard was obtained at negotiations. 3.67 IDA funding will be made available from the beginning for financing washing stations to be constructed by private entrepreneurs through arrangements, terms anu conditions generally similar to those of the on-going Small Enterprise Apex Project. The only variations from the already agreed procedures will include: (a) technical assistance for the preparation of all washing station sub-projects will be provided by the Implementation Unit of the Office du Cafe (para 4.12) instead of the Chamber of Commerce under Apex; (b) the first two sub-projects (instead of five) from each financial intermediary will be subject to prior IDA review and approval; (c) the procurement procedures will be those of the Coffee Sector Project; and (d) disbursements for 80% of local sub-project expenditures (instead of 70%). The specific arrangements for this financing will be the object of a Project Agreement between IDA and the BRB and of a Management Agreement between the BRB and the Government. The conclusion of the latter agreement will be a condition of Credit effectiveness. G. Procurement. 3.68 All goods and services financed under the IDA Credit will be procured in accordance with IDA guidelines and as outlined in paras 3.69 to 3.72 below. Assurances to this effect were obtained at negotiations. A World Bank mission visited Burundi in March 1986 to assess the country's procurement procedures and found that, on the whole, existing rules and procedures are acceptable by IDA guidelines. A notable exception concerned the matter of preference given to local firms under local competitive bidding conditions. This matter has been corrected by means of a side letter provided by the Government at negotiations providing assurances that it will nonform with IDA guidelines for goods and services to be financed by the Credit. 3.69 Prior IDA review of terder documents, bid evaluation, sward and contract signature will be required for all contracts exceeding US$ 100,000 equivalent. IDA-financed civil works related to washing station construction, to be built by either the public or the private sector, amounting to an eetimated cost of US$ 14.5 million, and to the construction of offices and laboratories, amounting to an estimated cost of US$ 0.35 million, will be procured through local competitive bidding (LCB) with 31/ Years 1 and 2: 100%; year 3: 802; year 4: 50%; year 5: 201. 32/This is the on-going practice in Kenya where a 1% levy on 'Payments to Planters' is earmarked for coffee research.  35 foreign firms allowed to participate. This procedltre will be appropriate as the isolated location of the country and the nature of the works (i.e. scattered over 14 to 22 sites annually, totalling about US$ 3.0 million per year) make it unlikely that firms not based in Burundi will be Interested in bidding. There are several local construction firms in the country (some of which are local representatives of large internaxional firms) with adequate expertise to undertake the works program. However, in cases where the number of bidders will be less than three, the Borrower will be required to consult with IDA before award. Moreover, since few of these firms have experience in the construction of washing stations 33/, and in order to minimize the risks of default, prospective bidders will be invited to pre-qualify. 3.70 IDA-financed vehicles and equipment will amount to an estimated cost of US$ 5.6 million; contracts for these will be grouped into packages of at least US$ 100,000 whenever possible. This will apply particularly in the case of washing station macninery and equipment, which will be purchased in annual lots, and vehicles. Contracts whose value exceeds an estimated US$ 100,000 will be procured through international competitive bidding (ICB). LCB will be permitted for contracts whose estimated value will range between US$ 20,000 and US$ 100,000. Contracts whose estimated value will amount to less than US$ 20,000, for an agregate amount not to exceed US$ 0.6 million, could be procured through international or local shopping on the basis of a minimum of 3 quotations. IDA-financed technical assistance, whose estimated cost will amount to USS 2.2 million, will be contracted following IDA guidelines. IDA-financed training activities, totalling an estimated US$ 1.2 million, will be subject to IDA approval. 3.71 Contracts for civil works, goods, equipment and services to be financed by other financiers will be subject to the latter's procurement procedures. These will include: (a) washing station civil works, for an estimated US$ 3.8 million by CCCE; ((b) offices and laboratories, for an estimated US$ 27,000, by FAC; (c) vehicles and equipment, for an estimated USS 0.9 million, by FAC and CCCE; (d) technical assistance, for an estimated USS 2.5 million by FAC, CCCE and CIRAD; and (d) washing station access road construction, for an estimated USS 1.0 million, by the Government of Belgium. Washing station civil works financed by CCCE are expected to be procured under LCB procedures similar to IDA's. 3.72 Procurement arrangements are summarized in Table 4. 33/Except for a few stations financed by CCCE, most washing stations in the country were constructed under force account.  36 Table 4 PROCUREMENT ARRANGEMENTS (US$ million) --- Type of Procurement --- Project Element ICB LCB Other N.A. Total Civil Works Land Aquisit. 0.30 0.30 (-) (-) Washing Stat. 22.90 22.90 (14.5) (14.5) Offices & Labs. 0.35 0.35 (0.27) (0.27) Vehicles & Equip. 6.9 1.2 8.1 (5.6) C_) (5.6) Tech. Assistance 4.7 4.7 (2.2) (2.2) Trainir.g 1.30 1.30 (1.22) (1.22) Increm. Op. Costs 8.6 8.6 (4.3) (4.3) Total 6.9 23.3 5.9 10.2 46.3 (5.6) (14.77) (2.2) (5.52) (28.0) Notet Figures in parentheses are amounts financed by IDA. H. Disbursements. 3.73 The proceeds of IDA Credit will be disbursed over 7 years (Annex 2) based on Burundi's disbursement profile. All disbursement requests will be fully documented with the exception of those supported by statements of expenditures (SOEs). Disbursements for expenditures up to US$ 20,000 equivalent per item or contracce will be against SOEs. Disbursements will be centralized in OCIBU/Office dts Cafe where documentation pertaining to SOEs will be retained. SOE related accounts and supporting documents will be available for inspection in the course of IDA supervision missions. All project related disbursement applications will be jointly signed by the Director and the finance director of OCIBU/Office du Cafe. 3.74 In order to facilitate the flow of funds and to ensure that they are available when needed, a special account will be opened in US$ in a financial institution acceptable to IDA, for all operations to be coordinated by the Office du Cafe. IDA will deposit an amount of US$ 1.5 million in the special account; this amount corresponds to an average of three months of expenditures eligible for IDA financing. Checks written on this account will be jointly signed by the Director and the finance director of OCIBU/Office du Cafe. Claims will be grouped for re- imbursements of not less than US$ 20,000 and IDA will replenish the account accordingly. Transactions for local currency expenditures will be effected  37 at the prevailing exchange rate of the day. For loans to private entrepreneurs to be financed by the Credit, disbursement procedures will be identical to those of the Apex project. 3.75 IDA disbursements by category, amount and percentages will be as shown in Table 5 below. Table 5 Schedule of Disbursements Catetorv US $ Million X of Expenditures 1. Civil Works of 14.1 Washing Stations (a) Public Sector 100% of foreign expenditures and 80% of local expenditures (b) Loans to private 100% entrepreneurs 2. Civil Works of 0.2 100% of foreign expenditures Offices and labo- and 801 of local expenditures ratories 3. Machinery and tools 4.0 for washing stations (a) Public Sector 100% of foreign expenditures and 80% of local expenditures (b) Loans to private 100% entrepreneurs 4. Office equipment 0.1 100% of foreign expenditures and furniture and 801 of local expenditures 5. Research inputs 0.1 100% of foreign expenditures and equipment and 80% of local expenditures 6. Vehicules 1.25 100Z of foreign expenditures and 80% of local expenditures 7. Technical 1.4 1001 Assistance S. Training 1.4 100% 9. (a) Research 1.2 1001 (1990-1992) Operating 80% (1993) Expenditures 502 (1994) 20% (1995)  38 (b) Operating 0.1 1002 of foreign expenditures Expenditures of and 60Z of local expenditures the Office du Cafe (c) Operating 1.6 1002 of foreign expenditures Expenditures of and 602 of local expenditures SOGESTALs 10. Refunding of PPF 0.65 11. Unallocated 1.8 Total 28.0 3.76 Government counterpart funds for civil works, equipment, vehicles and operating expenditures will be made available to OCIBU/Office du Cafe on a quarterly basis. Each 4nstalment, which will correspond to one fourth of the estimated annual contribution, will be deposited in a Project account administered by OCIBU/Office du Cafe. Assurances in this regard were obtained at negotiations. An initial deposit of FBu 35.0 million will be a condition of effectiveness of the IDA Credit. I. Accounts and Audits. 3.77 Independent auditors, to be recruited in the first half of 1990, under terms and conditions acceptable to IDA, will carry out the audit of OCIBU. Funding for this assignment has been provided in the PPF. A formal audit of OCIBU's accounts as of December 31, 1989 will be a condition of effectiveness of the IDA Credit. Audit requirements for the RDCs are governed under the Credit Agreement for the ASSP. 3.78 Separate project accounts will be maintained by each of the agencies responsible for the execution of the project (OCIBU/Office du Cafe, ISABU) or to be created in the course of the project (SOGESTALs, SODECO, Assets Management Facility). The financial and cost accounting systems of the latter (including the Office du Cafe) will be developed and established with the assistance of the financial specialist recruited in the course of project preparation (para 4.03). Assurances were obtained at negotiations that the above accounts will be maintained in accordance with sound and generally accepted accounting principles and practices acceptable to IDA. 3.79 All the above mentioned project accounts and the Special Account (para 3.74) will be audited by independent auditors acceptable to IDA. The auditors' terms of reference will include a specific audit of SOEs. Audit reports will be submitted to IDA within six months after the close of the financial year and IDA will be allowed to review all project accounts and records. Assurances to this effect were obtained at negotiations.  39 J. Environmental Aspects. 3.80 Discharge water from coffee washing stations contains pollutants which, if allowed to flow into streams and rivers in sufficient concentrations, can cause serious polution downstream from the stations. Preliminary studies of the potential problem have been carried out and reviewed by IDA. The conclusions do not recommend that new stations be equipped at this time with water recirculation systems. These systems are generally not needee in view of present throughput and are expensive as they may represent an additional 10 to 15Z to the investment costs of the stations. The recommendations, however, propose that the design of the future stations provide for the installation of such facilities should they be needed in the future and proposes a stringent monitoring system to be maintained in order to assess the level of pollution and, if needed, determine where and when recirculation systems should be installed. These recommendations will be implemented under the project (para 3.49). IV. PROJECT IMPLEMENTATION A. Orfanization and Management. 4.01 Project implementation will be coordinated by OCIBUIOffice du Cafe. The Office will have direct responsibility for implementing the Institutional Restructuring and Training components of the project. Public sector washing station construction will be implemented by a special Implementation Unit to be created within the OCIBU/Office du Cafe for the duration of the project. This unit will also provide technical assistance to PFIs and to private investors in the preparation of their washing station construction projects. ISABU will be responsible for implementing the research component of the project. B. Implementation of Specific Components. Institutional Reforms. .02 OCIBU/ Office du Cafe. Calendar 1990 will be a transition period during which OCIBU will prepare the transfer of its present industrial activities; the scheduled transfer date will be December 31, 1990. In the meantime it will also prepare to assume the wider ranging regulatory and supervisory responsibilities of the Office du Cafe, which will start operating as a separate entity as of January 1, 1991. To fulfill this double mandate, OCIBU will, for a limited period, cumulate both its current responsibilities as well as those of the future Office. This transitional status has been confirmed by a directive from the Minister of Agriculture. The directive instructs OCIBU to take the necessary steps to transfer some of its current activities to SODECO, to the Ministry of Agriculture and to the Mumirwa SOGESTAL. The ownership of the assets (the hulling factories, the washing stations and the central  40 workshop) and their related financial obligations will be transfered to the Assets Management Facility. 4.03 The tasks to be accomnlished during the transition period includet (a) a reorganisation of the accounting systems of OCIBU and of the RDCs; (b) a full inventory and evaluation of the physical assets of OCIBU and the RDCs whose operations are to be transfered to SODECO and the SOGESTALs, while the assets accounts and related documentation are to be trausfered to the Assets Hanagement Facility; (c) the closing of accounts and related statements of financial position to effect the transfer of activities and assets at the date of transfer (at latest, December 31, 1990); and (d) the design and establishment of financial and cost accounting systems for the new entities (Office du Cafe, SODECO, SOGESTALs and the Assets Management Facility). A financial specialist has been recruited to coordinate these tasks. He will be assisted by the financial staff and advisors of OCIBU and the RDCs; the existing position of financial advisor at OCIBU, financed by CCCE, will be maintained during 1990 as will those of the RDC financial advisors, financed under the ASSP. Additional assistance is scheduled to be recruited during the year to assist in carrying out the inventory and valuation of the assets. Audit requirements were discussed at para 3.77. 4.04 A legal specialist will be recruited during 1990 to assist in the preparation of the legal documents and in fulfilling other legal requirements leading to the establishment of the Office, the SODECO, the SOGESTALs and the Assets Management Facility. An important function of the legal specialist will be to formulate opinions and recommendations regarding the legal status under Burundi law of the entities to be established. His terms of reference will include the transfer of operating responsibilities from OCIBU and the RDCs to SODECO and the SOGESTALs as well as the transfer of assets to the Assets Management Facility. He will work in close liaison with the earlier mentioned financial specialist in formulating the leasing arrangements whereby the Assets Management Facility will lease the hulling factories and other facilities to SODECO and the washing stations to the SOGESTALs or interested private operators. He will also elaborate the modalities ana procedures by which the Assets Management Facility will be empowered to sell the assets under its trust to private investors. The terms of reference of the legal specialist will be discussed and finalized during the negotiations of the IDA Credit. 4.05 The decree which governs OCIBU will be rescinded at the time the latter will transfer its industrial activities to SODECO, SOGESTALs and the Assets Management Facility. This transfer will be effected at the latest by December 31, 1990. The decree formally constituting the Office du Cafe will be promulgated at the same time. The draft of the decree constituting the new Office will be reviewed by IDA prior to its final approval by Government. 4.06 SOGESTALs. The accounting, inventory and legal work leading to the establishment of SOGESTALs will be included as part of the assignment of the financial and legal specialists (paras 4.03 and 4.04). The work will inelude earmarking of RDC and OCIBU operating assets (other than washing stations and their related equipment) to be transfered to the SOGESTALs. Also included will be the modalities by which these assets will  41 be transferred to the new entities. The legal specialist will also have the responsibility for formulating the arrangements and modalities by which RDC and OCIBU staff will be transferred to the SOGESTALs. 4.07 To conform with Burundi law, the legal status of the SOGESvALs and SODECO will have to evolve to allow for the expected increase of private investor participation. In the initial stages, when more than 902 of their share capital will be owned by the Government, the companies will be registered as 'Soci;t;s de droit public". As the Government's ownership decreases, between 90Z and 10%, the companies will be registered as "Societes d'economie mixte de droit priv;". Finally, when Governmient participation will represent less than 102 of the companies' capital, the companies will be registered as 'Soci;t;s par action de droit priv;". The legal specialist of the project will need to review these changes and formulate his own recommendatiorts as well as proposals regarding the modalities by which capital shares will be made accessible to private shareholders. These modalities will include the procedures by which shares will be valuated (in the abs ance of an open market), bought and sold, and how their owners will be repiesented on the companies' Boards of Directors. 4.08 The decrees constituting the SOGESTALs will be promulgated in 1990. As in the case of the Office du Cafe, their drafts will be reviewed by IDA prior to their final approval by the Government. Ai.09 A specialist in management and finance will be recruited by the project to assist the SOGESTALs during the firs. two years of operations. Assurances were obtained at negotiations that this specialist be appointed by uecember 31, 1990. 4.10 SODECO. Much of the pr paratory financial and legal work to be done in the case of the Office d' Cafe and the SOGESTALs by the financial and legal specialists will also -pply to SODECO. The decree constituting the SODECO will be promulgated in 1990. Its final draft will be reviewed by IDA prior to Government approval 34/ . 4.11 Technical assistance to SODECO will include a management and finance specialist and a technical specialist for operating the sorting equipment. Both are financed by CCCE, are currently in post and will continue working with SODECO following its establishment. 4.12 Washing Stations. OCIBU's Implementation Unit will be responsible for the project's washing station component. The unit will be headed by the Coffee Technologist and support will be provided by a consultant civil engineer to supervise the implementation of the construction program. A basic design, including plans and estimated costs, of a typical washing station were prepared in the course of project preparation drawing from years of experience in the construction of such stations in Burundi. Further refinements were implemented at appraisal 34/ Assurances with regard to IDA's review of the statutes of the Office du Cafe, SOGESTALs and SODECO and of their promulgation by end 1990 have been included in a side letter on coffee strategy which was obtained from Government at negotiations.  42 (See Technical Anwexes for details). Work on station design aiming at Improving efficiency and reducing the construction costs of new stations and the operating costs of existing and future stations, will continue during the project under the direction of ISABU'e Processing Technology Division which will be created under the project. C. Work Programs. Monitoring and Mid-Term Review. 4.13 The economics department of the Office du Cafe (para 3.10) will be responsible for the preparation of: (a) annual progress reports; (b) annual work programs; and (c) a project completion report. Progress reports and work programs will be submitted for the approval of the Government and the financiers by September 30 of each year. Agreements on issuance of progress reports and the preparation of the project Completion Report were obtained at negotiations. 4.14 A joint review of all project activities will be carried out in 1992 by the Government and the project's financiers, giving particular attention to the implementation of sector reforms. A joint agreement on the recommendations of the review and a commitment by the Government to implement them will be obtained by December 31, 1992 (para 3.27). V. FINANCIAL ANALYSIS. 5.01 The analysis of the coffee industry was carried out in constant 1989 FBu on the basis oft the Bank's international coffee price projections which assume that the quotas and price levels of the International Coffee Agreement will not be resumed in the near future; Burundi's expected coffee production; coffee prices presently paid to farmers and to traders; estimated operating costs of the industry's coffee marketing and processing institutions; and projected annual costs of the Office du Cafe, coffee research, the maintenance of access roads, and inputs now being provided free by the Government. Cost of all investment capital has been included at an annual rate of 10.52, which is in line with the present cost of long- term funds. Cost of short term borrowings has been estimated at 9.5% per annum. An exchange rate of FBu 173 per US$ 1.00 has been used in the calculations. 5.02 The analysis indicates that the industry will incur financial losses of about FBu 2.6 billion (US$ 15.0 million) and FBu 0.9 billion (US$ 5.2 million) in 1990 and 1991, respectively, and then make positive financial contributions, increasing gradually from about FBu 0.9 billion (US$ 5.2 million) in 1992 to FBu 7.3 billion (US$ 42.2 million) by year 2000. The projected financial improvement is due to: (i) the expected increases in world market prices; (ii) the expansion of Burundi's coffee production; (iiI) the growing share of fully washed coffee; (iv) improvements in the efficiency of the coffee industry; and (v) the increase in the quality premium forecast to be earr.ed by Burundi fully washed coffee as a result of quality improvements expected to be achieved under the project.  43 5.03 While the analysis shows financial losses accruing to the industry in 1990 and 1991, coffee revenues exceed the cost of production, processing and marketing even at present low world market prices. The main reason for the projected overall lose of the industry stems from the administered producer price which is considerably above the on-farm cost of production. Present producer prices yield 231 and 211 financial rates of return depending on whether farmers market cherry coffee to washing stationls or sell semi-washed parchment coffee to traders. Farmers' cost of production, is estimated at 70X and 75% of the price now paid for cherry coffee and semi-washed parchment coffee, respectively. At lower producer prices, the industry would thus be self financing even at current depressed world coffee prices. 5.04 Government's policy has been to rule out producer price reductions and to compensate for world market price declines by drawing from the stabilization fund and adjusting the exchange rate, thus reducing producer prices in real terms. This policy is expected to continue to govern Government's decisious (para 3.33). VI. JUSTIFICATION, BENEFITS AND RISKS A. Justification. 6.01 Coffee has been and will remain for the next decade the single most important economic activity in Burundi, and one which in most years has been consistently profitable to its participants. The sector reforms which the project aims to introduce, i.e. private sector participation, increased competition and increased quality of the prodact, are therefore expected to have far-reaching implications on the efficiency and profit potential of the industry, and enable it to even better overcome the swings in world market prices. As such, it also provides a crucial sector specific complement to the structural adjustment and macro-economic reforms the Government has embarked upon with IDA support since 1986. B. Benefits. 6.02 The project will generate quantifiable benefits from three differents sources. The first will be the incremental value of coffee sales as a result of the shift from washed to fully washed coffee. The premium for Burundi fully washed coffee over the ICO indicator price for other mild arabica ex-dock Neow York, has been conservatively estimated to progress from '1 percent to 15 percent by PY5. This percentage applies to a base price which is the lowest observed over a 15 years period. The second source of benefits results from avoiding a likely discount on the price of semi- washed coffee. In a 'Without Project' scenario, total production will be 50,000 tons of clean coffee by 1995 of which perhaps 10,000 might be fully washed from existing washing stations. The 40,000 tons of semi-washed would increase problems associated with inadequate quality control and export quali-:y would fall as a restlt. This process of deterioration is well advanced in neighboring countries producing only  44 semi-washed coffee. Larger quantities of mediocre quality result in lower prices, specially in a free market. As a result it is assumed that in a "Without Project' situation the value of Burundi washed will gradually fall below the ICO other mild indicator and by 1995 the value would be at least 52 below. The justification for considering this scenario is the fact that additional plantings necessary to produce these quantities, are already in place. The "Without Project" situation will be worsened if export quotas were to be re';ntroduced. In this case Burundi will have to L sell substantial quantities of mainly semi-wasned coffee to non-quota markets and the average differential FOB between Fully Washed (which would mostly go to quotas markets) and semi-washed will likely be wider. The third benefit will be obtained from the increased efficiencies resulting from industry reforms and reflected through quantifiable improvements in the clean-to-parchment coffee ratio from 78 percent to 82 percent. 6.03 The potential impact of the following has not been quantifified: (i) reduction of farm labor (mostly from women) resulting from the suppression of hand pulping; (ii) resumption of ICA quotas which would result in a widening of the price differential between fully-washed and semi-washed. C. Economic Analysis. 6.04 It has been assumed for the economic analysis that: (a) the life of project investments would be 25 years with all the main investments taking place during the implementation period (the first five years); (b) all expenditures for capital investments, and incremental operating and maintenance costs relating to the project components have been included. Costs exclude price contingencies, taxes and duties but physical contingencies are included; (c) all investment costs are calculated on the basis of late 1989 constant prices taking into account the exchange adjustment of November 1989. The calculation of the base price for Burundi fully-washed arabicas is set out in Annex 3. Based on past sales it has been estimated that Burundi would earn in 1990 a premium of 11 percent over the ICO indicator. Due to quality improvements through the project the premium is anticipated to increase gradually to 15 percent. Projections of the ICO indicator prices are based on IBRD commodity price forecasts in a non-quota scenario; (d) the economic and financial benefits from quality improvement will occur for two classes of fully washed cofffe and two classes of semi-washed coffee (main grade, under grade). Throughput of the 100 additional washing stations is assumed to reach 105 percent of nominal capacity, as indicated by experience with the 74 washing stations currently in operation; and (e) a shadow exchange rate applies on the basis of a 10 percent foreign exchange premium to the value of all tradables. 6.05 The base case for the project has been estimated on the assumption that the ICA will continue without quotas. Benefits and costs streams used in the economic analysis are presented in Annex 4. On the basis of the above assumptions, the economic rate of return (EER) of the whole Project is 22 percent, with a net present value of about US$19 million. 6.06 Sensitivitv Analysis. Long term reduction in projected prices of 25 percent arising from permanent disruption of the market structure and/or  45 from the inability of the coffee industry to improve its marketing strategy would lead to an ERR of 15 Z. Cost escalations of 25 X accross all investments and incremental operating costs would reduce the ERR to 15 percent. 6.07 Switching values have also been estimated. In order to reduce the present value of the Project to zero at a 12 percent discount rate coffee prices would have to be 32 percent lower than assumed in the base case analysis. This is unlikely to happen during the life v. the project because of Burundi's strong comparative avantage in producing high quality arabica coffee and the increasing market demand for this type of coffee. Cost overruns would have to reach about 45 percent to reduce the net present value to zero at the same discount rate. The economic worth of investment is thus robust to price and costs assumptions. A return to the ICA quotas would not significantly change the ERR of 22%. D. Risks. 6.08 One of the main domestic risks of the project will be delays by Government in ensuring the effective implementation of the proposed reforms. However, the Government has show6x willingness to work on the basia of clear objectives and orderly steps to achieve them, and, as described in the report, a considerable degree of Government agreement to the major sector reforms was already obtained during project preparation, so that most of the reformt can be implemented early in the project. 6.09 Another element of internal risk concerns the skepticism on the part of many in the industry as to when private investors will begin to show interest. The general view is that the process will be gradual and will depend to a large extent on Government commitment, the success of the new payment system and the ability of the new entities and the washing stations to demonstrate their viability. It is not envisaged that private investment will start flowing in the industry until about late 1991 after the reforms have been implemented and have gone through a running-in period. The review in 1992 will provide a timely opportunity to assess the situation and amend the process as necessary. 6.10 A last element of risk will be the interest shown by private exporters in participating in the coffee auctions. Start-up of the private auction facility is expected to be slow: some exporters could be reticent to participate initially, preferring to await and see' before committing themselves. An important factor in its success will be the reputation which it will acquire over time on the basis of its regularity and consistency and the willingness of the Burundi authorities to facilitate the process by removing any remaining export constraints. Once a good reputation of the Burundi auctions is established there should be no problem in attracting buyers and increasing the quantities marketed.  46 VII. ASSURANCES AND RECOMMENDATIONS. 7.01 Assurances Obtained at Negotiations: (a) Final decision on location of washing stations on the basis of technical, financial and economic criteria acceptable to IDA (para. 3.46). (b) Procurement of goods and services in accordance with IDA guidelines (para 3.68). (c) Agrsement, by means of a side letter, to conform to IDA guidelines in the matter of local firms under local competitive bidding conditions (para 3.68). (d) Government's quarterly instalments of counterpart funds to be deposited to project account (para 3.76). (e) the following condition to be implemented by October 31, 1990: (1) the joint review by Government and IDA of the consultants' recommendations on the new pricinglpayment system (para. 3.34). (f) the following conditions to be implemented by December 31, 1990: (1) the establishment of the private auction facility on the basis the agreement between the Government, IDA and CCCE (para 3.26). (2) appointment of the training specialist (para 3.58). (3) agreement between the Government and CCCE relative to the latter's financing of the project (para 3.63). (4) appointment of a management and finance specialist for the SOGESTALs (para 4.09). (g) the following condition to be implemented by February 28, 1991: (1) the establishment of the new pricingipayment system (para. 3.34) (h) the following condition to be implemented by June 30, 1991: (1) the establishment of the Private Sector Promotion Committee under terms of reference and work program acceptable to IDA (para 3.40). (i) the following condition to be implemented by December 31, 1992: (1) agreement on and comritment to implement the recommendations of the 1992 joint review (paras 3.27 and 4.14).  47 (j) Accounts to be maintained in accordance with sound and generally accepted accounting principles and practices acceptable to IDA (para 3.78). (k) Audit reports within six months of financial year (para 3.79). (1) Progress reports and Project Completion Report (para 4.13). 7.02 Conditions of Effectiveness of the Credit. (a) Review and comments by IDA of the recommendations of the Preparation Comnittee (para 3.26). (b) The implementation of the staff conditions of service and the appointment of the Marketing Manager in BCC on the basis of competitions (para 3.29). (c) The appointment of the consultants to undertake the pricinglpayment system study (para 3.34). (d) The appointment of the research coordinator (para 3.53). (e) Agreement between the Governments of Burundi and Belgium relative to the financing of the washing station access roads (para 3.63). (f) Concluwion of a Management Agreement between the Government and the BRB on the arrangements to on-lend IDA proceeds to private entrepreneurs (para 3.67). (g) An initial deposit by Government of FBu 35.0 million to the project account (para 3.76). (h) Audit of OCIBU's 1989 accounts (para 3.77). 7.03 With the indicated assurances and conditions, the proposed project would be suitable for a Credit of SDR 21.3 million (US$ 28.0 million equivalent) to the Government of Burundi on standard IDA terms with 40 years maturity.  -48- Annex 1 BURUFDI PROJET FILIERE CAPE STATIONS DE LAVAGE PROGRAMME DE CONSTRUCTION Stations de lavage supplementaires Total St. L. a la SOGESTAL exist. Financement Financement Projet Cafe fin du oxistant projet 1989 1990 1990 1991 1992 1993 1994 Total NGOZI 21 2 2 2 5 4 5 3 20 45 KAYANZA 10 2 - 7 6 6 6 5 30 42 MUYINGA 6 3 - 5 4 3 - 2 14 23 KIRIMIRO 29 2 1 - - 3 5 5 13 45 MUMIRWA 8 5 5 - 5 6 6 6 23 41 TOTAL 74 14 8 14 20 22 22 22 100 196 IDA 14 15 16 16 19 80 CCCE 0 5 6 6 3 20  - 49 - BURUNDI Annex 2 COFFEE SECTOR PROJECT Estimated Schedule of Disbursements of IDA Credit (USS Million) IDA Project Quarter At the end Disbursement Fiscal Yr Year Endini of Ouarter Cumulative 1991 1 Jun. 90 0.0 0.0 Sept. 90 0.7 1/ 0.7 Dec. 90 0.2 0.9 2 Mar. 91 0.2 1.1 Jun. 91 1.0 2.1 1992 Sept. 91 1.0 3.1 Dec. 91 1.1 4.2 3 Mar. 92 1.1 5.3 Jun. 92 1.4 6.7 1993 Sept. 92 1.4 8.1 Dec. 92 1.5 9.6 4 Mar. 93 1.6 11.2 Jun. 93 1.8 13.0 1994 Sept. 93 1.8 14.8 Dec. 93 1.4 16.2 5 Mar. 94 1.4 17.6 Jun. 94 1.4 19.0 1995 Sept. 94 1.4 20.4 Dec. 94 1.3 21.7 6 Mar. 95 1.3 23.0 Jun. 95 1.1 24.1 1996 Sept. 95 1.1 25.2 Dec. 95 0.4 25.6 7 Mar. 96 0.4 26.0 Jun. 96 (,.6 26.6 1997 Sept. 96 0.6 27.2 Dec. 96 0.4 27.6 8 Mar. 97 0.4 28.0 IIPPF Refund  - 50 - Annex 3, Table 1 (1) COFFEE PROJECT Page I of 3 Projected Financial Contribution from Coffee in Burundi - Assusing that no International Coffee Agreement Exists - (in con;tant 1989 F8u'nillionl 1990 II 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001ff KEY PARAMETERS ---- ---- ---- --- - ---- ---- ---- ---- ---- ---- ------ Projected Exports (tons of clean coffee) (iW fully washed - main grade 28,800 15,420 18,170 21,100 24,250 27,500 29,120 29,950 30,290 30,290 30,290 30,.90 - under grade 3,200 1,710 2,020 2,340 2,690 3,050 3,240 3,330 3,370 3,370 3,370 3,370 Qii) washed - nain grade 7,200 26,880 25,930 23,900 20,750 17,510 15,880 15,050 14,710 14,710 14,710 14,710 - under grade 800 2,990 2,880 2,660 2,310 1,940 1,760 1,670 1,630 1,630 1,630 1,630 Total Exports 40,000 47,000 49,000 30,000 50,000 50,000 50,000 50,000 50,000 50,000 50,000 50,000 Proj. Prices (F8u000/ton f.o.b. Dar) 1/ lil fully washed - main grade 329.6 371.0 413.1 436.5 479.8 499.4 538.5 538.5 53C.5 558.1 577.7 577.7 - under grade 257.8 290.9 324.6 343.2 377.9 393.5 424.9 424.9 424.9 440.6 456.2 456.2 (ii) washed - eain grade 294.0 328.1 362.2 379.2 413.3 430.3 464.4 464.4 464.4 481.5 498.5 498.5 - under grade 229.3 256.5 283.8 297.4 324.7 339.3 365.6 365.6 365.6 379.2 192.8 392.8 Average Export Price 315.5 336.3 374.9 397.3 439.5 462.5 501.5 502.9 503.4 521.8 540.2 540.2 REVENUES 12,618 15,805 18,371 19,867 21,976 23,124 25,077 25,144 25,172 26,091 27,010 27,010 EXPENDITURES 8CC Operations 2/ Overhead Expenditures 3/ 45 45 45 45 45 45 45 45 45 45 45 45 Cost of Port Storage in Buiuebura 4/ 24 28 30 30 30 30 30 30 30 30 30 30 Transport Cost - Bujusbura - Kiga lincl. loading) 39 45 47 48 48 48 48 4 48 48 48 48 - Kigosa -forwarding fees USSI13.6/too) 94 111 115 119 lie 118 118 118 118 11 9 11 9 11 9 - Kigoma - Dar (payable in US$ 51 299 350 365 372 372 372 372 372 372 372 372 372 Oar -forward, fee lpa;able in US$) 6/ 125 147 153 156 156 156 156 156 156 156 156 156 Insurance Ipayable in US$) 7/ 30 37 44 47 52 55 59 60 60 62 64 64 Dank Comaission (0.552 of coffee valus) 69 87 101 109 121 127 138 138 138 144 149 149 Interest Expenditures - during storage at Bujumbura 8/ 342 390 406 419 420 422 423 423 424 424 424 424 - during transport to Dar 91 62 71 74 76 76 76 77 77 77 77 77 77 - during period until paym. r*cipt 10/ 100 125 145 157 174 193 I 199 199 207 214 214 Transport Losses (0.52) 11/ 63 79 92 99 110 116 125 126 126 130 135 135 Total 8CC Expenditures 1,290 1,514 1,618 1,675 1,721 1,748 1,790 1,792 1,793 1,812 1,831 1,831 1/ Based on prices in Table on Coffee Price ProjeKtione, and an exchange rate of Fla 173.0 per USf1.00. 2/ Projections presented here do not yet take into account the effect which the establishmnt of the planned parallel wketing channel will have on 8CC, and accordingly will have to be amnded in line with exports to be auctioned through the parallel system and BBC's adjustment efforts. 3/ Estimated on the basis of Actual expeneitures in 1969. 41 Calculated on the basis of an average storage charge of FBa5.75 per day and ton, and an average storage peried of 3.5 aonths. 5/ Cost are based on rail and road charges of US$40 and US$70 per ton, rspetively, and asmning that 90.0 2 of transport would be on rail. 6/ Based on fee of US$18.2 or US$16.5 per ton, depending on arrival by rail or road; 90.0 %of transport assumed to be by rail. 71 Cost are calculated at 0.227 and 0.39257 of the value of the differeat coffee qualities, depending on transport by rail or road, respectively. 8/ Estimated on the basis of past experience at 3.5 mnths; interest rates 9.5 1 P.a. 9/ Estimated at 20 days and 10 days for rail and road transport, respectively; internt rate 9.5 1 p.a., I 90.0 % rail transport. 10/ Estieated at one eonth, based on recrnt experience; interest at 9.5 2 pa. II/ Based on recent experience.  - 51 - Annex 3, Table 1 (2) Projected Fjnancial Cintribution froe Cuff. in krundi Palo 2 of 3 - Asiuaing that no International Coff e tnet Esists - (in constant 1"9 FEu'.ilIlinl 190 1791 M 1993 1994 1995 199 19 199 1M 2000 2001ff EXPENDITURES -.-- ---- ---- --- - ---- --- ---- _ _ _ _ SODECO Operations 12/ 1. Fixed Expenditures Permanent Staff 86 96 86 8t 86 86 86 84 86 96 806 6 Hired Staff I I 1 1 1 I I I I I I I Social Benefits 10 10 10 10 10 10 10 10 10 10 10 10 Total Personnel Cost 6 9 96 % 96 96 9 96 % 96 96 96 Repair and Raintenance 20 20 20 20 20 20 20 20 20 20 20 20 Supplies and Naterials 11 It 11 it It 11 It It It It 11 11 - tc I 5 5 5 5 5 3 5 Coffee Prosotion 20 20 20 20 20 20 20 20 20 20 2t 20 Insurance, Taxes and fees 22 22 22 22 22 22 22 22 22 22 22 22 Total Fixed Expenditures 174 174 174 174 174 174 174 174 174 174 174 174 2. Yariable Expenditures Labor 55 65 68 69 69 6 69 69 69 6 69 69 Packaging Naterials 86 101 106 too to8 108 106 108 106 1too10o 108 Repair and faintenance 46 55 57 58 58 58 59 58 59 58 58 58 Other Variable Expenditures 96 101 105 107 107 107 107 107 107 107 107 107 Short tern Interest 131 - (i) fully eashed coffee 143 77 90 105 121 137 145 149 151 151 151 ,51 - (iil seni cashed coffee 66 249 239 220 191 161 146 139 135 135 135 135 Total Variable Expenditure 483 646 64 67 654 640 433 630 629 629 628 628 3. Cost of Capital and Depreciation Cost of Long-tere Capital 141 420 575 575 575 575 575 575 575 575 575 575 575 Depreciation IS/ 247 345 34 345 345 345 345 345 345 345 345 f45 Totai Capital Cost and Depreciation 67 920 920 920 920 920 920 920 920 920 920 920 Total SUDECO Cost 1,324 1,739 1,758 1,761 1,747 1,734 1,m 1,723 1,M 1,M 1,722 1,722 12; Details are in Table on Projected SDDECD OWrating Cost. 13/ Based on an annu;l interest rate of 9.5 1, and estimted hort tero funding nes to finance pacmnt purcasxn and other operating expenditures for 2.0 mnsths if parchmns. is procred fr SOUSIMA and for 4.0 mnths if SODECO procures from traders, to shan it would normaly advance funds for their prchaing operations. 141 Based on present long tern capital need of FBu4billion which vill in 11 Increase to aboet F85.47biliion witb the expansion of the SITEGA hulling plant. The cost of capital has bee estinted at 10.5 2 pe anue. 15/ Based on present depreciation charges and the increase Pected fre the expansion of the SE611 hulling factory.  - 52 - Annex 3, Table 1 (3) Projected Financial Contribution fras Coffn in lurundi Page 3 of 3 - Assusing that no International Coffee Agrnleet Euists - (in constant 1989 F8u'nillion\ EIPEI3DITLIRES 1S90 1 1992 193 39S4 19S5 1996 19'7 1998 199 2000 2001ff SOSESTAL Operations 1. SOBESTAL Administration 16/ Petsonnei Cost 33 35 57 38 38 38 38 38 38 38 38 38 Leuacng of Otfice, Store, etc. 4 4 4 4 4 4 4 4 * * 4 4 Repair Ad Iaintenance it 12 12 13 13 13 13 13 13 13 13 13 Vehicle Operations 26 27 28 29 29 29 29 29 29 29 29 29 Other Ependitures 5 5 5 5 5 5 5 5 5 5 5 5 Total SGEBSTAL Adoiuistration 79 93 86 90 90 90 90 90 90 90 90 90 2. Vashing Station Operations 16/ Fixed Cost - Lease 17/ 300 399 484 569 6t3 161 761 743 761 761 761 761 -Personnel 120 151 1i7 205 235 264 2U6 264 266 266 266 2t6 - Repair, haintenance and llisc. 9 s 118 139 160 194 209 208 209 208 2 28 209 208 Yariable Cost - Procurnt of Cherry Coffee 181 3,123 3,920 4,501 5,224 6,001 6,802 7,202 7,406 7,497 7,407 7,497 7,487 -Labor 113 138 162 199 237 246 260 268 271 271 271 27i -Fuel and Lubricants 13 16 to 21 25 28 30 30 31 31 31 31 -gqs and rying Naterials 32 40 47 54 62 71 75 77 78 79 7o 75 - lisellan eas Operating Cost 30 13 15 17 20 22 24 24 25 25 25 25 - Tranport 191 71 86 102 I1l 136 154 163 I68 170 170 170 170 -Insurance (0.12 of vael 4 4 5 6 7 a 9 9 9 S 9 9 - Short Twr Interest 201 49 60 71 83 95 108 114 117 119 19 119 119 Total Washing Station Operations 3,928 4,845 5,722 6,648 7,644 8,674 9,112 9,335 9,423 9,423 9,423 9,423 3. Cost of Capital and preciation nt of Long-twreCapital 211 78 78 78 78 79 79 79 79 79 79 79 79 Deprciation 37 37 37 37 2 37 37 37 37 2 37 37 Total Capital Cost I Depreciation 115 1 I5 115 11 9115 3 1 1 90 135 115 Total SOSESTAL Cost 4,122 5,042 5,923 6,852 7,813 9,878 9,317 9,539 9,620 9,593 9,628 9,628 Procur. of SGeig-ad Parbmat Co. 221 1,904 7,108 6,056 6,321 5,498 4,629 4,198 3,979 3,989 38589 3,889 3,889 Bross Contribution of Burundi Coffee 3,979 401 2,215 3,259 5,207 6,136 8,041 8,111 9,141 9,076 9,941 9,941 Operating Cot of the Office de Cafe 72 72 72 n 72 72 72 72 72 72 72 72 Cost of Coffee Reuarc 112 112 312 332 1 12 U 3 112 312 332 332 112 112 Naintence and epir of Coffee ads 13 17 20 23 26 29 29 29 29 29 29 29 Coffee Inputs Provided Free by Bovrat. 4S0 490 490 490 490 49 4 49 4 490 490 490 Met Contribution of Eurundi Coffee 3,291 (2901 1,522 2,562 4,507 5,432 7,343 7,408 7,43B 9,373 9,238 9,238 16/ Details are in Tables on SOSESTAL leadqurtr Ct, and lashing Station Operating Cost. 17/ Lea* cost for an uanhing stations have bn calculated basd an construction cost etimate lincl interet during constra. and land cost), a 30 year atrtzation period, and intenrt & ainl. cost of 10.5 t pa. Far nisting stations, a 202 Ioer leas cost has be assud. 19/ Basd on cherry coffe procuren. price of Flu 36.0 per kg and an average outturn ratio as indicated hov. 19/ Based on transport cost of Flu 19 p km and ton ad a average transWort distance linclud. return tripl of 230 kb. 20/ Basd on teao mths borroing needs to finance the prosur;at of cherry coffn frau fareers, and an interest rate of 9.5 2 per annie 21/ Based on long term capital need In the aount of capital ependitures and permannt eorkiag capital rfquiresnts estimated at nine months operating costs, excl. espenditures for cherry coffe prorent ad leasing cost. The cost of capital has been estimated at 10.5 I p.c. 22/ Rased on parchunt procuremet price of F8u 181 per kg, paid by SODECO to traders.  - 53 - Annex 3, Table 2 _m PRWEC1 Projection of Coffel lpert Prices 11 (Mitbout in International Coffen grenont) (Price in constant 19I9 UIS p tan of clean coffee) Im 1991 IM 199 4 1995 99 IW7 19 9 2000 2001ff Projeted IC0 Indicator Price lotherUili arabicas, n-dock IN York) 1,900.0 2,100.0 2,300.0 2,400.0 2,600.0 2,700.0 2,900.0 2,S00.0 2,900.0 3,000.0 3,100.0 3,100.0 Avrage Price Preitue for Baruned Cffn i fully washed coffn 2/ - gaia grade 209.0 252.0 m.o 336.0 390.0 405.0 435.0 435.0 435.0 450.0 465.0 465.0 - uandr grade 1212.9) (218.41 (220.9) 1211.2) 1208.0) (216.0) (232.0) (232.01 1232.0) (240.0) (248.0) (248.0) (ii) ss%i ashed coffee 31 - in grade 0.0 0.0 0.E 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 - unde grade 1390.0) (420.01 (460.01 t0) (1520.0) (540.01 (5100) (580.0) 1580.0) (O. (420.0) (620.0) TransWrt, Insufance, Pandling If.o.b. Dar es Salue to x-dok Ibn Yorki fi) Transport of ccAtaineri2d c-ffne froe Oar to Nou York 41 125.0 125.0 125.0 125.0 125.0 125.0 125.0 125.0 125.0 125.0 125.0 125.0 (ii) Lading cbarges m York 47,0 47.0 47.0 47.0 47.0 47.0 47.0 47.0 47.0 47.0 47.0 47.6 (iii) Ad valorem chars incl. Inw- rence (0.5) and interest 112): 1.51 of value of coffee - fll'? uashed coffn - gin grad 31.6 35.3 39.0 41.0 44.9 46.6 50.0 50.0 50.0 51.8 53.5 53.5 - under grade 25.3 21.2 31.2 32.9, 35.9 37.3 40.0 40.0 40.0 41.4 42.9 42.8 - sci mabed coffn - eal trade 2B.5 31.5 34.5 36.0 39.0 40.5 43.5 43.5 43.5 45.0 46.5 46.5 - under grade 22.9 25.2 7.6 21.9 31.2 32.4 34.8 34.8 34.0 36.0 37.2 37.2 Projected Coffee Etpot Price - --------------- --- ----------- -. (f.o.b. Oar a alas quvalent) (a) fully Ra coffne - min grade 1,905.4 2,144.7 2,3U.0 2,523.0 2m.2 2,89.4 3,213.0 3,123.0 3,113.0 3,226.3 3,339.5 3,339.5 - under grade 1,489.9 1,611.4 1,176.0 1,9.0 2,184.1 2,274.7 2,456.0 2,456.0 2,456.0 2,546.6 2,637.2 2,637.2 fii) cii mh coffee - uian grade 1,699.5 1,96.5 2,09.5 2,192.0 2,99.0 2,487.5 2,4.5 2,684.5 2,684.5 2,793.0 2,81.5 2,1.5 - undr grade 1,325.2 1,482.9 1,60.4 1,719.2 1,876.8 1,955.6 2,113.2 2,113.2 2,113.2 2,192.0 2,270.8 2,270.8 11 Projectins are bas n l1D coedity price forcattb. 21 Bad an put sln, it hs n sntiuated that endi d earn in 190 fir its sain grad fully washed coffN a prosius of 11.0 2. We to quality iwroits the preii is nticipated to iwreae graully to 15.0 t in 1995. Under grades ll at 807 of the gin grade prices. 3! Baeud n past sales, it has km etitiod that hiuedi sJ mce coffe (gain grade) mod contine to ell at a price sisilar to the IC0 Indititor Price. Under grade mo ull at a 202 discout. 41 Calculated on thr basis of a 20 foot containr holding 16.8 tans of eln coffee.  - 54 - Annex 3, Table 3 BURUNDI COFFEE PROJECT REVENUES AND COST OF PRODUCTION PER TON OF COFFEE (F8u) FULLY MASHED 1/ SENlI WASHED I/ Revenues 322,440 207,538 Cost 2/ N Marketing 32,541 31,120 - Hulling 38,659 42,467 - Mashing 65,939 0 - Herchants Intermediation 0 16,456 - Office du Cafe 1,800 1,C00 -Coffee Research 2,800 2,800 -Repair & haint. of Coffee Roads 735 0 -Governeent Coffee Input Srants 10,875 17,750 Production 154,268 165,949 Total Cost 307,517 278,342 Net Revenue 14,923 9,196 1/ The calculation takes into account that both the production of fully washed and sedi washed roffee, yields about lOZ unoer grade quality coffee. Revenues and those cost which vary vith the export value of coffee, have been calculated based on a projected Burundi coffee export price derived from an assumed ICO Indicator Price of USS 1,900 per ton, and an estimated price premium for Burundi coffee of 11.0 Z, and of 0.0 Z for the main grades of fully washed and semi washed coffee, respectively. 2/ Estimated based on an annual production of 32,000 tons and 8,000 tons of fully washed coffee and of semi sashed coffe, respectively, and coffee exports in the like amounts. Interest expenditures for sesonal credit have been based on a 9.5 Z p.a. rate of interest. Capital costs have been estimated at 10.5 2 per annum.  - 55 - Annex 3 - Table 6 -~~~~~~. . UN__ .UFFEE