Sri Lanka - Second Dairy Development Project

Document of The World Bank FOR OFmFCIAL USE ONLY 2 -CC Report No. 5089-CE STAFF APPRAISAL REPORT SRI LANKA DAIRY DEVELOPMENT II May 13, 1985 South Asia Projects General Agriculture Division fThis document has a restricted distribution and Imav be use by reeipients only in the performance of | their oflicisl duties. Its contents may not otherwise be disclosed without World Bank authorization. |  CURRENCY EQUIVAIENTS US$1 = SLRs 25 SLRs 1 = US$0.04 ACRONYMS AND ABBREVIATIONS ADB - Asian Development Bank AI - Artificial Insemination AIIUL - Anand Milk Union Ltd. ASC - Agrarian Service Centers BC - Bank of Ceylon CB - Central Bank CRCS - Comprehensive Rural Credit Scheme CWE - Cooperative Wholesale Establishement DAPE - Department of Animal Production and Health DDF - Dairy Development Foundation DPA - Dairy Producers' Associations FAO - Food and Agriculture Organization PRR - Financial Rate of Return GOSL - Government of Sri Lanka GDP - Gross Domestic Product ICB - International Competitive Bidding LCB - Local Competitive Bidding LPD (lpd) - Liters per Day gpd - Grams per Day MILCO - Milk Industries of Lanka Company MPC - Milkshed Producers' Companies MRID - Ministry of Rural Industrial Development NLDB - National Livestock Development Board NMB - National Milk Board PB - People's Bank PPF - Project Preparation Facility pa - Per Annum pl - Per Liter RCAB - Rural Credit Advisory Board RDP - District Rural Development Projects SLRs - Sri Lankan Rupees SOE - Statement of Expenditure UHT - Ultra High Temperature VMPA - Village Milk Producers' Associations VMPC - Village Milk Producers' Companies WFP - World Food Program WEIGHTS AND MEASURES 1 millimeters (mm) = 0.03937 inches 1 hectare (ha) = (10,000 m2) = 2.471 acres 1 kilogram (kg) = 2.205 pounds 1 kilometer (km) = 0.6214 miles 1 liter (l) = 1.057 US liquid quarts 1 meter (m) = 3.281 feet 1 square kilometer (sq km) = 0.386 square miles GLOSSARY maha - Northeast monsoon season (October to February) poonac - residue from coconut oil extraction tank - impounding reservoir for irrigation Technical - artificial insemination and veterinary services services provided to milk producers yala - Southeast monsoon season (April to September) FISCAL YEAR January 1 to December 31  FOR OMCIAL USE ONLY STAFF APPRAISAL REPORT SRI LANKA DAIRY DEVELOPMENT II Table of Contents Pate No. I. BACKGROUND .................................................... 1 A. Introduction . . ................................1.... B. Economic Setting . ................................ ........ 1 C. The Role of Agriculture in the Economy ................... 2 D. The Role of Livestock and Dairying in the Economy ........ 3 The Livestock and Dairy Sector ........................ 3 Milk Production and Processing ..... ................... 3 Milk Imports ....... ................................. 3 Consumption and Demand ................................ 4 Milk Prices, Subsidies, and Tariffs .................... 4 E. Institutional Support to Livestock ....................... 5 Responsibility for Development ......... ............... 5 Livestock Research ...... .............................. 6 Livestock Extension .................... ............... 6 Credit ................. 6 Dairy Cooperatives .................................... 7 F. Livestock Production Systems ............................ 8 The "Estate" System .... ............................... 8 The Mid Country Smallholder System .. .................. 9 The Coconut Triangle System .... ....................... 9 The Dry Zone System ................................... 9 The Urban Jaffua System .. .............. 9 The Smallholder Buffalo System .. ...................... 10 G. Development Activities and Project Rationale . ............ 10 District Rural Development Projects ................... 10 Bilateral Livestock Development Projects . .............. 10 ADB-Assisted Projects .... ............................. 10 Bank Involvement in the Dairy Sub-Sector and Project Rationale .................................. 11 This report is based on the findings of an appraisal mission to Sri Lanka in November 1983. Mission members included Messrs. Knudsen and Hardison, Ms. Estoque (Bank), Messrs. Halse, Madhavan, Punjrath and Wijesekera (Consultants). The mission had preliminary discussions on dairy commodity assistance with USAID and EEC representatives in Colombo. Ms. Ai-Chin Wee and Mr. Brian Johnson (Bank) gave valuable assistance in the preparation of the economic and financial analyses. IThis docu"nnt hasa restrced ditbution and may be usd by epients only in the perfonnmance of| thir ofrida duties. Its contents may not otherwise be dicosed without World Dank autboization.  Table of Contents (continuation) PaRe No. 1I. THE PROJECT ................................................. 12 A. Project Area ............................................. 12 B. Project Design ............* . ........................ 13 C. Objectives ........................................................ 15 D. Project Components ....................................... 15 Organization and Management ... ...... ............. 15 Dairy Plants and Chilling Centers .................. ... 16 Technical and Extension Services ...................... 17 Training .............................................. 18 Technical Assistance ....................................... 18 Monitoring and Evaluation ..... ** ....................... 19 III. PROJECT COSTS AND FINANCING ...... . ....... 19 A. Project Costs .....19 B. Financing ....... ................ 22 On-Lending Terms and Conditions . . .. 23 C. Accounts and Audits .... 24 IV. PROJECT ORGANIZATION AND IMPLEMENTATION .... 24 A. Essential Features of Corporate Entities .. 24 General .................. 24 Dairy Development Foundation (Guarantee) Limited (DDF) .. .. .. .. ..... 25 GZeneral ..................... 2.5 Objectives .................. ..... 25 Income and Turnover Taxes and Cuatoms Duties 25 Profits and Dividends .......................25 Milk Industries of Lanka Company Limited (MILCO) ...... 25 Fu-nctions ....................25 Share Capital ..........*........ 26 Income and Turnover Taxes .... .. 26 Milkshed Producers' Companies Limited (MPC).. 26 Functions ................ 26 Share Capital . . ...................... 27 Distribution of Profits .............. 27 Income and Turnover Taxes . 27 Village Milk Producers' Companies (Private) Limited (VNPC) ......27 Functions ............... ..... 27 Share Capital ........ .......... 27 Income and Turnover Taxes . ........... 28 B. Inter-Ministerial Coordination . ....... 29 Steering Committee ............................... 29 C. Procurement and Disbursement . ........... 30 Procurement ..... .. ................ * . 30 Disbursements .....................31 Project Special Account ........................ ... 32 D. Pre-Investment Activities ......... .. ...... ......... . 32 E. Monitoring and Evaluation ........................... 33  - iii - Table of Contents (continuation) PaRe No. V. MARETING. PRICES, TARIFFS AND SUISIDIES ...................... 33 A. arketic g .... . ............ e .... . ..... . 33 B. Prices .................................. ........... o ....... 33 C. Tariffs . . ........................... .................. 34 D. Subsidies .....o..... * * ...... o ..... ........ ..... ... 34 VI. FINANCIAL^ VIABILITY .................. o.................................. 35 A. Dairy Development Foundation .................... ...oo ... 35 B. Milk Industries of Lanka Company .................... ..... 37 C. Milkshed Producers' Companies ............................ 38 D. Village Milk Producers' Companies .. ........ . ....... 39 VII. PROJECT JUSTIFICATION ADD RISKS .............................. 40 A. Milk Production and Procurement .......................... 40 B. Project Viability and Economic Returns ................... 41 C. Other Impacts ......................... o................. .......... 0.0 43 D. Project Risks . .. ................ ....... .................. o . 44 VIII. RECOMMDATIONS ....... ..... ............. ........ ...... .. 45 mT TABLES 1.1 Distribution of Cattle and Buffaloes in Sri Lanka ............. .. 8 2.1 Technical Assistance . ......... ... ........................ ..... 19 3.1 Project Cost Summary ......... ................................. 21 3.2 Project Cost Summary by Expenditures ......................... 22 3.3 Financing Plan ...o..... ..... o.... ..........0...... ...................... 23 4.1 Procurement ...... -..............-....-............................... 30 6.1 Flow of Funds and Summary of Financial Statements for DDF ..... 36 6.2 Flow of Funds and Summary of Financial Statements for MILCO .... 37 6.3 Flow of Funds and Suary of Financial Statements for the MPC.. 39 6.4 Summary Income Statement for the VMPC .......................... 40 7.1 VMPC aud Incemental Milk Production and Procurement .... o...... 41 7.2 Economic Rate of Return .................. .................. . 42  - iv - Table of Contents (continuation) PaRe No. LIST OF ANNEXES Annex I - Project Costs Annex II - Organization Charts and Implementation Schedules Annex III - VMPC, Milk Production and Milk Procurement Annex IV - Financial Statements and Projections Annex V - Economic Analysis Annex VI - Disbursements Annex VII - Terms of Reference of Consultants Annex VIII - Documents in Project File NAP World Bank Number 18398  SRI LANKA SECOND DAIRY DEVELOPMENT PROJECT Loan and Proiect Summary Borrower: The Democratic Socialist Republic of Sri Lanka. Beneficiary: Dairy Development Foundation. Amount: US$38.0 million (including loan interest during construction). Terms: Repayment in 20 years, including 5 years' grace, at the standard variable interest rate. Relendinu Terms: The Government would relend the loan proceeds to the Dairy Development Foundation for 15 years, including 5 years' grace at a variable interest rate coinciding with the Bank's lending rate, plus a fee of 2 percentage points to cover foreign exchage risk. Proiect Description The project seeks to provide support for the long-term development of dairying in Sri Lanka by assisting in the establishment of an organizational structure for the dairy industry; increasing rural employment and incomes; improving efficiency in production, collecting, chilling, processing, and marketing milk; increasing the supplies of domestically produced and hygienically processed fluid milk; providing technical assistance and training; and strengthening the monitoring and evaluation of development projects in the livestock sub-sector. About 180,000 farm families are expected to benefit from increased on-farm employment, and more evenly distributed incremental income derived from increased milk production. Risks arise from the substantial institutional changes to be effected by the project, and farmer acceptance of the new village milk companies.  -ii- Estimated Costs: al Local ForeiRn Total (US$ Million) Land 0.8 - 0.8 Civil Works 8.6 1.9 10.5 Machinery and Equipment 4.4 15.6 20.0 Vehicles 0.6 1.0 1.6 Technical Assistance 0.8 2.7 3.5 Training - 0.9 0.9 Technical Services 1.9 5.3 7.2 Organization and Management 3.6 - 3.6 Support to VMPC 8.8 0.2 9.0 Working Capital hJ 4.0 - 4.0 Recurrent Costs c/ 7.8 - 7.8 Total Base Costs 41.3 27.5 68.8 Physical Contingencies 1.7 1.3 3.0 Price Contingencies 19.7 9.7 29.4 Total Project Costs 62.7 38.5 101.2 Interest During Construction - 10.0 10.0 Total Financing Required 62.7 48.5 111.2 FinancinR Plan: Local Foreign Total (US$ Million)---- Donated Commodities 39.5 17.4 56.9 IBRD 11.1 26.9 38.0 Netherlands 3.3 4.2 7.5 Short-term Loan (GOSL) 0.4 - 0.4 Funds from Operations 5.8 _ 5.8 Equity Investments 2.6 - 2.6 Total 62.7 48.5 111.2 a/ Including taxes and duties of US$4.2 million equivalent. b/ Incremental working capital for MILCO and MPC and initial working capital requirement for DDF. c/ Incremental costs related to DDF's operations and support to MlLCO and MPC. Estimated Disbursements: IBRD FY FY86 FY87 FY88 FY89 FY90 (US$ Million) Annual 0.6 14.5 13.7 5.6 3.6 Cumulative 0.6 15.1 28.8 34.4 38.0 Rate of Return: 23Z map: IBRD 18398  SRI IANKA DAIRY DEVELOPMENT II I. BACKGROUND A. Introduction 1.01 Although the major activities of the proposed project are oriented to dairying-the production, collection, processing, and marketing of milk-the project seeks a more far-reaching impact. Milk, although in short supply in Sri Lanka, is a means of bringing to rural areas development in its broadest sense. On one hand family based production and incomes would increase, and on the other an organization and management structure would emerge as a consequence of the formation under the project of autonomous rural based companies. In addition, the project would incorporate those dairy cooperatives which meet certain established criteria. The companies and cooperatives, besides being responsible for milk procurement and provision of technical and advisory serv- ices to livestock owners, would encourage the growth of leadership and manage- ment as similar organizations have done elsewhere in South Asia. With dairying being labor intensive and involving those with little opportunity for using their labor elsewhere, this organizational structure would encourage the par- ticipation of many of the poorest of rural households. Through the village units and the concomitant organizational structure, rural producers would have the opportunity to participate in the management, and share in the profits, of a processing and marketing chain that extended far beyond its rural base. Fur- thermore, through increased efficiency in the production, collection, processing and marketing of milk, farmers would be offered incentive prices to produce more milk and consumers would have increased access to high quality milk and milk products at fair market prices. B. Economic Setting 1.02 Sri Lanka is endowed with favorable resources and diverse agro-climatic conditions that permit wide-ranging agricultural pursuits. It also possesses a highly literate population and has achieved considerable success in curbing population growth. Sri Lanka comprises an area of about 65,500 sq km, with a population of about 15 million growing at about 1.7% p.a. GDP per capita is about US$320. About 75% of the population live in rural areas and are directly or indirectly dependent on agriculture. The climate is characterized by nearly constant temperatures and large variations in rainfall. The island can be divided into two major zones: the Wet Zone (average annual rainfall over 2,285 mm) and the Dry Zone (average annual rainfall belov 900 mm) which covers about 64% of the island. A narrow transition band between the Wet and Dry Zones is referred to as the Intermediate Zone. About three quarters of the population reside in the Wet Zone with an average population density of about 700 per sq km, compared to only 80 per sq km in the Dry Zone. The combined Wet and Inter- mediate Zones occur in the southwest quadrant and cover about 35% of Sri Lanks's land area, including about 45Z of the cultivated land. 1.03 Until the late 1970s, economic growth in Sri Lanka was slovw Real per capita GDP grew only 2.6% during the period 1970-76 mainly due to a combination  -2- of inadequate investment, ineffective and inappropriate government policies, adverse weather, and a sharp rise in the cost of imported fuel. This disap- pointing economic performance resulted in severe unemployment. In late 1977, the newly elected Government i'iitiated a liberalization strategy with the objec- tive of increasing the growth rate of production and employment, savings and investment, and exports. To achieve these objectives, the Government introduced policies which included exchange rate unification and depreciation, import trade liberalization, interest rate increases to encourage savings, full cost pricing for most imported commodities, and the establishment of a free trade zone to attract export oriented foreign private investment. The economy reacted vigorously to these changes with real GDP growth accelerating to an average of 7.3% p.a. during 1977-79, subsequently slowing to 5.0% between 1981-83. The primary sectors of significant growth have been paddy production, services, garment exports, and construction, all of which benefitted from liberalization increases and the increased level of investment. Furthermore, employment growth has been substantial in both the public and private sectors. Unemployment has been reduced further by out-migration in response to job opportunities abroad. As a result, the unemployment rate fell from 24% of the labor force in 1975, to 14.8Z in 1978, and 13.1% in 1982. Fixed investment likewise has risen from 14-15Z of GDP in the early 1970s to 30% in recent years. 1.04 However, underlying this improved performance were certain key struc- tural issues which the Government needed to confront to maintain medium-term growth prospects. Due to a marked jump in public capital expenditure and a disappointing revenue effort, the budget deficit increased from 1l.lZ of GDP in 1977 to 19.8% in 1980, declining to an average 13.7% during 1981-83. The budget deficit is expected to decline further to about 1OZ in 1984. Due partly to this budgetary deficit, and partly to a combination of external developments, and an incentive structure biased against exports, the current account deficit in the balance of payments also increased to a record 19.8% of GDP in 1980, leveling off at an average 13.7% during 1981-83. The Government has taken steps to reduce the budget deficit and is actively considering policy measures designed to stimulate exports and efficient import substitution activities. If progress continues to be made along these lines, the current account deficit in the balance of payments should decline to approximately ;Z towards the end of the decade with the debt service ratio declining to below 15Z in 1990 after tem- porarily rising to almost 17% due to existing debt repayment commitments. C. The Role of Agriculture in the Economy 1.05 Agriculture plays a major role in the economy of Sri Lanka, accounting for about 252 of GDP, about 50% of the labor force, and 60% of export earnings. About 95Z of the rural population is -nvolved directly in agriculture related activities. Sri Lanka's wide variations in precipitation, topography, and soils permit cultivation of a wide range of crops. Out of some 2.24 million ha under permanent cultivation (34% of land area), paddy accounts for 0.65 million ha, coconuts 0.50 million ha, tea 0.24 million ha, rubber 0.19 million ha, and other perennial crops 0.04 million ha; the remaining 0.62 million ha comprise mixed rainfed farming, mostly as small gardens around homesteads. An additional 1 million ha are under shifting cultivation (chena). Investment and incentive policy changes introduced by GOSL in the late 1970s for the agricultural sector were a response to disappointing growth in the early and mid 1970s. Average agricultural growth declined from 3.4% p.a. during the 1960s to 1.9% p.a. over  -3- the period 1970-77. Agricultural performance during the period 1978-83 improved substantially with growth averaging 3.8% p.a. This overall result was buoyed by an impressive increase in paddy production averaging 5.0% p.a., and was depressed by poor growth in the tree crops sector, averaging -0.96% p.a. Climatic vagaries, however, continue to plague Sri Lankan agriculture. A mar- ginal decline in paddy production in 1982 of 3.3% was attributable to an 1l1 drop in the 1981-82 maha crop due to severe drought, mitigated by a 1982 yala crop some 12% better than the previous year. Similarly, drought conditions during the 1983 yala crop were very serious, with normally planted areas either uncultivated or damaged. The 1982-83 maha crop (1.79 million tons), in con- trast, was a record, being about 30% above the previous year's depressed produc- tion. D. The Role of Livestock and Dairyinu in the Economy 1.06 The Livestock and Dairy Sector. The livestock population of Sri Lanka comprises about 1.7 million cattle (of which about 17% are either exotic or crossbreds), 0.9 million buffaloes, 0.5 million goats, 0.3 million sheep, and 0.75 million pigs. The contribution of livestock to the gross value of agricul- tural production is estimated at 8%, which is considerably lower than the actual contribution since this estimate does not take into consideration the value of manure, hides and skins, and draft power. The sector contributes significantly to employment generation and makes a major contribution to the economy by utilizing much land which would have very little alternative use. The dairy industry is comparatively small, representing only about 2% of GDP; however, dairying is an important activity for many smallholders and landless estate laborers. 1.07 Milk Production and Processing. Domestic milk production is estimated to range between 400,000 lpd (from national food consumption survey) and 700,000 lpd (from numbers of cows and buffaloes if milked). Of the total milk produced, the Dry Zone accounts for about 40%, the Wet Zone Hill Country about 25%, the Intermediate Zone, excluding the Coconut Triangle, about 15%, and the Wet Zone Low Country aud the Coconut Triangle about 10% each. Some 70% of the milk produced is consumed at home or marketed locally, either directly or through middlemen, and only about one third of the total is procured by the organized dairy sector. Collection is highly skewed towards the Wet Zone Hill Country where about 65% of the milk produced is procured. Comparable figures for other areas are 25% for the Dry Zone, 15% for the Coconut Triangle, 15% for the Inter- mediate Zone, and 10% for the Wet Zone Low Country. Milk processing is carried out in the four plants (total capacity of about 200,000 lpd) operated by the National Milk Board (NMB), one plant (capacity about 100,000 lpd) operated as a joint venture between NMB and a multinational company, and a dairy products plant (products include milk powder and baby food) with a reported capacity of 300,000 lpd recently completed by the multinational firm. Thus, the country's combined processing capacity is about 600,000 lpd. Current throughput (locally produced milk plus imported commodities) is about 265,000 lpd. 1.08 Milk Imports. In 1983, Sri Lanka imported about 27,000 tons of dairy products, which provided the liquid milk equivalent of a further 400,000 - 500,000 lpd. The Netherlands, New Zealand, and the U. S. accounted for about 80% of the imports. The total value of milk imports has increased from SLRs 55 million in 1970, to SLRs 380 mill-on in 1978, and SLRs 870 million in 1983. The  -4- Cooperative Wholesale Establishment (CWE) under the Ministry of Trade and Ship- ping is responsible for the importation of most milk products; however, private entities may also import. Sri Lanka presently receives little commodity aid in the form of dairy products. 1.09 Consumation and Demand. Average urban per capita consumption of milk and milk products is estimated at about 76 grams per day (gpd) compared to an average rural per capita consumption of 48 gpd. Consumption varies considerably between income classes with the poorer urban classes consuming an average of about 40 gpd and the upper income classes consuming up to 130 gpd. About 39% of the milk and milk products is consumed as fresh and pasteurized fluid milk, while the remaining is consumed as reconstituted milk powder (33Z), infant milk (21%), and other milk products (7%). In general, the consumption of dairy products in Sri Lanka is well below levels found in some other South Asian countries. According to estimates made by the Bank's dairy sector review mis- sion, jJ average per capita demand for milk and milk products in Sri Lanka is expected to rise by an average 1.2% annually during the project period. This increase, combined with a projected annual increase in population of 2.2%, vould give an overall increase in demand of 3.4% annually. Judging from observed increases in dairy imports (para 1.08), these projections appear conservative. Demand for milk and milk products is projected to grow from an estimated present level of about 840,000 lpd to approximately 1.2 million lpd by 1991, and 1.4 mil- lion lpd by 1996. Rising import prices would somewhat temper this demand as would a significant structural adjustment that reduces real per capita income growth rates. However, the vigorous promotion of dairy products by domestic and international companies would tend to expand demand beyond that predictable by price and income alone. While milk prices in Sri Lanka have nearly doubled since 1982 (para 1.10), the quantity and value of milk imports have increased substantially in recent years (para 1.08). This indicates that whatever dampen- ing effect price increases may have had on the demand for milk products, they have been outweighted by other factors (increases in income and population) leading to net increases in demand. These trends are expected to continue. The income elasticity for the demand for milk and milk products is about 0.6 to 0.8 at mean consumption levels, indicating that a 10% growth in per capita income would result in about a 6% to 8% growth in per capita milk consumption. 1.10 Milk Prices. Subsidies. and Tariffs. The Government controls both producer and consumer prices of milk bought and sold by NMB, which, in turn, controls all sales of liquid milk, except small quantities traded in rural markets. Up to March 1983, producer milk prices generally ranged from SLRs 2.00 to 2.80 per liter, depending mainly on fat content. In March 1983, the Govern- ment increased the producer milk price to about SLRs 3.20 per liter and the consumer price to SLRs 6.07 per liter. In September 1984, the producer milk price was further increased to SLRs 3.70 per liter and the consumer price to SLRs 8.00 per liter. While the consumer price remains at SLRs 8.00 per liter for 3.2% fat milk, the current producer price is SLRs 4.30 per liter for 4% fat milk as a result of a recent Government decision. Bulk imported whole milk powder (the main substitute for processed fluid milk), repackaged in Sri Lanka, currently retails on average at SLRs 8.90 per liter equivalent after payment of jJ Sri Lanka Dairy Sector Review, World Bank Report No. 4365-CE, Volumes I and II, April 18, 1983.  -5- a 15Z import duty. The present retail price set by the largest processor of domestically produced whole milk powder is SLRs 8.40 per liter equivalent. MM operates at a financial loss and has a negative cash flow which is financed by Government subsidies. These losses occur in large measure because of insuffi- cient margins between the Government controlled prices of raw milk and finished products, lov plant throughput combined with a high proportion of fixed costs, and overstaffing. The size of the subsidy has been reduced in recent years as a result of above-mentioned consumer price increases and increased efficiency in lNHBs operations. In 1983-84, the average annual subsidy payment to NMB amounted to SLRs 23 million, compared to SLRs 45.9 million in 1982. The Govern- ment had expected that continued improvements in the efficiency of MBB opera- tions would permit the elimination of the subsidy by the end of 1984. The necessary improvements in MBB were not forthcoming and in the November 1984 budget the subsidy was raised from SLRs 0.375 to SLIs 0.50 per liter of raw milk processed as fluid milk. The Government, however, remains committed to the elimination of the subsidy. There is at present a need to ensure that some domestic raw milk is incorporated into locally produced whole milk powder (which also incorporates reconstituted imported skim milk powder). Without this out- let, some rural producers would lack an effective market for their milk. Also, markets distant from milk processing plants would be mainly dependent upon imported milk products. To make local raw milk an attractive ingredient for local milk powder processors, the Government presently pays a subsidy of SLRs 1.00 per liter on milk absorbed by these processors. The subsidy appears to be necessary in order for domestically produced whole milk powder to be competitive. The total subsidy to NMB budgeted for 1985 is SLRs 33 million. Until 1983, CWE also received a subsidy to cover its operating losses. The subsidy amounted to about SLRs 5 per kg of powdered milk; the annual cost to the Government was about SLRs 100 million. In March 1983, the consumer price of milk powder imported and repackaged by CWE was raised sufficiently to eliminate the subsidy. The Government has, thus, initiated the process of economic rationalization of the dairy industry and made substantial progress in diminish- ing distortions in pricing which have mainly benefitted the higher income con- sumers and had little effect on productivity at the farm level. Until January 1985, import duties on whole milk powder, skim milk powder, and butter oil were 5%, 7.5%, and 12.5%, respectively. Effective January 1, 1985, the tariff on all these products was increased to 15%. E. Institutional Support to Livestock 1.11 Responsibility for Development. Organizational and administrative responsibility for livestock and dairy development rests with the Ministry of Rural Industrial Development (MRID). Within the Ministry there are four separate entities involved. These are: (a) the Department of Animal Production and Health (DAPH) (provision of veterinary and extension services to producers, operation of breeding farms, training and research, organization of dairy cooperatives); (b) the National Livestock Development Board (NLDB) (development projects, establishment and maintenance of livestock breeding stations and farms, and organization of dairy cooperatives); (c) the WMB (milk collection, processing, and marketing); and  -6- (d) the National Oils and Fats Corporation (principal supplier of compound feeds to the livestock industry). The responsibilities of these bodies are not clearly delineated, and in some areas, for example in dairy development and formation of dairy producer coopera- tives, there is considerable overlapping of functions between the DAPH and the !LDB. Still another agency, the CWE under the Ministry of Trade and Shipping, is involved in dairy development, at least indirectly, since this body is responsible for the import of most dairy products. Because the CWE has been required in the past to cross-subsidize imported milk, it has indirectly influenced the price paid to producers and the losses sustained by the private and public dairy industry. The dairy plants and chilling centers controlled and managed by the NMB have not operated satisfactorily. Trade union problems, over staffing, low productivity, substantial losses, and inaccurate accounting sys- tems have plagued the operation of these plants. 1.12 Livestock Research. The principal institutions engaged in livestock research are the Veterinary Research Institute, the Faculty of Veterinary Medicine and Anima. Science of the University of Peradeniya, and the Department of Animal Husbandry of the Faculty of Agriculture of the same university. However, the research output is not large in relation to the numbers of qualified staff at these institutions. Moreover, there appears to be little or no joint research by inter-disciplinary teams aimed at solving the development problems of the livestock sector. The organization of livestock research is highly centralized and there is no regional research conducted which could lead to the formulation of more realistic regional development plans making maximum use of local resources. 1.13 Livestock Extension. The Department of Animal Production and Health is the principal agency of Gover,nent entrusted with providing extension serv- ices to the livestock sector. Unfortunately, as presently constituted and organized, the Department is far more animal health than animal production oriented and in consequence is trnable to make a major direct impact on national animal production per se. Furthermore, the interaction which must exist between extension and research if the production problems of the livestock owner are to be adequately addressed is not apparent. 1.14 Credit. Institutional rural credit in Sri Lanka is principally dis- bursed by two nationalized commercial banks, the People's Bank (PB) and Bank of Ceylon (BC). PB operates through its banking divisions, the Cooperative Rural Banks. BC provides credit services through its main branches, with repre- sentation maintained at Agrarian Service Centers (ASC). National rural credit policy is established by a Rural Credit Advisory Board (RCAB) chaired by the Governor of the Central Bank (CB). CB's Department of Rural Credit functions as the Board's secretariat and executive. The total amount of institutional rural credit disbursed nationwide declined in nominal terms from SLRs 156.8 million (1981) to SLRs 153.9 million (1982). This decline was primarily the result of a drop in lending under the Comprehensive Rural Credit Scheme (CRCS) by PB. High default rates continue to be an unsatisfactory feature of rural lending. From 1967 to end-1982, initially under the New Agricultural Credit Scheme and sub- sequently under the CRCS, only about 482 of amounts disbursed was recovered. A rescheduling scheme with a view to improving recoveries is now underway. In  -7- addition to institutional credit, there is evidence that a substantial share of rural credit comes from non-institutional sources. 1.15 Dairy Cooperatives. Dairy cooperatives in Sri Lanka date back to about 1930; however, the major force in the cooperative movement, the Cooperatiie Union in Colombo, was organize4 during the 1940s and continued to work effi- ciently until the inception in 1954 of the State-owned NMB. The NMB started selling milk in Colombo at prices lower than those of the Cooperative Union, which was run on sound financial principles, and eventually drove the Union out of business in 1958. The dairv cooperative movement resurged in 1979 under the IDA-assisted dairy project (Cr. 504-CE). 1.16 At present there are 100 milk cooperatives in Sri Lanka, 90 of which are classified as small primary societies and 10 as large primary societies. Cooperatives of the former type cover a relatively small area (generally one village) and their by-laws do not provide for elected branches. Some 50 of these societies, known as Dairy Producers Associations (DPA), were organized by the NLD3 under the first dairy project and are members of the Coconut Tri- angle Milk Producers' Cooperative Union. The large primary societies, organized by both the NLDB and the DAPH, have adopted by-laws similar to the existing multi-purpose cooperatives and consist of central units with varying numbers of branches which are supervised by elected branch committees. The central units of these societies are frequently located at the headquarters of the Assistant Government Agent, and their executive comittees generally consist of four government nominees and five members selected from among the producers by an electoral college. Four of the ten large primary societies each cover an administrative district. Because of its more centralized nature, distance of the Central Unit from the villages, and the presence of government nominees on its board, producer influence and cooperative autonomy are limited, and the large primary society is often viewed as another agency of the Government. The dairy cooperative societies to be formed under the ADB-assisted livestock development project (para 1.29) would be of the large primary type. 1.17 The total membership of the primary milk societies is about 23,000 and, since 1979, these societies have averaged about 31 million litres of milk col- lected annually (about 54% of the total milk collected by NMB). Although many of these societies have operated with considerable degree of autonomy and with producer-based participatory management, they have had little or no involvement beyond the village level, i.e., in milk processing and marketing. Furthermore, while both types of cooperatives offer advantages in terms of economies of scale in mill- collection and transport compared to the private collector, both place the producer in a disadvantageous position in terms of bargaining for price increases and for the provision of production inputs. As a consequence, producer milk prices are only about 54% of retail prices (para 1.10) vith margins between collection, chilling, transport, and processing being extensive. Likewise, retail price increases have a significant likelihood of not being fully reflected in producer prices but may result instead in increased margins and losses between producer, cooperative, chilling center, processing plant, and consumer. Thus, a reorganization of the existing cooperative system, which would allow producers wider participation and greater control, is needed (paras 2.02 to 2.06).  -8- F. Livestock Production Systems 1.18 The island is divided into seven major agro-ecological zones on the basis of annual precipitation, rainfall distribution, and altitude, and these are further subdivided into 24 agro-ecological regions, each with its unique combination of rainfall pattern, elevation, land form, temperature range, and types of soil. XI The combination of these factors determine the prevailing land use, cropping pattern, crop calendar, production constraints and develon- ment potential in different regions. 1.19 The total number of milk cattle and buffaloes in 8ri Lanka has been recently estimated at 552,000. These are distributed among the three rainfall areas of the country as follows: Table 1.1: DISTRIBUTION OF CATTLE AND BUFFALOES IN SRI LANIA Cattle Buffaloes Number Z Number Z Wet Zone 82,900 20 24,000 18 Intermediate Zone 65,800 15 7,700 6 Dry Zone 272.000 65 99.500 76 Total 420,700 100 131,200 100 The Dry Zone is the area of greatest potential for further expansion in milk production and procurement. 1.20 There are six major and distinctive milk production systems in the country. Each system has evolved in response to the natural resources, markets, and alternative income opportunities available to the producers. Five of the systems are based on cattle and the sixth on buffaloes. To a large extent one system tends to dominate in a particular agro-ecological zone, however, within such zones other production systems are also found. 1.21 The "Estate" System. The Mid and Hill Country of the Wet Zone and the wetter parts of the Mid and Hill Country of the Intermediate Zone are largely planted to tea and rubber. Most of the land is held by large estates now operated by state plantation agencies but which was privately owned prior to land reform. A few small (maximum 20 ha) private estates remain. Dairying is carried out by estate laborers (perhaps 10-20X of the total) to supplement their income. Because of a land shortage, all cattle are entirely stall fed with jj Mean air temperature ranges from 160C to 280C. The 75% expectancy value of annual rainfall from 508 to 3,175 mm. Land form varies from mountainous, steeply dissected hills to rolling, undulating and flat. There are 17 major soil groups and associations.  -9- concentrates (mostly poonac, with some rice bran) and forage cut from roadsides and non-arable land on the estates. Estate laborers are generally limited to a maximum of two COws per family. 1.22 The Mid Country Smallholder Systen. Within the Mid and Hill Country of the Wet and Intermediate Zones there are numerous smallholdings generally comprising very small areas (0.1 to 1 ha) of paddy land located in the valleys and associated with larger areas (1 to 3 ha) of upland, part of which is planted as a home garden. The steeper uplands are planted to fruit trees (jackfruit, durian, mango), beverages (tea, coffee, cocoa), and spices (pepper, cloves, cardamon, and nutmeg). A proportion of the upland areas comprises areas of natural grassland and tea land abandoned because of low yields. Such land is generally badly eroded, low in nutrients, quite acid, and suited only to pasture and forestry. Dairying is still very much a secondary activity with crop cul- tivation taking priority. A typical smallholder with access to pasture has about four purebred or European crossbred cows vhich are fed cut green fodder and concentrates. Animals are stall fed throughout the year. Management prac- tices are generally similar to those followed by cattle oNners under the 'Estate" System, although slightly lower management standards result in somewhat lower milk yields. 1.23 The Coconut Triangle System. The Coconut Triangle covers an area of some 0.5 million ha with annual rainfall ranging from 1,000 mm in the north to 1,800 mm in the southwest. It receives rain from both the northeast and south- west monsoons. Virtually all the lowest lying land is used for paddy cultiva- tion and the higher land for coconuts and home gardens. The coconut growing area comprises about 0.35 million ha and, prior to land reform, most of this area was held by large estates. There are two distinct production systems. The first is based on small herds of indigenous cattle which are grazed under coconuts on partly improved pasture and fed small quantities of coconut poonac. The second, and more common, is found on the larger private estates (up to 20 ha) and involves the grazing of crossbred cattle on improved pastures planted under coconuts. 1.24 The Dry Zone System. The Dry Zone extends over about 4 million ha or about 62Z of the total area of the country and accounts for about 25% of agricultural holdings with 0.6 million ha being distributed among some 0.6 million- smallholdings. The remaining 3.4 million ha is mostly state land which is used by the smallholders for extensive grazing. The grazing areas consist mainly of vast extents of secondary forest and scrub resulting from centuries of shifting cultivation. However, there are limited areas of reasonably productive natural grasslands adjacent to the major rivers and tanks. These areas, known as vil- las, are generally overstocked and are increasingly being transformed into irrigated farm land. The extensive grazing of livestock presently practiced in the Dry Zone is likely to be progressively replaced by more intensive livestock rearing within integrated small farms. Present cattle herd size averages about -40 head with about 40Z of the herd being cows. These state lands also provide grazing for buffalo herds in areas adjacent to paddy production. 1.25 The Urban Jaffna System. Of the 65,000 ha under cultivation in the Jaffna district, rice accounts for about 60%, coconuts 9%, chillies 7%, onions 6%, and groundnuts, sesame, pulses and maize account for the remaining 8%. The Jaffna district is reported to have a cattle population of about 150,000 head of  -10- which about 50,000 are located on the peninsula and the remainder on the main- land. Of these 50,000, some 12,000 are either crossbreds or improved breeds with a high proportion of European blood. The dairy industry on the peninsula is largely based on the 12,000 improved cattle. Individual producers keep two to six cows which are stall fed on concentrate, roughage (mainly straw), fodder trees, and occasionally some green forage. 1.26 The Smallholder Buffalo System. Buffaloes are mainly associated with the provision of draft power for paddy cultivation in the Low Country and in the Dry Zone. More than half of the total population are located in five dis- tricts: Kurunegala 18%, Anuradhapura 12%, Hambantota 8%, and Polonnaruwa and Batticaloa with 7% each. An estimated 70% of the buffaloes are located on holdings of two ha or less. Most of the buffalo milk is a by-product from the provision of draft power. Few buffaloes are used solely for milk production. A fairly typical herd numbers about 18 animals, of which 7 are cows with about 4 in milk at any time. The herd grazes waste (state) land during the day and returns to a night paddock in the evening. No additional feed is provided. G. Development Activities and Project Rationale 1.27 District Rural Development Proiects (RDP). The IDA-supported Kurunegala RDP (Cr. 891-CE) was the first district level RDP implemented by GOSL in 1979. This was followed by the IDA-financed Second (Cr. 1079-CE) and Third (Cr. 1363-CE) RDP in Matale and Puttalam districts and Mannar and Vavuniya districts, respectively, and by the Badulla RDP (Ln. 111-CE) financed by IFAD. In addi- tion, there are RDP underway in Hambantota and Nuwara Eliya districts with financial assistance from Norway and the Netherlands, respectively. Each of the RDP has a rather small livestock component implemented, in most cases, by DAPH. 1.28 Bilateral Livestock Development Proiects. Tnese include the Sri Lanka-Netherlands and the Sri Lanka-Swiss Livestock Development Projects and the Sri Lanka-West German Goat Development Project. The first began in 1978 with a gift of 900 exotic dairy animals to Sri Lanka and includes activities in the areas of dairy development, mainly in connection with the development of two NLDB farms, and training in dairy farming; treatment of straw to increase its feeding value; and improvement of animal draft power, mainly in the areas of the Mahaweli Development Program. The Swiss-supported project covers the dis- tricts of Polonnaruwa, Batticaloa, Trincomalee and Amparai, and inter alia is involved in the establishment and operation of district dairy cooperatives (para 1.16). 1.29 ADB-Assisted Projects. The Anuradhapura Dry Zone Agriculture Project, which also receives support from IFAD, was initiated in January 1981. The livestock component of the project is concerned with the strengthening of veterinary and AI services, expansion of two NLDB livestock farms, provision of credit for the purchase of draft animals, and the establishment of a small feed mill and three milk chilling centers in Anuradhapura district. In January 1983, GOSL began the implementation of an ADB-financed livestock development project. The project covers 10 districts and the main components are dairy development, improvement of animal draft power, upgrading of pig production, development of poultry, livestock education and training, improvement of feed and fodder supplies, strengthening of veterinary services, improvement of marketing facilities, and the provision of credit. Under the dairy component,  -11- which covers the districts of Jaffna, Kegalle, Gampaha, Colombo, Kalutara and Galle, A.I. services would be strengthened through the importation of breeding animals and semen, and assistance would be given in the establishment and opera- tion of dairy producer cooperatives (para 1.16). According to ADB's latest review mission, the project is some six months behind schedule. 1.30 Bank Involvement in the Dairl Sub-Sector and Proiect Rationale. The Bank's involvement in the dairy sub-sector in Sri Lanka began with a 1971 Agricultural Sector Mission which concluded that a large scale dairy development program was feasible. In early 1973, a project was prepared by a Sri Lankan interagency team with assistance from the FAO/IBRD Cooperative Program. Appraisal took place in mid-1973 of a multi-component project which included credit and technical services for on-farm development, supply of equipment for expansion of milk collection, credit to cooperatives, importation of dairy heifers, and the establishment of demonstration units for calf rearing and forage production and management. 1.31 As a result of slow progress, the project was reformulated in 1978. Emphasis was shifted from providing credit to commercial dairy farms to the formation of dairy cooperatives (para 1.15) based upon the Anand pattern found to be so successful in India. It was also agreed that the Government would raise the price of milk to producers and stabilize concentrate feed prices. In November 1978, the Government raised the procurement price by 38%. In addition, subsidies were offered to producers of SLRs 120 to SLRs 2,500 per acre for fodder production, and heifers were sold from calf rearing units at about 30% below costs. Meanwhile, consumer prices were subsidized resulting in annual operating deficits of the NMB ranging from SLRs 33 to 96 million. In the liberalization of the economy in the late 1970s, export prohibitions on poonac, the main ingredient of concentrate cattle feed, were lifted, raising prices to international levels and reducing the profitability of dairying. 1.32 After 1978, further dialogue between IDA and the Government on milk pricing and subsidies had little effect and progress of the project remained generally slow, though formation of DPA, and milk collection from them, was sufficiently successful to demonstrate the validity of this approach. In Decem- ber 1981, the project closed after a one year extension. Approximately US$4.9 million of the US$9 million credit remained unutilized and was cancelled. It was clear from the outcome of the project that issues on institutional struc- ture and pricing, and project performance could not be separated. Consequently, it was agreed that, before embarking on further livestock investments, a joint GOSL/IDA review mission would study the livestock and dairy sector. 1.33 Between November 14 and December 3, 1982, a sector mission jl visited Sri Lanka to review and discuss with Government organizational and pricing )j This mission and the subsequent follow-up mission received substantial assistance from the India National Dairy Development Board which has suc- cessfully guided smallholder dairy development throughout India through the formation of village cooperatives under four IDA-financed projects. A full description of the sector issues reviewed and discussed is presented in the Sri Lanka Dairy Sector Review, World Bank Report No. 4365-CE, Volumes I and II, April 18, 1983.  -12- issues vithin the dairy industry. A smaller, follow-up mission returned between February 27 and March 4, 1983, to continue this policy dialogue. X} The review confirmed the potential of the dairy industry and outlined three policy options for its development: (a) Continuing with a Government-based milk board and processing industry; (b) Inducing marginal changes in organization and the supply of technical inputs whilst maintaining the NMB; and (c) Initiating substantial institutional changes leading toward privatiza- tion of the industry. After review of the findings of these missions, GOSL selected the third option and invited preparation of a project along the lines suggested by this option. Responding to this invitation, and taking advantage of the data gathering and analytical work accomplished in the sector review, the Bank mounted a joint preparation/appraisal mission in November 1983. 1.34 The rationale for continued Bank involvement in the development of the Sri Lankan dairy industry is: Ca) The potential of dairying to bring benefits to large numbers of the rural poor whilst also achieving significant foreign exchange savings in an economically efficient way; (b) The need for radical institutional changes in the industry if this potential is to be realized; and Cc) The ability the Bank has to guide these changes based on considerable experience with dairy development in the Region, in particular with the apparently very successful cooperative dairy development program in India. II. THE PROJECT A. Project Area 2.01 The proposed project would cover the entire country which would be divided into five milksheds for operational convenience: (a) Jaffna, including the districts of Jaffna, Hullaitivu, Mannar, Kil- linochi, and Vavuniya, all of which are located in the Dry Zone; (b) Colombo including the districts of Colombo, Kalutara, Kegalle, Matara, Ratnapura, Galle and Gampaha, all (with the exception of about 40Z of Gampaha district) of which are situated in the Low or Mid Country Wet Zone; (c) Kandv. including the districts of Kandy, Matale and Nuwara Eliya. Kandy is located in the Mid Country Wet Zone; Matale in the Mid Country Inter- mediate Zone; and Nuvara Eliya in the Hill Country Wet Zone;  -13- (d) Polonnaruwa- including the districts of Polonnaruwa, Anuradhapura, Puttalam, Kurunegala and Trincomalee, all of which are in the Dry Zone, except Kurunegala which is located in the Low Country Intermediate Zone; and (e) Monarakala. including'the districts of Monaragala, Badulla, Amparai, Hambantota and Batticaloa, all of which are in the Dry Zone, except Monaragala district which is partly situated in the Intermediate Zone. B. Proiect Design 2.02 The present institutional structure for milk processing and marketing in Sri Lanka has not operated according to expectations despite subsidies and public investments in milk plants and chilling centers (paras 1.10 and 1.11). Furthermore, pricing policy has tended to favor the relatively better off urban consumer at the expense of the rural producer. As a consequence, dairying has had very little incentive to expand and has not attained the economic and dis- tributional impact that it could achieve. To ensure that both rural producers and urban consumers gain from a program of dairy development, major institu- tional and organizational changes which result in increased milk production and promote efficient processing and marketing are required. The proposed project therefore would give primary emphasis to these aspects, and the project design draws heavily on the experience gained in cooperative dairy development in India, particularly over the past decade. 2.03 The system of dairy cooperative development .in India, known as the AMUL or Anand pattern, has been developed over the last 35 years in the Kaira Dis- trict of Gujarat State and tested widely throughout the country with con- siderable success. Essential elements of the AMUL model involve the formation of a cooperative society of milk producers at the village level, the selection of a board of management which sets the policies of the society within the framework of a general set of by-laws, and the appointment of a local person as the paid secretary of the society. Each morning and evening the society buys milk, on a commission basis, from all producers in the village who wiuh to sell it, and makes payment to each producer based on the volume and fat content of milk. The society also sells compound cattle feed and provides an artificial insemination and veterinary first-aid service to village cattle. All milk producers in the village are eligible to be society members irrespective of cast or sex. Membership requires agreement to sell milk to the society only, to subscribe to at least one share of the society, and to have one vote per member with only one member per household. A further key element in the AMNL model is that the 400 or so societies within a 50-75 km radius are members of a collec- tively owned milk union. Milk from the constituent societies is brought to the union dairy plant where it is pasteurized and sold as fluid milk or processed into milk products. Each milk union has a board of directors elected by the producer-members. The board employs a professional manager who is the union's chief executive and professional staff who provide the society members the technical inputs needed to sustain and increase milk production. Thus, the unions organize mobile veterinary clinics, artificial insemination services, and supplies of balanced cattle feed which are marketed to producers through their societies. The milk unions within a state are joined together tc form federa- tions, enabling all member-unions to benefit from shared processing, markLting,  -14- financing and investment programs managed by specialists employed by the federa- tions. Thus, the AMUL model is a three-tier structure of village dairy coopera- tives, unions of village dairy cooperatives, and federations of milk unions; the whole system is owned by the primary producers of milk and operates to their benefit. 2.04 In order for producers to receive an incentive price to produce more milk and for consumers to be protected from unduly high retail milk prices, maximum efficiencies must be sought in milk collection, processing and market- ing. To ensure these efficiencies it is recommended that Government divest itself of the management of investments in milk processing and marketing vhich are essentially commercial activities better handled by the private sector. As long as these activities are carried out by the public sector, pricing and marketing (as well as production) remain political issues vith Government taking the brunt of criticism for any inefficiencies or price increases. In addition, a private company, unlike a public enterprise, would have full autonomy in pricing and management and could pay sufficient wages to attract high caliber staff (paras 5.02 and 7.14). 2.05 For dairy producers to receive the benefits of higher prices, a reor- ganization and expansion of the existing cooperative system is needed (paras 1.15 to 1.17). In particular, producers should be granted greater con- trol and they should be permitted to enter into partnership with entrepreneurs in building the dairy industry. Therefore, the proposed project would establish an organizational infrastructure patterned after the India AMUL model except that the various entities created (i.e. village producer units, milkshed units, milk processing and marketing unit, and the apex organization) would be registered under the Companies Act and not under the Cooperatives Act. This would have the advantage of embodying the most desirable features of smallholder dairy cooperatives yet minimizing public sector influence and would place each of the entities under the same requirements of accountability and procedures as any other private sector firm. The proposed institutional changes would ensure that the technical inputs necessary for increasing milk output and improving the efficiencv of production are provided. They would also aLlow producers to participate in the profits obtainable in milk processing and marketing and vould provide a mechanism by which incentive prices could reach the farmer. 2.06 Given the substantial institutional and infrastructural development and the indivisibility of investments, the project would be implemented over seven years. Experience with similar projects in India has shown that a relatively long implementation period is necessary to construct and get into operation milk processing plants and to allow adequate time for the new institutions to build sufficient expertise to operate efficiently vithout outside assistance. To reduce risk and to ensure rapid implementation following loan approval, certain pre-investment activities have been completed Cpara 4.25). In addition, a mid-term review (para 4.27) would be carried out at completion of the third project year to ensure necessary adjustments required for effective implementa- tion.  -15- C. Obiectives 2.07 The primary objectives of the proposed project are: (a) To create an institutional structure to provide technical support and financing for the long-term development of dairying in Sri Lanka; (b) To increase rural employment and incomes through expanding milk produc- tion and by providing autonomous organizations for the efficient collec- tion, chilling, processing and marketing of milk, and by allowing producers to share in the management and profits of each of these activities. The producer s share of consumer payments for processed milk, currently about 54% (para 1.10), is targeted to reach 60% with the project; (c) To increase the supplies of domestically produced and hygienically processed fluid milk to urban areas; (d) To provide technical assistance and training to support the development of dairying; and (e) To strengthen and support the monitoring and evaluation of projects in the livestock sector. D. Proiect Components 2.08 To accomplish its objectives, the proposed project would implement the following components over seven years. Organization and Management 2.09 The project would support the establishment of an organizational struc- ture for the dairy industry consisting of: (a) Dairy Development Foundation (DDF), a non-profit guarantee limited company to provide finance for investments in dairying, as well as training, technical assistance, research and development, and management support in organizing and developing milk production, collection, processing and marketing; 1j (b) Milk Industries of Lanka Companv Limited (MILCO), owned by producers and private investors, to lease and manage 4 dairy plants currently owned by the EMB, to lease the existing NMB chilling centers, and to invest in about 3 new milk plants and about 10 new chilling centers; AJ This would be a company limited by guarantee with no share capital. Such companies are generally incorporated for charitable or quasi- charitable purposes. It would not obtain working capital from or pay dividends to its members. The liability of each member is limited to the amount (not to exceed SLRs 100) he guarantees to contribute in the event of liquidation of the company.  -16- (c) Five Nilkshed Producers' Companies (NPC). owned by producers, to lease from MILCO and manage NMB chilling centers as vell as new chilling facilities. The MPC would support village companies and cooperatives through the provision of services, and vould collect milk from these units and transport it to chilling centers and/or dairy plants; and (d) About 1.950 Village Milk Producers' Companies (VMPC), owned by producers, to collect milk from and provide technical and extension services to producers. 2.10 The project would establish these companies, and project costs would include the costs of organization and incorporation, as well as costs of initial staff, office materials and equipment, vehicles, and Spearhead Teams to under- take the establishment and efficient operation of VHPC (para 2.15 (a)). In the case of DDF, the project would finance operating expenses for the duration of the project. Assurances were obtained that GOSL would lease to MILCO NMB's dairy plants and chilling centers and abolish the IMB by September 30, 1986. It would be a condition of effectiveness that the lease agreement between NMB and MILCO had besn entered into. The better qualified NMB staff would be absorbed by DDF, MILCO and the MPG; the remainder would be granted normal redundancy payments by GOSL. 2.11 The existing dairy cooperatives, as well as those formed in the future, would participate in the project provided they are small primary cooperatives; they are financially viable; have management boards, the majority of whose members have been elected from among the members of the cooperatives by secret ballot; refrain from hiring, employing or otherwise renumerating employees of Government; and agree to abide by the same rules of conduct and procedures as the VMPC. The cooperatives, which would be referred to as Village Milk Producers' Associations (VMPA), would buy shares in the MPC and have the same benefits and obligations under the project as the VNPC. Dairy Plants and Chilling Centers 2.12 To modernize, expand, and improve the efficiency of milk chilling, processing, and marketing, the project would: (a) Refurbish the existing Colombo milk plant; (b) Construct a pasteurization and URT milk plant in Colombo of about 200,000 lpd capacity; (c) Construct a pasteurization and dairy products plant in Killinochi or Vavuniya of about 60,000 lpd capacity; (d) Construct a feeder-balancing dairy jJ in Amparai of about 100,000 lpd capacity; and j/ During the lean season, a feeder-balancing dairy transfers practically all of the milk it receives to urban dairies; in the flush season, it transfers only a portion to urban dairies and converts the surplus milk into dairy products.  -17- (e) Construct new chilling centers in the following areas: Haputale, Baliela, Nawalapitiya, Chenkaladi, Ruwanwella, Vavuniya, Chilaw, Horana and Tissamaharama (about 10,000 lpd capacity each), Kotagala and Norwood (about 20,000 lpd capacity each). Refurbishing of the old Colombo plant for the reconstitution of donated com- modities would be carried out as a part of the Netherlands assistance. To ensure that the designs and layouts of the new dairy plants and chilling centers are completed in a timely manner, the Government gave an assurance that the Dairy Plant Engineer would be appointed by December 31, 1985. Government also gave an assurance that not only the final designs but also acquisition of the building sites for the processing and chilling facilities whose construction is scheduled for Project Year 2 would be completed by September 30, 1986. Technical and Extension Services 2.13 To help increase the productivity of the national dairy herd, the project would provide the following support to milk producers; (a) Artificial insemination (AI) services using imported frozen semen; (b) Routine and emergency veterinary services; Cc) Para-veterinary first-aid services; and (d) Production inputs such as planting materials, compound cattle feed, and minerals; and ae) Advisory services on fodder and pasture production, animal management and feeding practices, crop-livestock integrated farming, calf rearing, and milking hygiene. 2.14 The provision of these services would be at three levels. Firstly, centralized services provided by MILCO to MPC which would inclua_ bulk importa- tion, storage and distribution of frozen semen, as well as the production, storage and distribution of liquid nitrogen. Secondly, services provided by the MPC to the VMPC/VMPA which would include distribution of semen on a regular basis; the operation and maintenance of training centers and central animal husbandry and veterinary units (including diagnostic laboratories), with sub-units located at sub-district level; and veterinary services, for vhich a fee would be levied for each emergency call attended. And thirdly, those provided by the VHPC/VMPA to their members which would include AI services for cattle, animal first-aid care, and a feed and livestock extension program through trained village level staff (para 2.15 (e)) who would maintain daily contact with farmers. VMPC/VMPA members would pay a fee for each AI performed. The fees charged for veterinary and AI services would cover at least direct costs. Some 200 farms owned by progressive farmers would be selected and used for demonstration of improved methods of animal husbandry, pasture and fodder production, crossbreeding, calf rearing, crop-livestock integration at the farm level, and hygienic milk production.  -18- Traininra 2.15 An important aspect of the project would be the training of farmers and staff of DDF, MILCO, the MPC, and the VMPC;VMPA. The following training programs are envisaged under the project: (a) Spearhead Teams. Each MPC would have a Spearhead Team who would be entrusted with the task of organizing VMPC. Each team would comprise about 20 individuals and would include veterinarians, extension officers, and agricultural graduates, as well as graduates in other disciplines. Each team would undergo a comprehensive training program of about four months in the organization of milk yroducers at the National Dairy Development Board in Anand, India. (b) Trainers. To staff the training centers to be established in each milkshed, 15 trainers would be required initially. These would be recruited and trained at the National Dairy Development Board. The duration of training would be about four months. tc) FellowshiDs. Selected personnel of DDF, NILCO and the MPC would be provided with overseas fellowships to study organization and management, project planning, dairy plant construction, dairy plant operation and management, quality control, marketing, monitoring and evaluation, and livestock and fodder production. Most fellowships would be for four months duration and the project would provide a total of about 100 man months of overseas training. (d) Farmers' Study Tours. About 500 progressive milk producers selected by the Spearhead Teams would visit Anand and other milksheds in India to observe the operation of milk producers' organizations. The tours would be 7-10-days in duration. (e) VMPCIVMPA Staff. The staff would be trained at the MPC training centers in artificial insemination, animal first aid, fodder production, dairy production, milk testing, basic record keeping and accounting. The duration of training would be about two months. The centers would also be responsible for the training of project farmers. DDF would have overall responsibility for supervision and administration of the program. Technical Assistance 2.16 Technical assistance would be essential for planning and implementation of the proposed project. Since considerable specialist expertise is available from professors in Sri Lankan universities, retired GOSL staff, private sector managers, and in personnel of some Government departments, local expertise would be used vherever possible. However, in certain areas, such as dairy engineering and farmer organizations, such experts are not available locally and would need to be recruited internationally in the early years of the project. The proposed project would provide a total of 408 man months of technical assistance as indicated in Table 2.1.  -19- Table 2.1: TECENICAL ASSISTANCE Months Management and Manpower Development 24 Dairy Engineering 48 Dairy Processing 48 Quality Control 24 Dairy Marketing 24 Monitoring and Evaluation 24 Planning 24 Farmers Organization (5) 120 Livestock Production 24 Feed and Fodder Production 24 Liquid Nitrogen Production and Distribution 24 Total 408 2.17 The average man month cost of expatriate technical assistance is estimated at US$8,000 which would include travel, subsistence and allowance. if Terms of reference for all technical assistance are given in Annex VII. DDF, after consultation with MILCO and the MPC, would be responsible for selection of consultants and for administration of the technical assistance program. Monitorin and Evaluation 2.18 The project would undertake monitoring and evaluation separately. Within DDF, a project monitoring unit would be established to monitor physical and financial progress of the project. Within KRID, an existing technical monitoring unit would organize evaluation studies to independently assess the impact of the project. To enable it to do this the project would strengthen the unit through the provision of about 12 man years of technical staff, 21 man years of support staff, equipment, and vehicles. III. PROJECT COSTS AND FINANCING A. Project Costs 3.01 Total project cost is estimated at SLRs 2,780 million (US$111.2 mil- lion), including SLRs 104 million (US$4.2 million) for taxes and duties on imported vehicles machinery and equipment, SLRs 250 million (US$10 million) for interest during construction, and SLRs 140 million (US$5.6 million) for incremental working capital. Cost estimates are based on end-CY1983 prices jj Based on the assumption that a high proportion of consultants would come from India.  -20- updated to March 1985. The foreign exchange component is estimated at SLRs 1,224 million (US$48.5 million), which represents 44S of total project costs. Physical and price contingencies over the project period amount to SLRs 809 million (US$32.4 million), or 29Z of total costs. Physical contingen- cies have been estimated at 5Z of base costs, except for vehicles for which no physical contingencies have been provided. Price contingencies have been applied at 8% in 1985; 9Z in 1986-1988; 7.5% in 1989; and 6% in 1990-1992 for foreign costs. For local costs, rates have been applied at 12Z in 1985; 11% in 1986; 10% in 1987; 9% in 1988; 7.5Z in 1989; and 62 in 1990-1992.  -21- 3.02 Ihe project cost estimates (Annex I) are sumarized in Tables 3.1 and 3.2. Table 3.1: PROJECT COST SDKARY SLIs Million US$ million x Z Total Local Foreian Total Local Foreian Total F.E. Base Costs DDF 101.6 96.7 198.3 .4.1 3.9 7.9 49 12 MILCO 383.4 503.0 886.4 15.4 20.1 35.4 57 52 Milkshed Producer Companies af Polonnaruwa MPC 12'. ' 20.7 145.8 5.0 0.8 5.8 14 8 Colombo MPC 112.8 17.7 130.5 4.5 0.7 5.2 14 8 Jaffna MPC 122.6 21.7 144.4 4.9 0.9 5.8 15 8 Kandy NPC 101.1 16.0 117.1 4.0 0.6 4.7 14 7 Monaragala HPC 83.4 12.1 95.5 3.3 0.5 3.8 13 6 Sub-total 545.0 88.3 633.3 21.8 3.5 25.3 14 37 Monitoring & Eval. 1.9 0.6 2.5 0.1 - 0.1 24 0 Total Baseline Costs 1,031.9 688.6 1,720.5 41.3 27.5 68.8 40 100 Physical Contingencies 42.6 32.9 75.6 1.7 1.3 3.0 44 4 Price Contingencies 492.2 241.5 733.7 19.7 9.7 29.4 33 43 Total Project Costs 1,566.8 963.1 2,529.9 62.7 38.5 101.2 38 147 Interest During Construction - 250.0 250.0 - 10.0 10.0 100 Total Financing Required 1,566.8 1,213.1 2,779.9 62.7 48.5 111.2 44 aJ No direct costs for the VNPC/VMPA are shown since these are included in the MPC costs. Of total project costs, an estimated SLRB 74 million (US$2.9 million) would be incurred before project effectiveness to initiate the organization of project entities, staff training, provide for the refurbishing of the old Colombo Dairy Plant in order that the initial donations of dairy comodities could be recon- stituted and combined with domestically procured milk, and carry out studies (paras 4.25 and 4.26).  -22- iTklo 3.2: PROJECT COST SUMMARY BY EXPKDITURUS SJis Nillion 15$ Million z Z Total Local Foreian Total Local Foreign Total F.. Base Costs Land 19.7 - 19.7 0.8 - 0.8 - I Civil Works 214.0 47.8 261.8 8.6 1.9 10.5 18 15 Machinery & Equipment 109.1 389.4 498.4 4.4 15.6 20.0 78 29 Vehicles 14.3 24.7 39.0 0.6 1.0 1.6 63 2 Technical Assistance 18.9 67.5 86.5 0.8 2.7 3.5 78 5 Training - 21.8 21.8 - 0.9 0.9 100 1 Technical Services 48.2 132.1 180.3 1.9 5.3 7.2 73 10 Organization & Mgt. 91.4 - 91.4 3.6 - 3.6 - 5 Support to VHPC/VXPA 221.4 5.3 226.7 8.8 0.2 9.0 2 13 Recurrent Cests 195.5 - 195.5 7.8 - 7.8 - 11 Working Capital 99 4 99.4 4.0 - 40 - - Total Base Costs 1,031.9 688.6 1,720.5 41.3 27.5 68.8 40 100 Pbysical Contin. 42.6 32.9 75.6 1.7 1.3 3.0 44 4 Price Contingencies 492.2 241.5 733.7 192_L _I9 29.4 33 43 Total Project Costs 1,566.8 963.1 2,529.9 62.7 38.5 101.2 38 147 Interest During Construction - 250.0 250.0 - 10.0 10.0 100 Total Financing Required 1,566.8 1,213.1 2,779.9 62.7 48.5 111.2 44 B. Financina 3.03 The proposed Bank loan of US$38 million would finance about 34% of total project costs, including about 552 (US$26.9 million) of foreign exchange costs and about 18S (US$11.1 million) of local costs. Monetized donated commodities are expected to finance about 51% (US$56.9 million) of project costs. EEC has indicated that it vould be prepared to donate about 13,000 tonnes of skin milk powder and 6,500 tonnes of butter oil over the life of the project subject to a year to year review of project requirenents. UFP has also indicated that it would be prepared to donate about 13,000 tonnes of skin milk powder. The Netherlands Government has expressed a willingness to provide approximately US$7.5 million equivalent which vould cover 7S of project cost. It would be a condition of effectiveness that the cofinancing and the comodity assistance agreenents have been properly executed between the Borrower and the respective parties. Financing of the remaining cost (US$8.8 million, including incremental working capital estimated at U8$5.6 million) would be through a government short-term loan to DDF of US$0.4 million equivalent, US$2.6 million in equity investments in project entities by shareholders, and US$5.8 million in funds generated from operations. The proposed financing plan is shown in Table 3.3. US$0.6 million of the total US$2.9 million pre-project costs would be financed from the Bank's Project Preparation Facility (PPF). Advances from the PPF would cover training (US$0.2 million), technical assistance (US$0.2 million), and organization and management (US$0.2 million) costs. Dutch funds and commodity  -23- assistance would finance US$1.5 million and US$0.4 million, respectively, of pre-project costs. This support for pre-project activities would, hovever, not provide DDF vith readily available cash for initial operation start-up costs. on a direct payment basis, the PPF vould finance only technical assistance and overseas training, while disbursements for local costs would be against state- ments of expenditure. The Dutch assistance would be made available as a grant package (equipment and technical assistance) and the sale of donated commodities vould require some time before funds would be generated. Until such funds are generated, DDF would need cash funding of about SLRs 10 million (US$0.4 million) as short term loan from the Government. An assurance vas obtained that Govern- ment would provide the funds required for project start-up operations under terms and conditions satisfactory to the Bank. Table 3.3: FINANCING PLAN (US$ Million) Local Foreign Total Donated Co odities 39.5 17.4 56.9 IBRD Loan 11.1 26.9 38.0 Netherlands 3.3 4.2 7.5 Short-Term Loan (GOSL) 0.4 - 0.4 Funds from Operations 5.8 5.8 Equity Investments 2.6 2.6 Total 62.7 48.5 111.2 3.04 On-lending Terms and Conditions. The Government would be the Borrower and vould repay the Bank loan over a 20-year period, including a 5-year initial grace period, vith interest during construction capitalized for the first four years, at the standard variable interest rate. GOSL would execute a Subsidiary Loan Agreement with DDF under which it would on-lend the Bank funds to DDF at the standard variable interest rate plus two percentage points for the foreign exchange risk, vith repayment over a 15-year period, including a 5-year grace period. The Government would bear the foreign exchange risk. DDF would enter into Participation Agreements with NILCO, the MPC and VMPC/V)APA providing for the on-lending of part of the funds lent to it by GOSL. It would be a condition of effectiveness that the Subsidiary Loan Agreement and the Participation Agreements between DDF and MILCO and DDF and the MPC are satisfactory to the Bank and have been properly executed. DDF would extend long-term loans to other project entities for the financing of commercial (e.g. construction of milk plants and chilling centers) as well as non-commercial (e.g. staff training, technical assistance, technical services) activities. It would also make short-term loans for working capital. On-lending of funds by DDF would be at 14% for long-term loans and 16Z for vorking capital loans. Repayment of long-term loans would be over 15 years vith a 3-year grace period on repayment of principal. At the current interest rate on IBRD loans of about 10, this would yield a 4% to 6% margin to DDF on on-lent funds. In Sri Lanka current commercial interest rates for medium- and long-term agricultural loans range  -24- from 11.5% to 16Z depending on the reputation of the client and the nature of the loan. Longer term loans provided by the National Savings Bank and the Development Finance Corporation of Ceylon are at about 17%. Under the proposed project, lending rates of 14% and 16% vould be positive in relation to medium-term inflation rates (para 3.01). On-lending rates of DDF would be reviewed semi-annually to ensure that they remain positive in real terms. An assurance to this effect vas obtained. C. Accounts and Audits 3.05 The entities established under the project would be limited liability companies incorporated under the Companies Act No. 17 of 1982. This act requires annual audit of accounts by a member of the Institute of Chartered Accountants of Sri Lanka, registered under the Companies Act to undertake audits. Each project entity would prepare separate accounts vbich would be audited annually by independent auditors acceptable to the Bank. Assurances were obtained that DDF would furnish to the Bank certified copies of its audited financial statements within nine months of the end of the financial year, and that the Participation Agreements between DDF and the other project entities (para 3.04) would spell out in a manner satisfactory to the Bank the disposition of audited financial statements of MILCO, the MPC, the VHPC and the VNPA. The audit reports would certify that funds disbursed against statements of expendi- ture (SOE) have been used for purposes of the project. 3.06 To ensure that the auditors of VMPCIVHPA accounts carry out their work in accordance with generally accepted auditing standards, and that there is a uniform approach to auditing, a Draft Questionnaire and Validation Guidance Program has been prepared and is available in the Project File. In addition, auditors appointed by the Directors of DDF to carry out its audit would also review on a regular basis the work of VMPC/VMPA auditors. Also, DDF and MILCO vould make joint appointments to an Audit Committee which would act as a sub-committee of their Boards of Directors. This Audit Committee would consist of serving Directors with financial background, Chief Financial Officers (one of whom would act as Secretary), and independent auditors from professional accounting firms. The purpose of the Audit Committee would be to review the overall progress of audits and to gain further insights into the financial positions of the varluus project entities. Assurance was obtained that an Audit Committee with composition and terms of reference satisfactory to the Bank would be formed within six months following loan effectiveness. IV. PROJECT ORGAJIZATON AND IMPLEMENTATION A. Essential Features of Corporate Entities General 4.01 The corporate entities which would be vested with the responsibility of implementing the project are: (a) Dairy Development Foundation (Guarantee) Limited (DDF);  -25- (b) Milk Industries of LankA Company Limited (MILCO); (c) Milkshed Producers' Companies Limited (HPC); and (d) Village Milk Producers' Companies (Private) Limited (VNPC). 4.02 The charts in Annex II summarize the functions and financial structures of these entities. In the Project File, Draft Memoranda and Articles of Association are given. The ViPA participating in the project would have equal status with the VMPC. Dairy Development Foundation (Guarantee) Limited (DDF) 4.03 General. DDF would be formed to spearhead the development of the dairy industry of Sri Lanka. It would be the focal point for implementation, and all funding would flow through it. DDF would provide all initial funding for the establishment of the other project entities, including long-term loans and working capital. Machinery and equipment as well as technical inputs (e.g. semen) for the project would be passed on to the implementing entities at full cost to be repaid through long-term loans. DDF would be the recipient of com- modity assistance and would sell the skim milk powder and butter oil to MILCO at market price. The funds received would be used to help finance the project. IBRD funds would pass from GOSL to DDF for on-lending to MILCO, NPC and VMPCIVMPA. Funds from the Netherlands Government would also pass through DDF. 4.04 Objectives. As stated in its Memorandua of Association, the primary objectives of DDF are: (a) "To promote by all such means as are available to the Company the development throughout Sri Lanka of the dairy industry in all its aspects and to achieve the maxim beinefit to the people of Sri Lanks therefrom including attainment of increase in production of milk and all other related dairy products in its several forms; and (b) To assist and aid the policy of the Government of Sri Lanka in develop- ing the dairy industry." 4.05 Income and Turnover Taxes and Customs Duties. The Company has been classified as being formed for charitable purposes and therefore would be exempt from income tax under Section 8 (d) of the Inland Revenue Act No. 28 of 1979. The objectives and functions of DDF have been discussed vith the Couissioner of Inland Revenues and the Company meets the basic criterion for exemption from income tax. DDF would also be exempted from turnover tax and from customs duties on imported dairy commodities under Section 19 (a) of the Customs Ordinance. 4.06 Profits and Dividends. No profits and/or dividends would be declared and paid at any time during the existence of the Company (para 2.09 (a)). The assets at the time of liquidation vould be realized and the cash so derived donated to an approved charity as per the Companies Act No. 17 of 1982. Milk Industries of Lanka Company Limited (MILCO) 4.07 Functions. The functions of the Company would be:  -26- (a) The purchase, processing and marketing of fluid milk and dairy products. It would set the pricing from producer to retail level (para 5.02); (b) Reconstitution or recombination of donated and commercially procured milk products; (c) Expansion of the capacity and continued modernization of the milk processing industry; and (d) Provision of professional management under contract vith concerned organizations to undertake the management of dairy plants, other milk processing facilities, and any other entities in the dairy sector or its related subsectors, such as feedmills. These functions would include the leasing and operation of existing dairy plants now owned by NMB, the leasing of IMB chilling centers, and the construction of new chilling centers. In the initial years, before the new dairy plants become operational, recombination of donated skim milk powder and butter oil vould be done at the refurbished Colombo plant. Commodities received before this plant is ready vould be repackaged under contract and sold directly. 4.08 Share Capital. The company would be a public limited liability company whose shares would be quoted by the Colombo Brokers Association and thus be freely marketable. Fifty one percent of the share capital would be offered for subscription to the public (public issue by prospectus) and 49% vould be pur- chased by MPC as they generate funds for investment. The share issue would be underwritten by DDF. As stated in MILCO's Articles of Association, voting rights of shareholders are proportional to the shares held; in the case of the MPC, the shares held vould be in proportion to the value of the milk supplied to MILCO during the previous year. Dividends vould be distributed on the basis of share ownership, and an assurance was obtained that dividends would not be paid until MIlCO-s debt:equity ratio had declined to 60:40. NILCO would have a nine-member board consisting of three directors appointed by DDF, three MPC representatives selected by lot from among the five elected MPC chairmen, and three elected public shareholders. During the first year of operation, the three MPC representatives would be appointed by Government. Prior to the issuing of shares, and thus the election of public shareholders, the functions of the board would be carried out by Government appointed company promoters consisting of three DDF directors and four representatives from the private sector. 4.09 Income and Turnover Taxes. Under the provisions of the Inland Revenue Amending Act No. 14, Section 17 (A), of 1984, dairy development companies have been granted a tax holiday for five years. Thus, MILCO vould not be liable for normal taxes of 50Z of taxable income for this period. As it vould be a quoted public company, withholding tax vould also not be liable on dividends dis- tributed to shareholders. MILCO would be required to pay turnover tax at 1%. Milkshed Producers' Companies Limited (MPC) 4.10 Functions. Five MPC vould be established to cover all districts of Sri Lanka. These companies would:  -27- (a) Lease from MILCO and manage NMB chilling centers as well as new chilling centers constructed by MILCO; (b) Collect milk on behalf of MILCO from the VNPCIVNPA and transport it to chilling centers and/or dairy plants; (c) Provide technical and extension services to milk producers; and (d) Engage in the training of VMPC/VMPA staff and farmers. The MPC would also invest in MILCO on behalf of the VMPC/VMPA. Eventually the HPC would hold 49% of the outstanding share capital in MILCO. Since the MPC would be owned by the VMPC/VMPA in their respective milksheds, the 49% share represents producer holdings in the processing company. 4.11 Share Capital. The VMPC/VMPA would be encouraged to invest in their MPC in proportion to their shares of milk sold to the MPC. Voting rights of VMPC/VMPA would be in proportion to the value of milk sold to the MPC in the previous year. 4.12 Distribution of Profits. In the last month of each financial year, the Boards of the Companies would ascertain whether surplus funds had been gener- ated. Any such funds would be distributed on the basis of share ownership by the V1PC/VMPA. 4.13 Income and Turnover Taxes. As the Companies would be fully producer-owned, they would offer services to VMPC/VMPA members at cost so as to minimize their income and imposition of taxes. As dairy development companies, they would be exempt from taxes for a period of five years (para 4.09). After that period, any profits would be subject to income tax at 50Z. Turnover tax would also be payable at 1Z. Village Milk Producers' Companies (Private) Limited (VMPC) 4.14 Functions. The functions of these companies would be: (a) Collecting, testing, and arranging payment for milk from individual producers; (b) Providing technical services (artificial insemination and para-veterinary services) to their members; (c) Disseminating to their members extension information on fodder and pasture production, animal management and fceding, calf rearing, and milking hygiene; and (d) Investing on behalf of their members in their respective HPC. The VMPA would perform the same functions for their members as the VNPC. While the VMPC/'MPA would sell most of their milk to MILC0, a small quantity (about 5% of the total) would be retailed locally by the village units. 4.15 Share Capital. These companies would be incorporated as private com- panies. The share capital of a VNPC would be SLRs 500,000 divided into 50 Class  -28- A shares of SLRs 10 each and 49,950 Class B shares of SLRa 10 each. The issued share capital of a VMPC would at no time exceed SLRs 500,000 in order that such companies could qualify as small companies under the Internal Revenue Act. The Spearhead Teams vould organize the registration of VHPC according to a plan agreed between the MPC and DDF. This organizational work of the MPC would be done under contract with DDF, The contract would make the MPC responsible for ensuring that each VMPC would: (a) Have not less than seven members; (b) Have only one member per family; (c) Have as members only those who own milk producing cattle or buffaloes; (d) Purchase from each member only milk produced by the member; (e) Sell to each member only one Class A share carrying voting rights; (f) Sell to all members such Class B shares as they may wish, as and when these are issued; (g) Issue Class B shares and so organize its finances as to permit it to purchase shares in its MPC in proportion to the total milk sold to the MPC; and (b) Inform the MPC when the number of members plus qualifying applicants for membership approaches 50 so that such companies could be converted into public companies by passing an appropriate resolution and filing a statement in lieu of a prospectus. As long as the VMPC remain private companies, the right to transfer shares would be restricted so that no Class A share would be transferred to a household which already held a Class A share in the Company. Class B shares would, however, be freely transferable among producers supplying milk to the VNPC. In the last month of each financial year, the VNPC boards would ascertain whether surplus funds had been generated. 50Z of any such funds would be distributed among stockholders on the basis of share ownership. The balance of 50Z would be dist=ibuted on the basis of the value of milk supplied to the VMPC by each member. 4.16 Income and Turnover Taxes. The companies would be exempt from taxes for a period of five years under current law (para 4.09). Since the VMPC would be fully owned by producers, they would deliver services to producers at or near cost so as to minimize income, and thus tax obligations, when exemptions would no longer be applicable. 4.17 After the tax holiday, income tax would be levied on the taxable profits of these companies at the following rates:  -29- Up to SLRs 50,000 of taxable income 20% Next SLRs 100,000 30% Next SLRs 100,000 40% Balance 50% The companies would be liable to a turnover tax of 12. 4.18 DDF would enter into Participation Agreements with MILCO, the MPC, and the VMPC/VMPA providing for the onlending of funds by DDF (para 3.04). The Agreements between DDF, MILCO, and the MPC would also set forth their under- standings and responsibilities for project implementation. The MPC would enter into similar agreements with the VMPC/VMPA in which the latter would agree to abide by the rules and regulations governing milk procurement and the provision of technical services at the village level. Assurance was obtained that Par- ticipation Agreements, satisfactory to the Bank, would be exchanged between the MPC and the VMPC/VMPA. B. Inter-Ministerial Coordination 4.19 Steerinz Committee. As dairying is influenced by developments and policies in several ministries, a Steering Committee vould be established which would coordinate inter-ministerial activities and policies and serve as a deci- sion making body on policy matters affecting the dairy industry. This Steering Committee, as a sub-committee of the Committee of Development Secretaries, would formulate and recommend to the Cabinet, through the concerned ministry, policies and directives for Sri Lanka's dairy development. The Committee would not set milk prices (para 5.02). It would meet at least once every three months and NRID would provide the secretariat. The Committee would consist of the follow- ing members: Chairman: Minister, Ministry of Rural Industrial Development Convenor: Secretary, Ministry of Rural Industrial Development Members: Secretary, Ministry of Rural Industrial Development Secretary, Ministry of Finance and Planning Secretary, Ministry of Agricultural Development and Research Secretarv, Ministry of Lands and Land Development Secretary, Ministry of Trade and Shipping Secretary, Ministry of Mahaweli Development Secretary, Ministry of Food and Cooperation Secretary, Ministry of Local Government, Housing and Construction Chairman, Agricultural Development Authority Chairman or Chief Executive, DDF Chairman or Chief Executive, MILCO Assurance was obtained that the Steering Committee with the above composition and with terms of reference satisfactory to the Bank would be formed within three months after loan effectiveness.  -30- C. Procurement and Disbursement 4.20 Procurement. Proposed procurement arrangements are summarized in Table 4.1. Table 4.1: PROCUREMENT (US$ Million) Total Proiect Element ICB LCB Other N.A. Cost Land 1.0 1.0 Civil Works 14.7 14.7 (10.3) (10.3) Macbinery, Equipment, Vehicles 16.8 23.1 3.0 42.9 and Materials (13.4) (0.5) (13.9) Consultant's Services, Training, 9.5 9.5 Fellowships and Study Tours (2.5) (2.5) Staff Salaries and Operating 27.5 27.5 Expenses (1.3) (1.3) Working Capital 5.6 5.6 IBRD Financing Charges 10.0 10.0 (10.0) (10.0) Total Project Costs 16.8 37.8 13.5 43.1 111.2 (IBRD Financing) (13.4) (10.8) (2.5) (11.3) (38.0) aJ A/ Figures in parentheses are the respective amounts to be financed from the IBRD loan. 4.21 Procurement of goods and services financed by the Bank would be carried out by DDF. MILCO would procure the land for the construction of dairy plants and chilling centers in accordance with established Goverment procedures acceptable to the Bank. Civil work contracts (US$14.7 million) for construction of dairy plants, chilling centers, training centers, staff quarters, etc., would be individually small and dispersed in time and place. Thus, they would be unsuitable for ICB and would be contracted through LCB under procedures accept- able to the Bank. Major items of machinery, equipment and vehicles (US$16.8 million) would be bulked and procured by ICB in accordance with Bank guidelines. Minor equipment (US$0.5 million) costing less than US$15,000 per contract that could not be bulked into packages suitable for ICB, or which vould be required urgently for project implemerntation, would be procured after  -31- solicitation of quotations from at least three independent suppliers in accord- ance with procedures satisfactory to the Bank. Machinery, equipment and materials procured by Dutch aid (US$3 million) would be on a tied basis from Dutch suppliers. Bids for imported frozen semen (US$6 million) would be sought from at least three countries, to be selected in agreement with the Bank, wbere bulls of the required types are readily available. Hiring of consultants (US$4.2 million) for assistance in project implementation would be in accordance with Bank guidelines. Bids for civil works contracts costing over US$185,000 equivalent and equipment contracts over US$50,000 would be subject to prior Bank review. Other contracts would be subject to selective post award review. 4.22 Disbursements. Disbursements under the loan would cover: (a) Civil works - 90% (US$9 million); (b) Machinery, equipment, materials (including laboratory and veterinary equipment, chemicals, veterinary pharmaceuticals and semen) and vehicles (US$12 million): ti) directly imported - 100% of foreign expenditures; Cii) locally manufactured - 100% of local expenditures, ex-factory; Ciii) locally procured (off-the-shelf) - 65% of local costs; Cc) Consultants' services, training, fellowships and study tours - 100% (US$2 million); ud) Incremental staff salaries and operating costs (US$1 million) on a declining basis: Years 1 and 2 - 95% and Year 3 - 80%; Ce) Loan interest during construction (US$10 million); (f) Refunding of project preparation advance (US$0.6 million); and (g) Unallocated (US$3.4 million). Disbursements for small civil works contracts (i.e., less than SLRs 500,000) and for items under (d) above would be made against statements of expenditures certified by the Chief Executive of the executing entity. Documentation sup- porting the statements, including receipts and project records, would be retained in the relevant project office and wade available for review by Bank staff on request. Assurances were obtained that all documents supporting SOE would be audited annually by independent auditors acceptable to the Bank (para 3.06). Disbursements for civil works contracts for SLRs 500,000 or more and for items under (b) and (c) above would be documented in accordance with the Bank's standard requirements. Reimbursement applications would be submitted on a monthly basis or when such expenditures reach the equivalent of US$20,000. The withdrawal applications would be prepared by the appropriate executing entity and then forvarded to DDF for countersigning prior to submission to the Bank. Advances drawn from the Project Preparation Facility (para 3.03), with the service charges accrued thereon, would be repaid out of the proceeds of the proposed loan as soon as it is declared effective.  -32- 4.23 Proiect Special Account. To ensure that adequate funds are available for the prefinancing of small contracts and other project expenditrres and to facilitate the project's timely implementation, a Project Special Account in US dollars would be established in the Central Bank of Ceylon and operated under procedures acceptable to the Bank. Only one Special Account would be opened and only the Bank's share of project costs would be deposited into and disbursed out of the Account. The initial deposit to be made by the Bank in the Account would be US$1 million, estimated as the equivalent of four months' average expendi- ture. The Account would be operated by the Chief Executive of DDF, or his designate. 4.24 Approximately 80Z of total project expenditures are expected to be incurred by project Year 5. Since Government would provide no project financing and since funds generated by the sale of donated dairy commodities would not be immediately available, the flow of Bank funds to DDF would be accelerated in the early project years, and loan disbursements are expected to be completed by June 30, 1990, i.e. at a considerably faster rate than that indicated by the South Asia 10-year average disbursement profile ('nnex VI). However, total project expenditures would be spread over 7.5 years and would thus resemble rather closely the average profile. The 7.5-year period is in line with experience with similar type dairy projects in India. D. Pre-Investment Activities 4.25 To ensure more rapid implementation of the project, the Government has initiated a number of activities, some of which have been completed. These include: (a) Establishment of an inter-ministry Action Committee c'aired by the Minister of MRID and including representatives of the Ministries of Trade and Shipping and Finance and Planning; (b) Registration, under the Companies Act, of DDF, MILCO, and the MPC, and the appointment of the initial Boards of Directors (i. :luding the Chair- men) of these companies; (c) Acquisition of letters of intent from donors to participate in the project; and (d) Completion of arrangements for the initial staffing of DDF, MILCO and the MPC, including the organizational structure and job descriptions of key personnel. 4.26 Other pre-project activities currently underway include the folloving: (a) Recruitment and training overseas (India) of the first members of the Spearhead Teams and about 15 trainers who would be responsible for training local project staff and farmers in-country; (b) Initiation of the refurbishment of the Colombo dairy to enable it to recombine the donated dairy commodities; and  -33- (c) The carrying out of a market survey and z -dies in the field of animal feeding, animal breeding, and farmer training and extension. B. Monitorim and Evaluation 4.27 Given the innovative nature of the project and the substantial institu- tional and infrastructural development required, an essential feature of project organization would be a strong monitoring and evaluation system which would provide guidance for project implemertation and permit a realistic assessment of project progress. Two independent units, an existing unit within MRID, which would be strengthened under the project, and one to be established within DDF, would be responsible for these activities (para 2.18). DDE would organize and coordinate the collection of data which would be used to prepare regular progress reports, compare project progress with appraisal targets, and keep project management informed of implementation problems. Semi-annual progress reports would be prepared in accordance with a format to be agreed with IBRD. The MRID unit would prepare annual project evaluation reports for Bank's review based on independently collected data, as well as data provided by DDF, and would be responsible for preparation of the Project Completion Report. Assuran- ces to this effect were obtained. A mid-term review of project activities would be carried out by IBRD in cooperation with DDE, MRID and the other project financiers at the completion of the third project year to ensure necessary adjustments required for the attainment of project obiectives (para 2.07). V. MARKETING, PRICES. TARIFFS AND SUBSIDIES A. Narketing 5.01 The main marketing strategy of the proposed project would be to increase the consumption of fluid milk in both urban and rural markets. However, in the marketing of fluid milk in rural areas, the cheapest and simplest processing system-pasteurization--does not give the product sufficient shelf-life to be distributed by the wholesale-retail network. In addition, the present milk marketing system does not provide the consumer with fluid milk of satisfactory quality, convenience, and price, and, consequently, about one half of the milk presently consumed is purchased in the form of powder. MILCO's basic marketing aim would be to distribute pasteurized milk in the larger urban markets closer to the dairy processing plants, and supply long-life (UHT) milk to more distant markets. The private distribution system which presently handles packaged foods is both enterprising and efficient and would be used to retail fluid milk and milk products produced under the project. B. Prices 5.02 Under the project, MILCO would be granted autonomy in milk pricing from producer to retail level. The ceiling on consumer prices would be determined by imported dairy commodities. Differences between prices paid to producers in different regions of the country would be based on differences in the marginal costs of milk collection and transport and not on differences in the regional cost of milk production as this would tend to discourage production in low cost  -34- areas and encourage it in relatively higher cost areas. The pricing mechanism should be based on carrying back through the marketing, processing, collection and transport chain the retail price deducting at each stage the actual ma-ginal costs incurred. Retail prices would be based on sound economic principles applied to milk processing and marketing. While Government would always have the prerogative to intervene in extraordinary situations, an assurance was obtained that MILCO would be granted autonomy in milk pricing. one of the objectives of the project (para 2.06 (b)) is to increase the producer's share of consumer payments for Processed milk to at least 60Z through more efficient collecting, processing, and marketing. The financial prices used are given in Annex IV, Table 8. Annex V, Table 2, gives the economic prices used. C. Tariffs 5.03 With constant domestic milk production, it is estimated that dairy imports would increase from about 450,000 lpd in milk equivalents to about 730,000 lpd by 1991, and to nearly 1 million lpd by 1996. An incremental milk production of about 300,000 lpd by 1991 would be required to stabilize imports at current levels. The proposed project would result in an incremental produc- tion of about 80,000 lpd by 1991 and about 200,000 lpd by 1995. Thus, even with the project, commercial dairy imports are not expected to fall from present levels, and, in fact, may increase. 5.04 The project would promote the development of a more efficient dairy industry in Sri Lanka through the introduction of major institutional and organizational schemes (paras 2.02 to 2.06). However, improved efficiency in the production, collection, processing, and marketing of milk would not be apparent until these changes are in place and the industry is adequately staffed with trained personnel. Furthermore, since dairying is a long-term growth activity, reasonable stability in domestic milk prices is required, especially during the development period, in order to sustain a steady rate of investment. The dairy industry would, therefore, continue to need some protection from cheap milk imports dulring the life of the project. To provide this protection and stability, the Sri Lanka Dairy Sector Review (April 1983) suggested a variable tariff on dairy imports. The level of tariff would need to be set in the light of actual trends in international prices, as well as local milk production costs. The Steering Committee established under the project (para 4.19) would periodically review the price/cost data of locally produced and imported milk and recommend appropriate tariff levels to the Tariff Commission. To minimize the risk of misallocation of resources and consumer welfare loss, the Goverument would seek the concurrence of the Bank before raising the tariff above 25%. Assurance to this effect was obtained. D. Subsidies 5.05 The project seeks continued and sustained operational and product line improvements in Sri Lanka's milk processing industry, and the Bank would thus seek the removal of subsidies presently paid to NMB (para 1.10). Accordingly, it would be a condition of effectiveness that the SLRs 0.50 per liter subsidy on processed fluid milk had been removed. It is expected that the subsidy would be passed entirely to the consumer raising the retail price of fluid milk to SLRs 8.50 per liter. This should not unduly impair fluid milk's competitiveness  -35- since the existing price of imported whole milk powder is SLRs 8.90 per liter equivalent. Furthermore, assurance was obtained that the present subsidy of SLRs 1.0 per liter of milk used for the domestic manufacture of whole milk powder would be phased out within two years of loan effectiveness. Under the project, MILCO would take over the operation of NMB's existing milk powder factory and, through diversification into more profitable products, eliminate the basis of the subsidy on fluid milk processed locally into whole milk powder. VI. FINANCIAL VIABILITY 6.01 Detailed financial projections for all project entities are available in the Project File. For the financial analysis the detailed projections were consolidated and summarized in order to highlight critical areas. A. Dairy Development Foundation 6.02 All funds for the project would be channelled through DDF. External funds in the form of a grant, monetized commodities and the IBRD loan would form the basis of financing for the project. In addition, DDF operations would generate significant funds over the project period. The flow of funds and DDF's financial projections are summarized in Table 6.1.  -36- Table 6.1: FLOW OF FUr1 S AND SUMKARY FINANCIAL STATEMENIS FOR DDF (Millions of Current SLRs) Years 0 - 7 /a 8 - 10 Sources and Arnlication of Funds Source Operations (Net After Debt Service) 2.3 (130.5) Grant 187.5 - Monetized Commodities 1,422.7 - IBRD Loan 700.0 Short-term Loan 10.0 Total 2,322.5 (130.5) ADiDlication Long-Term Loan to MILCO 968.2 - Contracted Services via MNILCO 208.1 - MPC and VMPC Development 922.8 - DDF Equipment 12.9 - Total 2,112.0 - Net Cash Flow 210.5 (130.5) Income (Cumulative) Revenue 505.7 226.5 Expenses 664.5 391.1 Net Income (158.8) (164.6) Balance Sheet (End of Period) DDF Assets 1,110.5 649.5 Investment in Other Entities 1.130.9 1.130.9 Total Assets 2,241.4 1,780.4 Long-Term Debt (ThRD) 790.4 494.0 Equity 1.451.0 1O286.4 Total 2,241.4 1,780.4 /a Years 0-7 represent the project implementation period. DDF's main operational concern would be to ensure that sufficient funds are available to provide the necessary inputs to project operating units during implementation. More detailed financial projections (Annex IV, Table 1) indi- cate that net fund flows for DDF would be adequate. From Year 4 on, DDF vould have generated sufficient internal cash flow to cover operational requirements and could begin retiring its debt. However, it has been assumed that the long-term debt would have a 5-year grace period in order to provide some allowance for project implementation slippage. Debt servicing would start in Year 6. While DDY s cumulative profits are not positive, its cumulative cash  -37- flow is. At the end of Year 10, DDF vould have sufficient net assets to retire its remaining debt. 6.03 DDF would underwrite the share issue of HILCO, and thus it would have a contingent liability in the early years of about US$2 million for the equity inputs to NILCO. This liability would reduce as MILCO stock is taken up by the public and the VMPC/VMPA. The project envisages that 512 of MILCO stock would be held by the public and 49Z 'y the VMPC/VNPA. B. Milk Industries of Lanka ComDany 6.04 Detailed financial projections for IMILCO are given in Annex IV, Table 2, and are summarized in Table 6.2. Table 6.2: SUMMARY OF FINANCIAL STATEMENTS FOR MILOO (Millions of Current SLRs) Years 0 - 7 8 - 10 income (Cumulative) Revenue 12,401.9 11,466.2 Expenses 11i976.7 11.186.4 Net Income 425.2 aJ 279.6 Net Cash Flow 858.8 565.7 Balance Sheet (End of Period) Current Assets 617.6 901.7 Fixed Assets 710.0 424.1 Total 1,327.6 1,325.8 Current Liabilities 80.3 80.3 Long-Term Liabilities 584.0 343.1 Equity 663.4 902.4 Total 1,327.7 1,325.8 Statistics Internal Rate of Return (Z) 25.9 (overall) - Gross Profit vs Revenue (average 2) 12.5 9.4 Net Profit vs Revenue (average 1) 3.4 aJ 2.3 Debt Service Coverage (end of period) 1.8 1.8 Debt vs Equity (end of period) 0.5 0.3 a/ Exempt from income tax in Years 1-5.  -38- These projections indicate that NILCO from its incorporation would be a viable and profitable operation. Sales revenues are projected to increase substan- tially over the period from a level of about SLRs 800 million in Year 1 to over SLRB 4,400 million by Year 10. Profitability is relatively consistent over the period with gross profit averaging IlZ of sales and net profit after tax about 3Z. Net cash flows are positive for all operating years and after dividend payments and allowances for working capital would accumulate to about SLRs 902 million by Year 10. The financial rate of return (FRR) is estimated to be about 26Z. MILCO's total assets would increase over the period to SLRs 1,326 million- and owner's equity to SLRs 902 million in current terms. Debt to equity ratio would average about 60:40 falling to about 30:70 by Year 10 as debt is retired. Debt service coverage which averages about 1.4 from Year 4 on is satisfactory. 6.05 The FRR, while at an acceptable level of about 26Z, is very sensitive to changes in benefits or costs. A 5% negative change in either stream, without a corresponding change in the other, would cause the FRR to fall below 10%. This indicates that MILCO would have to pass most of any market price reductions on to the NPC and VMPC/VNPA. C. Milkshed Producers' Companies 6.06 These companies would be incorporated to provide services to the VMPC and earn sufficient revenues to cover cash outlays. Consolidated revenue for the EPC is projected to increase from SLRs 96 million in Year 1 to about SLRs 600 million in Year 10. Consolidated statements (Annex IV, Table 3) for the MPC indicate that estimated net cash flow would be positive. Cash flow from operations is expected to accumulate to SLRs 150 million over the first 10 years. Generated funds from DDF (2% of the resale value of donated commodities) would be passed on as a grant to the MPC for the purchase of shares in MILCO on behalf of the VWPC/VMPA. It is estimated that this transfer would allow the MPC to purchase about SLRs 27.5 million in MILCO equity.  -39- Table 6.3: SUMMARY FINANCIAI STATEMENTS FOR THE HPC (Millions of Current SLRs) Years 0-7 8-10 Revenue (Cumulative) 1,442.3 1,506.3 Operating Costs (Cumulative) 1,473.4 1,491.3 Net Operating Income (Cumulative) (31.1) 15.0 Net Cash Flow (Cumulative) 134.2 16.2 Balance Sheet (End of Period) Net Current Assets 134.2 150.4 Net Fixed Assets 156.2 76.6 Investment in MILCO 27.5 27.5 Total Assets 317.9 254.5 DDF Investment 347.4 347.4 Equity Investment 1.5 2.5 Retained Earnings (31.0) (95.4) Total 317.9 254.5 D. VillaRe Milk Producers' Companies 6.07 Detailed operating statements have been prepared for the VMPC and are contained in the project files. Significant growth in VMPC formation is projected with 220 in Year 1 increasing to 1,950 by Year 7. On a consolidated basis, revenues are projected to increase from a total of SLRs 89 million in Year I to SLRs 1,518 million by Year 7 (in current terms). Net cash flow from operations is expected to be positive for all VMPC from Year 1 onward. Total net cash flow is expected to total SLRs 1.6 million in Year 1, increasing to SLRs 28.5 million by Year 7. A summary of the consolidated income statement for the VMPC is presented in Table 6.4.  -40- Table 6.4: SUMMARY INCOME STATKHENT FOR THE VHPC (Millions of Current SLRs) Year 1 2 3 4 5 6 7 Revenue 88.9 237.7 458.5 710.7 917.1 1,209.1 1,518.5 Operating Costs 87.4 234.2 451.2 699.9 906.0 1.194.5 1.500.1 Net Profit 1.5 3.5 7.3 10.8 11.1 14.6 18.4 Net Cash Flow from Operations 1.6 4.0 9.1 14.5 17.2 22.7 28.5 Number of VNPC (Cumulative) 220 520 860 1,190 1,480 1,740 1,950 VII. PROJECT JUSTIFICATION AND RISKS A. Milk Production and Procurement 7.01 As summarized in Table 7.1, by Year 7, 1,950 VMPC are expected to be formed with a total membership of about 180,000. Incremental milk production is estimated at 106,000 lpd by Year 7, and about 208,000 lpd by Year 10. Increased milk production would result from increases in milk yields as a conse- quence of improved feeding and management and the partial conversion of the national herd from low-yielding indigenous animals to higher-yielding crossbreds, and modest growth in the herd due to increased calving rates and reduced mortalities. Milk procurement is projected to expand by about 173,000 lpd by Year 7 and 275,000 lpd by Year 10. Milk procurement under the project is expected to rise from the present level of about 27% to 46% of total production. This increase appears feasible given the village-based procurement system envisaged under the project. The difference in incremental production and procurement represents the gains in procurement from collecting milk in pre- viously untapped areas. Milk now produced in these areas is either consumed at home or is sold locally (para 1.07). 7.02 In a typical VMPC, the number of milk animals would increase by about 10% by Year 7 of the project, and by about 16% by Year 10. The composition of the VMPC herd would also change. While the percentage of buffaloes would remain fairly constant at about 20%, the proportion of indigenous cattle would drop from a present level of about 65% to 58% by Year 7, and 42% by Year 10. Crossbreds, on the other hand, vould increase from about 13Z of the total at present to 20% and 35%, respectively, by Years 7 and 10. Average daily milk yield per animal would also rise, and by Year 7 of the project it is expected to be 25% higher than the present yield (0.9 lpd) and by Year 10 about 60% higher.  -41- Table 7.1: VMPC AND INCREMENTAL MILK PRODUCTION AND PROCUREMENT Year 0 1 2 3 4 5 6 7 8 9 10 VMPC Formed - 220 520 860 1,190 1,480 1,740 1,950 1,950 1,950 1,950 VNPC Membership ('000) - 9.5 26.5 51.0 80.0 90.9 146.4 179.6 179.6 179.6 179.6 Incremental Milk Production ('000 lpd) - - 13.5 28.2 43.8 61.7 80.8 105.8 131.7 165.5 207.5 lucremental Milk Procurement ('000 lpd) - 27.1 39.5 62.5 87.6 112.5 138.3 173.3 201.2 234.7 275.4 B. Proiect Viability and Economic Returns 7.03 Analysis of the project's viability is based on three benefit streams: Ca) Benefits from the incremental procurement of milk and its movement from rural areas where it commands a low value to urban areas vhere it fetches a higher value; Cb) Benefits of incremental home consumption of milk; and (c) Benefits from incremental by-products, including sales of manure and stock. Incremental milk production has been estimated through the use of herd models for each milkshed which take into account increases in milk yield and growth of the herd due to increased calving rates, reduced mortality, and the partial conversion of the herd from low yielding indigenous animals to higher yielding crossbreds. Incremental milk procurement has been estimated on the basis of expansion in the number of VMPC, the increase in the number of producers supply- ing milk to VMPC, and the incremental milk production occurring in the VMPC area of influence. Home consumption of milk was taken as production net of procure- ment. In calculating benefits, the value of incremental procurement of milk was estimated from the increased projected sales of milk and milk products by the project's dairy plants. Economic prices for fluid milk were based on the Decem- ber 1984 CIF cost of imported whole milk powder adjusted for handling, process- ing and packaging costs. For milk products, current local prices were adjusted by conversion factors to estimate equivalent border prices.. Economic prices used are shown in Annex V, Table 2. A fertilizer conversion factor was applied in estimating the value of manure.  -42- 7.04 Investment costs include costs of primary producers as vell as costs of all project entities, i.e., VMPC, MPC, MILCO, and DDF. In the case of primary producers, these are based on incremental capital costs per liter of milk tines year-to-year incremental milk production of the first 10 years of the project, assuming replacements at the same levels for subsequent 10-year periods. Project entities' investment costs cover civil works, machinery and equipment, vehicles, technical assistance, training, technical services, and organization and management. 7.05 Farm models for the different milksheds and types of animals were employed to estimate operating costs under the existing technology of milk production and under improved technology. In the benefit streams, it is assumed that the new technology is fully adopted by a producer five years after he becomes a supplier of milk to a VNPC. Operating costs include costs for primary producers, VMPC, MPC, and MILCO. Producer operating costs are based on per liter production costs estimated for different milksheds under both existing and new technology conditions. The production costs were converted to economic values by a standard conversion factor of 0.9. Family and hired labor were shadow priced at 80% and 90Z of market values, respectively, to reflect current levels of unemployment. 7.06 Computations based on the above premises yielded an economic rate of return of about 23% (Annex V, Table 1). Sensitivity tests indicated that the project is more resilient to increases in investment and operating costs than to decreased sales revenues stemming from a decrease in price. A second base run was made on the assumption that the skim milk powder and butter oil donated for the purposes of the project would be a tied resource; that is, they would not be fungible nor forthcoming without the project, hence, their value is treated as an economic benefit. Under this assumption, the estimated economic rate of return is about 46%. Economic rates of return are summarized in Table 7.2. Table 7.2: ECONOMIC RATE OF RETURN Parameter Percentage Base Run - Project Assumptions 23 Sensitivity Tests Investment Costs Increased 20% 19 Operating Costs (of Processing and Marketing) Increased 20% 17 Revenues from Plant Sales (Proxy for Price Changes) Decreased 20% 11 Combined Effects of Investment Costs and Revenues from Plant Sales 8 Base Run with Donated Commodities Included in Project Benefits as a Tied Resource 46  -43- C. Other Impacts 7.07 The main institutional changes brought about by the project and the establishment of a vertically intergrated organizational structure with producer ownership in the entities' responsible for the collection, processing, and marketing of milk would: (a) Help ensure that the technical inputs necessary for increasing milk output and improving the efficiency of production are provided; (b) Allow producers to participate in the profits obtainable in milk processing and marketing; (c) Provide a mechanism by which incentive prices can reach the farmer; and (d) Help ensure that consumers have increased access to high-quality milk and milk products at fair prices. 7.08 Incremental on-farm employment generated during the first seven years of the project is estimated at 25,000 man years distributed among about 180,000 households. Additional employment along with management opportunities vould also be created through the establishment of the various project entities. In terms of gross income, it is estimated that in year 7 project participants would receive about SLRI 750 million as income from the sale of milk. With a average return to labor of about SLRs 3.40 per liter, this is anticipated to yield an increase in net income by year 7 of about SLRs 265 million, or about SLRs 1,470 per household. 7.09 As cattle and buffaloes are more evenly distributed the land, with even those too poor to own cultivable land often keeping animals, the incremental income from milk among rural households would tend to be more evenly distributed than with developmert projects involving crop production. With dairying expected to expand into the Dry Zone, vhere incomes and opportunities for alter- native production activities are particularly low, regional distributions of income would also improve. 7.10 As milk is an expensive source of both calories and protein, the low allocation of household expenditures to milk and milk products by low income households indicates a rational utilization of income since equally nutritious sources of these nutrients can be purchased at lower cost in other foods, thus resulting in a positive contribution to overall nutrition. Furthermore, in rural areas, dairying offers farmers the opportunity to gain good returns steadily over many months of the year. 7.11 Draft power is a major contribution of the livestock sub-sector in Sri Lanka; however, it is difficult to effectively tackle the problems of draft animals directly. On the other hand, it is considered that draft animals would benefit from the development of the dairy industry, and their general improve- ment through better feeding and veterinary care and through crossbreeding would be an important side benefit of dairy development.  -44- D. Proiect Risks 7.12 The proposed project would require substantial institutional change and like all such projects would have inherent risks. Change, while welcomed by many, is resisted by those with vested interests in the status-quo, despite the inefficiencies and the costs it often imposes on others and on a country's development. To minimize this risk, any ESB staff taken on by the new project entities would be rigorously screened by the entity concerned before employment. Experience vith many Bank projects has indicated that directing investments through inefficient organizations also has inherent risks and often results in projects that fail to fulfill their objectives. Such was the case with the Bank's involvement in the first Sri Lanka Dairy Project (Cr. 504-CE) where the institution (UHB) directly involved with the project was overstaffed, ineffi- cient, and heavily subsidized. As a consequence, and despite investments in technical inputs and technical assistance by the Bank and other donors, the development impact through increased milk production could not be realized as the institutional and pricing structure inhibited producers from adopting new practices. GOSL and the Bank both realize that such inefficiencies must be alleviated through institutional change despite the difficulties and risks involved in such change. The Government's recent pricing reforms (para 1.10) indicate their commitment to putting economic rationale into the sector. The proposed project also has the support of many in the private sector who would like to see the dairy industry achieve its potential through folloving commer- cial principles and the building of an efficient processing and marketing Sys- tem. 7.13 The project carries a financial risk. Monetized donated commodities are expected to finance about 51Z of total project costs. Both the EEC and WFP have indicated an interest in supporting the project through the provision of dairy commodities, and it is expected that such support would be provided throughout the project implementation period. Hovever, in the event of a shortage of commodities, the following sources of financing could be explored: borrowing by DDF from local development bank and other sources of commodity assistance. The Netherlands' Government has also indicated that it might provide funding beyond Year 3. If financing from these sources was not forthcoming, the project would have to be reduced in scope. 7.14 The project would also carry the inherent risks of a private sector development where uncertainties in supplies and markets put premiums on expert management. Since this would be the first attempt at implementing such a project, the appointment of top level staff for the project entities would be critical. The institutions established under the project, being private sector with high-valued throughput (by Year 7 of the project MILCO would process milk with a retail value of SLRs 1,350 million), would be able to attract and hold highly qualified management staff who are available in Sri Lanka. To ensure that DDF, MILCO, the MPC and the VMPC are adequately staffed, the project would give considerable emphasis to staff training (para 2.15). Financial analysis of the various project entities indicates that rates of return are relatively high and operating performance resilient to many of the uncertainties facing the firms (Chapter VI). 7.15 MILCO is envisaged as a high volume/low margin operation and thus is highly sensitive to changes in revenue or costs, especially if one changes without a corresponding change in the other (para 6.06). On the reverse side,  -45- the basic determinant would be the international price level of dairy com- modities. At the present time, these international prices are very depressed, and as it is these depressed prices which have been used in the analyses, the probability of major price falls undermining HILCO's viability is low. However, to guard against deleterious movements in international prices, provision is made for the imposition of a variable tariff (para 5.04). This would protect against a fall in international prices of about 16Z from the levels used in the analyses. Any further fall in international prices would have to be cushioned by reducing the procurement price offered producers. On the cost side, there is no reason to believe that MILCO's costs will move out of line with the general cost/price relationships assumed in the analyses. 7.16 Although the project would not be implemented by Goveroment agencies, Government commitment to the project would be critical, particularly in the early period to ensure the establishment of sound implementing agencies. 7.17 Farmers' responsiveness to the proposed project is a risk. They may be reluctant to accept the unfamiliar, private VHPC. In this connection, the effectiveness of the MPC Spearhead Teams (para 2.15 (a)) becomes all important. 7.18 Lastly, timing of the project could be considered a risk. Given the present world prices of dairy commodities (skim milk powder and butter oil), it could be argued that this is not the proper time to implement the project since milk producers in Sri Lanka cannot compete with imported dairy products without some type of Government intervention (para 5.04). Furthermore, increas- ing the local production of milk in effect simply adds to the vorld surplus of dairy products. On the other hand, recent events indicate that this may be an opportune time to undertake the project. Both EEC countries and the US have taken what appear to be effective measures to reduce surplus production of milk. Given the time required to increase milk output through the development of a national herd of genetically superior dairy animals, it seems obvious that now is the time to initiate the development of Sri Lanka's dairy industry rather than waiting until world surpluses of dairy products disappear, especially so since even with present prices the project is economically viable (para 7.06). In any event, since the project is heavily dependent upon donated dairy com- modities for financing, it must be carried out while surpluses of such com- modities exist. VIII. RECOMMENDATIONS 8.01 At negotiations, Government gave assurances that: (a) NNB's dairy plants and chilling centers would be leased to MILCO and the NMB would be abolished by September 30, 1986 (para 2.10); (b) The Dairy Plant Engineer would be appointed by December 31, 1985 (para 2.12); (c) The acquisition of building sites and the final designs for the process- ing and chilling facilities scheduled for construction in Project Year 2 be completed by September 30, 1986 (para 2.12);  -46- (d) Funds required for project start-up operations would be provided under terms and conditions satisfactory to the Bank (para 3.03); (e) Interest rates on loans by DDF to MILCO, the MPC and VMPC/VMPA would be reviewed semi-annually and adjusted as necessary to ensure that DDF's on-lending rates remain positive in real terms (para 3.04); (f) DDF would furnish to the Bank certified copies of its audited financial statements within nine months of the end of the financial year, and that Participation Agreements between DDF and the other project entities would spell out in a manner satisfactory to the Bank the disposition of the audited financial statements of MILCO, the MPC and the VMPC/VKPA (para 3.05); (g) An Audit Committee with terms of reference satisfactory to the Bank would be formed within six months following loan effectiveness (para 3.06); 'h) MILCO would not distribute dividends until its debt:equity ratio declined to 60:40 (para 4.08); (i) The MPC and the VMPC/VNPA would exchange Participation Agreement, satis- factory to the Bank, setting forth their understandings and respon- sibilities under the project (para 4.18); (j) The Steering Committee with terms of reference and composition satisfac- tory to the Bank would be formed within three months after loan effec- tiveness (para 4.19); (k) All documents supporting SOE would be audited annually by independent auditors acceptable to the Bank (para 4.22); (f) DDF would prepare semi-annual progress reports while the monitoring unit in MRID would prepare annual project evaluation reports and would be responsible for preparation of the Project Completion Report (para 4.27); tm) MILCO would be granted autonomy in milk pricing (para 5.02); (n) Concurrence of the Bank would be sought before the tariff on imported dairy products is raised above 25% (para 5.04); and (o) The subsidy on milk used by NMB for the local manufacture of whole milk powder would be phased out vithin two years of loan effectiveness (para 5.05). 8.02 Conditions of loan effectiveness would be that: (a) The lease agreement between NMB and MILCO concerning NMB's dairy plants and chilling centers has been entered into (para 2.10); (b) The cofinancing and the commodity assistance agreements have been properly executed (para 3.03);  -47- (c) The Subsidiary Loan Agreement between GOSL and DDF and the Participation Agreements between DDF and MILCO and DDF and the MIPC have been properly executed (para 3.04); and (d) The subsidy on processed fluid milk has been removed (para 5.05). 8.03 With the above assurances and covenants, the project would be suitable for a Bank loan of US$38 million to the Democratic Socialist Republic of Sri Lanka for a period of 20 years, including 5 years of grace.  - YF - -7,1- X,7,1en^ z.1WS k  - 49 - ANNEXES  ' I I&-  11 IRV simal co Is~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~1 I- kill In he: ISP.~~~~~~~~~~~~~~~~~~~II n fuml[VOI Lil FIAIVIII so 3".. 1, 464: S,~~~~~~~~~~~tit,. 044 1 will4V106 so II I 2611 6S I - 0164 p015. .. .:2 0128111t1014120i 00 10 . 16 11:6 III~~~~~~~~~~~~~~~~~~~~~~~ ... ....... .... .... .... ...... ..... .I.. ...... ..... ..... ..... ..... ..... ... .. I 0111111ta MORI illlt3. 2:I 121- S.-212 4 11.2; 14.5... .... .... .... ....... 1*ielel 041465114 810 loollOtlI :.tge@ itt . . . . . . 7.203 6 4,022.6 18.6 ..... ...T ...... ........... .................... FUJol INVIIIIJ 2411 6 C0 . SIS1128 112 .S I ..... ..4.. ...4:4 Fly I . 41 6.5h:,6 OlelItti 2.6210 . . . . . . . ,.oua 0 tile . . . . . . . 2,294.2~~.. ..... .... ..... .... ....... & 11111114140115(4116~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 ,1 ; ,1 11 magMa "I 16. I I S I S I . t6 21S0II 2S1 .l S II 26 26I 2S 1 1111. 1 4 114.6 46. 04 46 OI allIMMI 0: 1 I 0 6 A 0 1 114 11 1 0I42 14 22 :4 I04 le.IiAS42.1 111.4 11, lI Il. 8. 6, S M.u. 0145:14 SIe1SCIlO Al 06 I I I I I I I toS a 11.i 1211 2606s 2106 2104 2i006 264e6 2150 5 22I 1. sK 116.4 210.6 alto4 ML$. Sit.6 en)4. 4n., oa1:.s . 611614 tAhI 1 1 1 1 1 1 I .2. Ilm M02 028 111.0 8.0 61M.4 0ll0 alto$ 6it 0 6al0 1672440 1080 6226~ I. ,2. .0 I:s: .o .0. to g 640 U26 1.111lv 0 . 1 22 4 21 11 26 16 22 11 21 el 24 .2 46110 1126. 1. 564.0 164101 61.25is 6606.4 ill, .3: if 6t0lCfbIlI 5it 11 0IV I.1 1 2 2 1 2 2 I 226 S2 21 St. 16 11 I 16 216 24 7 .1 4 III 2 ill Il14l I 1. $. 44 111 .0* KpUI:.12 1165liii 1611 1 I 2 1 2 2 1 2 2 22.: 1 1262 61 2216IMI M 2 S 6 3276 216