Sri Lanka - Tree Crop Diversification (Tea) Project

Document of The World Bank h~CP FOR OFFICIAL USE ONLY Report No. 1936a-CE SRI LANKA TREE CROP DIVERSIFICATION PROJECT (TEA) I STAFF APPRAISAL REPORT April 3, 1978 Agriculture Division B South Asia Project Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS US$1 = Rupees (Rs) 16 Rs 100 = US$6.25 WEIGHTS AND MEASURES 1 mile (mi) = 1.609 kilometers (km) 1 acre (ac) = 0.405 hectares (ha) 1 inch = 25.4 millimeters (mm) 1 pound (lb) = 0.4528 kilograms (kg) ABBREVIATIONS ADD - Agricultural Diversification Division AOO - Area Operations Officer DCS - Dairy Cooperative Society FD - Forestry Department FOO - Field Operations Officer GM - General Manager GOSL - Government of Sri Lanka GRC - Gurugoda-Ritigaha Catchment HAMT-Farm - Higher Altitude Mixed Tree Farm JEDB - Janatha Estates Development Board LAMT-Farm - Lower Altitude Mixed Tree Farm LC - Land Commissioner LRC - Land Reform Commission MAL - Ministry of Agriculture and Lands MKC - Mahaoya-Kudaoya Catchment NAC - Nilambe-Atabage Catchment NADSA - National Agricultural Diversification and Settlement Authority PTU - Project TechLnical Unit, National Livestock Board SCD - Soil Conservation Division SPC - State Plantations Corporation SRI LANKA FISCAL YEAR January 1 - December 31 GLOSSARY Clonal or vegetatively propagated tea outyields seedling tea because it is selected from superior mother bushes and the bushes are genetically identical; seedling tea produces bushes of varied yield potential and will always contain some poor yielding material.  FOR OFFICIAL USE ONLY SRI LANKA TREE CROP DIVERSIFICATION PROJECT (TEA) I TABLE OF CONTENTS Page No. I. SECTOR BACKGROUND .................................. 1 II. PROJECT GENESIS AND OBJECTIVES ..................... 3 III. PROJECT AREA ....................................... 5 Location ........................................... 5 Climate ............................................ 5 Topography, Soils, Soil Erosion .................... 5 The Population ...................................... 5 Employment and Incomes ............................. 6 Social Infrastructure .............................. 6 Agricultural Setting ............................... 6 Land Tenure, Size of Holdings and Present Land Use . 7 IV. THE PROJECT ........................................ 7 A. General Description ........................... 7 B. Project Activities ............................ 9 C Implementation Schedule ....................... 13 D. Cost Estimates and Financing .... ............. 15 E. Procurement and Disbursements .................. 17 F. Accounts and Audit ............................ 18 G. Environmental Effects ......................... 19 V. ORGANIZATION AND MANAGEMENT ........................ 19 A. Project Organization .......................... 19 B. Project Implementation ........................ 21 C. Agricultural Support Services .... ............. 24 D. Preparation of Future Projects .... ............ 25 This report is based on findings of an IDA appraisal mission to Sri Lanka in October/November 1977 comprising Messrs. Thornley, Schorosch, Pekonen, Sinha (IDA) and Jessup and Robinson (consultants). This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  -2- Table of Contents (continued) Page No. VI. AGRICULTURAL PRODUCTION ............................ 25 VII. PRICES, MARKET PROSPECTS AND MARKETING .... ......... 27 VIII. FINANCIAL ANALYSIS . ................................. 29 Farm Budgets .............. .......................... 30 Farmer's Income ........... .......................... 31 Project Rent .............. .......................... 32 Project Cost Recovery ...... ........................ 32 IX. BENEFITS, ECONOMIC JUSTIFICATION AND RISK .......... 33 Production Benefits ....... ......................... 33 Employment Effects ....... .......................... 34 Income Distribution ....... ......................... 34 Foreign Exchange Impact ...... ...................... 34 Other Benefits ......... ............................ 34 Economic Analysis ....... ........................... 35 Sensitivity Analysis and Project Risks .... ......... 38 X. RECOMNENDATIONS ........ ............................ 40 ANNEXES 1. Cost Estimates, Expenditures, Disbursements and Allocation of Credit Proceeds 2. Proposed Land Use and Farm Budgets 3. Economic Analysis 4. Prices, Market Prospects and Marketing 5. Government Policy Statement on Agricultural Diversification 6. Selected Documents and Data Available in the Project File  -3- Table of Contents (continued) LIST OF TABLES IN MAIN TEXT Table No. 3.1 Species Composition of Mixed Forest Gardens 4.1 Present and Projected Land Use 4.2 Planting Materials and Phasing 4.3 Summary Cost Estimate 6.1 Project Output in 1990 8.1 Summary of Farm Budgets 8.2 Period of Financial Contribution from NADSA 8.3 Labor Requirements 8.4 Farmers' Income 8.5 Cost Recovery 9.1 Economic Costs/Benefits - Total Project 9.2 Economic Analysis of Project Components 9.3 Sensitivity Analysis of Project Economic Rate of Return LIST OF FIGURES Figure No. Figure 4.1 Implementation Schedule 5.1 Organization Chart of NADSA LIST OF MAPS IBRD 13383   SRI LANKA TREE CROP DIVERSIFICATION PROJECT (TEA) I I. SECTOR BACKGROUND The Agricultural Sector 1.01 The Role of Agriculture in the Economy. Agriculture plays a major role in Sri Lanka's economy, accounting for 37% of GDP, 55% of total employment, about 80% of export earnings and a large share of public revenue. Approxi- mately 80% of the country's total population lives in rural areas. Thus, economic growth in the country depends largely upon performance in the agri- cultural sector. During the 1960's, considerable progress was made in in- creasing food production as GOSL pursued a firm policy of increasing food self-sufficiency. Paddy production expanded from 840,000 tons to 1.6 M tons (6.5%/year) as a result of area expansion (175,000 ha), increasing yields - from 725 kg/ha to 1,050 kg/ha - and increased cropping intensity. After 1970 area expansion continued at a similar rate but yields and cropping intensity declined so that production grew at only 0.7%/year. Major reasons for the fall-off in productivity were farm power shortages - reflecting foreign exchange constraints - and high fertilizer prices. However, by 1977 Sri Lanka was 68% self-sufficient in rice compared with 50% in 1960. 1.02 While GOSL's basic policy of promoting food production was sound the tree crop sector, which in the early 1960's accounted for over 90% of the value of merchandise exports, has been neglected. During the 1960's production grew at only 1% per annum, and that mainly from rubber which was benefiting from substantial replanting in the early 1950's. Since 1970 conditions in the tree crop sector have further deteriorated, with output of tea, rubber and coconuts all pointing downward. Declining quality has also been evident, especially in the tea industry. The disappointing production trends in the tree crop sector were primarily a reflection of declining profitability, associated with lower prices and high taxes, the uncertainties associated with GOSL land reforms and inadequate foreign exchange allocations for vital inputs such as fertilizer and spares and replacements for machinery and vehicles. 1.03 Land Reform. The first phase of land reform, initiated in 1972, limited private holdings to: (i) 25 acres if exclusively paddy land or (ii) 50 acres if not exclusively paddy land, with paddy land not to exceed 25 's acres. In the first phase takeovers, implemented in 1973 and 1975, about 7t'- 225,000 ha of land were vested in the Land Reform Commission (LRC). Of this -7 about 60% was under tree crops which included about 23% of the total tea 7 area, 15% of the rubber area, and about 10% of the coconut area. The State Plantations Corporation (SPC), under the Ministry of Plantation Industries (MPI) was given about 40 of the larger well-maintained tea and rubber estates, covering about 10% of the tree crop land taken over. The management of most of the remainder was placed in the hands of cooperatives.  -2- 1.04 In the second phase, initiated in October 1975, all land owned or possessed by Sri Lankan and foreign companies, which included the country's major tea and rubber estates, was vested in the LRC. About 65,000 ha were assigned to SPC and nearly 915,000 ha to the Janatha Estates Development Board (JEDB) which was established for the purpose under the Ministry of Agriculture. This meant, that over 60% of the tea estate land, 30% of the rubber estate land and 10% of the coconut estate land were vested in the LRC. In late 1977 the management of much of this land was significantly rationalized. JEDB was transferred to MPI and the Upcountry Estate Development Board and Electorate Level Cooperatives, which had mismanaged the estates they con- trolled, were disbanded and their estates divided between JEDB and SPC. 1.05 The Tree Crop Sector. Within the agricultural sector the tree crop sector plays a strategic role. It provides a major contribution to value added in agriculture, the bulk of the country's export earnings, and a considerable amount of employment. In 1976, tree crops, of which tea, rubber, and coconuts are the most important, contributed Rs 2,546 M or about 29% to Sri Lanka's total value added of agriculture and about 11% of GNP. In addi- tion, processing of tea, rubber and coconuts added about Rs 780 M to GNP. In the export sector, the three major tree crops accounted for about 70% of the country's total export earnings in 1976 while tea alone accounted for 44%. Over one million people are employed in the tree crop sector, which represents a quarter of the total labor force and one third of the total number of people employed. Of this labor force, 700,000 are employed in the tea sector, 200,000 in the rubber sector, and 100,000 in the predominantly smallholder coconut sector. 1.06 The Minor Export Crop Sector. In this sector, besides tea, rubber and coconuts, the country produces numerous other crops, such as cinnamon, cardamom, citronella, cashew, cocoa, pepper, cloves, nutmeg and coffee. They are commonly referred to as minor export crops and are all, with the exception of cashews, wet zone crops. In 1976, about 76,000 ha were estimated to be under minor export crops, their value added being about Rs 125 M, and pro- viding employment for over 10,000 people. 1.07 Performance of the Tree Crop and Minor Export Crop Sector. Perform- ance of the tree crop sector in general has been disappointing in the 1970s as evidenced by the declining levels of production. Over the period 1970 through 1976, production of tea fell 7.4%, rubber 0.4% and coconuts 16.6%. Measured at 1959 constant factor prices, their contribution to GNP in 1970-76 also declined. Decline in production can be attributed to a variety of rea- sons, including severe drought, declining profitability due to low prices and high taxes, fragmentation of management and neglect of investment because of nationalization and general economic stagnation, which seriously affected the availability of vital inputs, especially of fertilizers. As one of the most conspicuous signs of the current problems, the annual rate of replant- ing in the tea sector dropped from about 2,800 ha in 1970 to only about 1,100 ha in 1976, while in the rubber sector it declined from about 4,100 ha to 2,000 ha. Coconut production has also met serious difficulties which have been associated with prolonged adverse weather conditions and the fact that production is based largely on numerous smallholdings having limited access  -3 - to credit, inputs and extension advice. However, performance of the minor export crop sector has been encouraging. From 1970 to 1976, mainly due to real price increases, the value added of minor export crops increased by about 150% in constant prices, but its contribution to GNP is still less than 1%. This sector has substantial potential for future development. 1.08 GOSL Tree Crop Sector Strategy. GOSL has come to realize how vital the tree crop sector is to the economy. Efforts to streamline the organiza- tion and management of the nationalized tea and rubber lands are underway. GOSL is also taking positive steps to improve policies on recruitment, staff- ing, and compensation of personnel for SPC and JEDB to promote flexible, com- mercially-oriented business practices. To provide government with a policy framework in the tree crop sector on sales and export taxes and on subsidies for fertilizer and replanting a "Tree Crops Incentive Study," a component of the Agricultural Development Project (Credit 595-CE), is presently being undertaken. Furthermore, to formulate future policies for the Tea Industry covering desirable production levels, organization, management, replanting, marketing strategies, etc. an overall master plan for the tea industry, with the assistance of the Canadian International Development Agency (CIDA), is underway. Finally, in order to promote production of minor export crops, a Department of Minor Export Crops was set up in 1973. 1.09. To implement its rehabilitation policies (sese para 5.22) for the tree crop sector, GOSL is formulating plans for land that should be diversi- fied out of its present land use. The Ministry of Agriculture and Lands (MAL) through its Agricultural Diversification Division (AD])) has been entrusted with this task. The executing agency for these programs of rehabilitation, diversification, and settlement of degraded lands will be the National Agri- cultural Diversification and Settlement Authority (NADSA), which was estab- lished early in 1978 under the State Agricultural Corporations Act of 1972. II. PROJECT GENESIS AND OBJECTI'VES 2.01 In 1967 the Government of Sri Lanka (GOSL) appointed a Commission to review the effects of a replanting program aimed at replacing old seedling tea with high-yielding clonal material. The Commission concluded 1/ that the program was necessary for the Sri Lanka tea industry to remain competitive in the world market; as a corollary, it recommended that it would be necessary to diversify some tea land to other crops, to prevent excessive supply and depression of the world tea market. 2.02 GOSL determined that the region best suited for diversification was the mid-country 2/ since (a) yields and quality of tea there were on 1/ Sessional Paper No. XVIII (1968). 2/ The 'mid-country' is variously defined as that part of the wet-zone total area of 1.5 million ha having an average annual rainfall of 2,000- 3,700 mm and lying between elevations 300 and 1000 m or, for tea industry statistical purposes, those tea estates lying between 600 and 1,200 m.  - 4 - average lower than elsewhere and (b) since many estates were therefore neglected soil erosion was accelerating and these lands would soon be vir- tually useless unless remedial action was taken in this very densely populated area. 2.03 In 1969/70 GOSL requested UNDP/FAO technical assistance to develop a strategy to ensure optimal land use, based on technical and economic fea- sibility studies of alternative crops. A UNDP project covering five years, "Agricultural Diversification of Uneconomic Tea and Rubber Areas," was ini- tiated during 1970 to study the problems and make recommendations. From 1973 this project focused on diversification out of tea on LRC held tea estates and settlement of such land in three river catchments areas, Nilambe- Attabage (NAC), Gurugoda-Ritigaha (GRC), and Mahaoya-Kudaoya (MKC). 2.04 Following this study and at the request of GOSL, the FAO/World Bank Cooperative Program helped in preparation of a project for Bank group financ- ing and, in May 1976, an IDA mission appraised the project. The project as then formulated was unsuitable for IDA financing. The key problems were: incomplete identification of land available for settlement; uncertain market prospects for the crops to be diversified into, mainly spices; no cost re- covery proposals; unsatisfactory organizational proposals; and the inclusicn of social and physical infrastructure proposals (hospitals, schools, markets, roads) that could not be justified economically. 2.05 As a result GOSL prepared a modified project which is the subject of this appraisal report. The major objectives of this new project are to: - develop an institution which can implement a program of rehabilitation, diversification, and settlement by landless families of degraded lands, mainly nationalized tea estates, and - halt the serious soil erosion that is ongoing under uneconomic tea, by introducing more appropriate farming and land use systems (perennial tree crops, pasture, and forests) that will - increase national agricultural production and income. 2.06 The project is of a pilot or test nature that hopefully will be the first in a series of diversification projects. It should generate data on the practical aspects of modifying land use and settling former estate workers and landless people. Additionally, the project would provide technical data on farming systems and crop/livestock mixes under controlled conditions. While it would not be possible to replicate the exact crop mix of this project in other projects, the majority, of crops envisaged for use in later projects would be introduced under the proposed project.  III. PROJECT AREA Location 3.01 The project is in the densely populated mid-country near Kandy, 90 km northeast of Colombo (Map 13383). It encompasses parts of three catch- ments, Nilambe-Atabage (NAC), Gurugoda-Ritigaha (GRC), and Mahaoya-Kudaoya (MKC) in Kandy and Kegalle District. The total cultivable area is 42,000 ha and project works would be carried out on 61 scattered tea estates totalling about 11,500 ha of which 9,000 ha are cultivable. Climate 3.02 Rainfall is high and fairly well distributied averaging 2,800 mm in NAC to 4,000 mm in GRC. Temperatures depend on the altitude which ranges from 500 to 2,000 m in the three catchments and are relatively constant, the mean of the coldest month being rarely less than 20 IC. As rainfall is seldom limiting and the soils fairly uniform, elevation and thus temperature largely determines the crops and varieties of crops that can be grown. Topography, Soils, and Soil Erosion 3.03 Topography. The main features of the catchments are the steeply dissected and mountainous terrain and the wide range in elevation from 500 m to 2,000 m. Most slopes exceed 30%, many exceed 70%, and some are precipi- tous with sheer rock outcrops. The valley bottoms are narrow, flat and alluvial and used almost entirely to grow paddy. 3.04 Soils. The soils are reddish brown or yellowish brown podsolic soils, friable with varying degrees of stoniness. They vary in depth, but are generally truncated through erosion. Under natural cover these soils would have a pH of about 5.5 in the top soil, but are now acidic, often down to pH 4.5 or less, due partly to the removal of top soil and partly to the extensive use of sulphate of ammonia as a fertilizer on tea. 3.05 Soil Erosion. Following many years of poor financial returns, and consequent neglect, most of the tea has very high vacancy rates. This, coupled with neglect of soil conservation, has tended to accelerate erosion with the development of many gulleys and extensive layers of surface gravel due to the removal of finer materials. The resultant peak flows and silt loads in the rivers are serious. The Population 3.06 General. In the project area, the total population is approximately 230,000 and total cultivable area about 42,000 ha; t.hus 5.5 persons, the average family, per cultivable ha. Most of the population is located in villages (60%), the rest in Village Expansion Schemes (18%) and on estates (22%). Average age of the population is low; in 1976, 50% were less than 20  years old. On the 61 project estates the population is about 16,000 people but only half of them are Sri Lankan citizens; the remainder are Tamils whose future citizen status is in doubt, pending completion of the Tamil repatria- tion program. 3.07 Employmcnt and Incomes. The estimated unemployment rate is about 30%. The total labor force increases by about 5,000 annually and few are able to find permanent work. The official daily wage rate for unskilled workers in agriculture in 1976 was Rs 7.5 (US$ 0.50) per male and Rs 6.5 (US$ 0.40) per female worker; but, in practice, many receive less and get only casual work. In consequence, many exist on less than Rs 2,400 (US$150) per family per year, the absolute poverty line below which basic nutritional standards cannot be maintained. A detailed socio-economic survey of the Gurugoda-Ritigaha catch- ment in 1975 revealed that the average family income was only Rs 1,400. On the 61 project estates about 7,000 people are employed as laborers, slightly more than half being women tea pluckers. The average family income of these estate workers is estimated at Rs 3,000. Social Infrastructure 3.08 The project area is well served by a network of over 150 km of major roads, 300 km of minor roads, and a cheap and effective bus service to even the smallest villages. Water supply comes mostly from perennial streams and springs with a limited number of public wells. Hence, water borne diseases are relatively common. Electricity is generally available. There are also numerous retail shops, and regular markets in larger villages. There are over 40 junior and 20 secondary schools and some advanced training institutions in Kandy and Kegalle. The area is adequately served by 9 rural hospitals, 16 central dispensaries and 42 estate dispensaries. Agricultural Setting 3.09 The three catchments, like the whole mid-country, are characterized by a dual agricultural economy. The plantation sector (tea, rubber, coconut) is based on relatively large estates with the labor force, mostly Tamil origin, living in so called labor lines, that are primitive barracks in very poor conditions. The smallholder sector, in turn, is dominated by the Sinhalese living in scattered small villages. 3.10 The smallholder sector is based predominantly on a mixed forest garden system where a wide variety of perennial crops are grown:  - 7 - Table 3.1: Species Composition of Traditional Mixed Forest Gardens (Crop density about 600 trees plus vines per ha) Arecanut Papaya Jak Fruit Vegetables Coffee Avocado Pepper Cocoa Coconut Breadfruit Banana Yams Tea Mango Cloves Cardamlom Nutmeg Citrus Most smallholders also have access to some paddy in the valley bottoms. Mixed forest gardens average 0.5 ha in the project area and are generally very poorly managed with no regard to scientific agricultural husbandry practices. 3.11 The plantation sector has undergone major changes. Many estates have changed hands in the last 15 years with each successive owner anticipat- ing nationalization. Consequently, estates were run to maximize short-term returns which resulted in rapid deterioration, with virtually no expenditures on maintenance or renewal. Under the Land Reform Act of 1972, the maximum private land holding was set at 20 ha per adult faraily member with any excess being acquired by the state. It is the acquisition of such excess land that would be diversified and settled under the project. On the 61 estates, average yields of made tea leaf are now about 510 kg/ha, declining by about 15% annually, compared to average yields of 1,100 kg/ha on better managed public sector estates (SPC and JEDB). Land Tenure, Size of Holdings and Present Land Use 3.12 Land ownership patterns as well as size distributions of village lands and estate holdings vary considerably between the three catchments. On average about 40% of the area is village lands and settlements and the rest is mostly tea land owned by the Land Reform Commission and administered by the Ministry of Plantations' State Plantations Corporation (SPC) and Janatha Estates Development Board (JEDB). Individual private holdings in the village areas are small, with more than 50% less than 0.5 ha. Though most Sri Lankan estates are more than 400 ha, most of the 61 project estates acquired by the Land Reform Commission (LRC) in 1972, range between 30 and 400 ha, mostly below 100 ha. IV. THE PROJECT A. General Description 4.01 The four-year project, covering 11,500 ha, would diversify 4,600 ha out of tea, 4,900 ha out of scrub land and 1,800 ha out of degraded forests  - 8 - into more appropriate land use. In areas where the land is suitable for s->ttlement the project would develop two types of mixed tree farms, some tea and some cardamom farms; where it is not, forest plantations and conservation forest would be established. Albout 4,500 farms of 0.8 to 1.6 hectare would be >marcated, planted with appropriate perennial crops and allocated to landless families from the project catchments. Simple settler houses would be built in clusters, for which water supplies and, where necessary, access roads would be constructed. The project would be carried out by the National Agricultural Diversification and Settlement Authority (NADSA), which has been set up by GOSL to carry out this and similar projects. NADSA would implement all project works using the prospective settlers as its labor force. 4.02 The project would comprise: 1. Development of Mixed Tree Farms (a) 2,550 Lower Altitude Mixed Tree Farms (LAMTF) below 750 m, on 2,800 ha; (b) 1,300 Higher Altitude Mixed Tree Farms (HAMTF) with pasture and dairying between 750-1,000 m, on 2,000 ha; (c) 390 Cardamom Farms (CF) on 700 ha; (d) 230 Tea Smallholder Farms on 180 ha. 2. Social Infrastructure. Construction of simple houses located in clusters, access roads and village water supplies. 3. Forestry. Establishment of about 1,800 ha of pines, 1,900 ha of eucalyptus and 2,100 ha of conservation forest. 4.03 The project would change the present land use as follows: Table 4.1 Present and Projected Land Use Present Future (ha) (ha) Tea 4,800 2,800 LAMT Farms Scrub Land 4,900 2,000 HAMT Farms Degraded Forest 1,800 700 Cardamom Farms 180 Tea Smallholder Farms 3,700 Commercial Forest 2,100 Conservation Forest All land presently under tea would be used for LAMT farms, HAMT farms and tea smallholder farms. Scrub land would be converted into RAMT farms (230 ha) and the balance into commercial and conservation forest. Degraded forest lands would be used for establishing cardamom farms (700 ha) and for commercial forest.  -9 - B. Project Activities Development of Mixed Tree Farms 4.04 Soil Conservation. Of the project lands designated for farms, 2,300 ha are protected against soil erosion by existing soil conservation structures. The remaining area would need bench terraces protecting 450 ha, graded stone walls protecting 700 ha, lock and spill clrains protecting 950 ha, and grass strips covering 125 ha. The extent of mechanical structures would be kept to a minimum and agronomic soil conservation measures would be used wherever feasible. In the areas designated for forest. and pasture, no con- servation structures would be needed, but forest plantings would be on the contour and in the pasture areas strips of selected high producing grasses would be planted across the slopes. 4.05 For each farm a house and vegetable plot of 0.2 ha would be pro- vided on land with less than 35% slope. For half of this area, bench terraces protected with grass plantings would be constructed. Farms would mostly be on land with slopes of over 35%. Where the slope is less than 35%, grass strips would be planted on the contour. To establish these grass strips, tea bushes would not be uprooted but cut and chemically killed to avoid soil disturbance. Where the slope is over 35%, either graded stone walls would be constructed where stone is available at the site, or lock and spill drains, where there are no stones. To erect these mechanical structures tea bushes would be uprooted from a strip 1 m wide along the structure lines. 4.06 For planning and supervising all technical aspects of the soil conservation works, the project would provide additional staff (13 profes- sionals plus support staff), vehicles (2 4-wheel drive and 10 motorcycles), and equipment (field, cartographic and office equipment) to the Soil Conser- vation Division (SCD) of the Department of Agriculture. The actual work would be done under the supervision of NADSA staff by the prospective settlers, for which the project would provide tools. For cost details see Annex 1, Table 2. 4.07 As soon as the soil conservation works were completed, the Survey Department would demarcate 4,500 homesteads and farms for which 20 surveying gangs (1 surveyor, 7 laborers) would be required. Beacons and maps would be provided (see Annex 1, Table 3). 4.08 Soil Conditioning. To increase the very low natural fertility of the highly eroded soils, green manuring would be carried out on both the home- steads and farms before and after planting. Green manure would be provided through guatemala grass established with fertilizer on 10% of the area and the subsequent cutting and compositing of the cut material. For costs see Annex 1, Table 5, Page 1. 4.09 Planting Material. To develop the 4,500 mixed tree farms, high grade planting material (about 5 million seedlings, cuttings and splits of  - 10 - over 20 species) would be required. This material would be produced mainly by two existing project nurseries with expanded facilities: one at the Minor Export Crop Research Institute at Matale and the other in the project area on one of the former estates. Small field nurseries would be set up where necessary. Coconut seedlings would be bought from the Coconut Research Institute nursery. Ginger rhizomes would be selected within Sri Lanka or, if necessary, imported from India and multiplied at the project nursery. The amount of planting material, the species composition and the time needed are shown in Table 4.2. Costs for planting material are shown in Annex 1, Table 5, Page 2. A detailed plan for the production and purchase of planting material has been prepared by NADSA and is on file. Table 4.2: Planting Materials and Phasing (in Thousands) Year 1 Year 2 Year 3 Species 1978 1979 1980 Totals Coconut - 57 57 Arecanut - 132 132 Gliricidia - 215 215 Pepper - - 692 692 Albizzia - - 113 113 Breadfruit - 10 3 13 Jak Fruit - 26 - 26 Citrus Trees - 64 - 64 Coffee - 710 552 1,262 Clove - 12 24 36 Nutmeg - - 16 16 Banana - 87 - 87 Papaya - 124 - 124 Avocado - 11 - 11 Cardamom 663 663 1,326 Ginger 460 tons 460 tons Grass Sets: Guatemala 767 tons 155 tons 922 tons Brachiaria - - Guinea B. 360 tons 360 tons Legume Seed: Stylosanthes 518 kg 518 kg 4.10 Homestead Development. Each participating family would be provided with a small plot of 0.2 ha where a simple house would be constructed and vegetables grown. On these plots, depending on subsistence requirements, market prospects, temperature range and rainfall, a variety of vegetables such as tomatoes, eggplants, beans, onions, cabbage, capsicum, sweet potatoes, squash, garlic, lettuce, radishes, etc. would be grown. Whenever the terrain  - it - allows, paddy would be grown. To provide shade and later fruit for the family a few jak or breadfruit trees and some citrus trees would also be planted. 4.11 Lower Altitude Mixed Tree Farms (LAMTF). 2,550 LAMTF, each about 0.8-1.0 ha, would be developed on land below 750 meters. In selecting the species mix, the aim would be to reach an optimal plant composition from an agricultural and economic point of view. Specifically, early maturing species for initial farm income and to provide a balance of shade and light for the various crops would be included, together with as many high value crops as possible, and fast growing support trees for one of them, pepper. 4.12 To meet these criteria, the species selected are: banana, to provide shade for other seedlings and initial income; papaya and avocado for early income, coconut, arecanut and breadfruit to provide shade and food; gliricida as support tree for pepper; albizzia as a major tree for shade and canopy adjustment; coffee, cloves, nutmeg, pepper, and ginger as cash crops. 4.13 Higher Altitude Mixed Tree Farms (HAMTF). 1,300 HAMTF, each about 1.2 ha, would be developed on land between 750 m and 1,000 m in elevation. The species grown in HAMTF would be different. No papayas, coconuts, avocadoes and nutmeg would be included, and banana would only be grown as a quickly develop- ing shade tree. The proportion of coffee (robusta) would be much higher than in the LAMTF. Cash crops would include pepper, coffee, cloves and ginger. Also compared to the LAMTF, shade trees would be decreased and wind break trees increased. Farm development costs (LAMTF and 'HAMTF) are in Annex 1, Table 5. 4.14 All HAMT-farmers would have access to 0.3 ha of pasture which would be planted under the project, and they would be provided with one house cow. The pasture plot would be developed within 1 km of the homestead where the cow would be kept and the cut and carry system of supplying fodder would be adopted; of the pasture area of each farm, about 75% would be planted to grass (Brachiaria brizantha and Panicum maximum) and about 25% to a legume crop (Stylosanthes humilis). Alternate strips of Brachiaria and Panicum would be planted across the hill slopes. Planting materials, seeds, and fertilizer for establishing the pastures would be provided by the project. For costs see Annex 1, Table 6, Page 2. 4.15 Since the farmers on HAMTF would need additional income until their longer term tree crops came into bearing, about 1,650 dairy heifers (6-9 month old) would be provided under the project by the beginning of year 3. The Project Technical Unit (PTU) of the Ministry of Agriculture, which is handling livestock procurement and technical services under the IDA financed dairy project (504-CE) would be responsible for importing the animals if local supplies are insufficient and for follow-up actions. Additional staff (1 livestock officer, 3 livestock extension officers, 1 veterinary surgeon, 2 field assistants, clerks, typists and drivers), vehicles (1 4-wheel drive, 6 motorcycles, 1 truck) and incremental operating expenditures would be provided to the Project Technical Unit. (See Annex 1, Table 6).  - 12 - 4.16 Cardamom Farms. 390 cardamom farms, each about 1.6 ha, would be developed on the limited areas suitable in the Gurugoda-Ritigaha Catchment. Cardamom planting material wouLd be selected from high yielding malabar and mysore varieties and interplanted with fast growing shade trees (glyricidia) and higher canopy shade trees (albizzia). Costs are shown in Ani.n.x 1, Table 5 page 3. 4.17 Tea Smallholder Farms. 230 tea smallholdings, each about 0.8 ha, would be formed on existing land where vegetatively propagated (VP) tea yielding above 5,000 kg/ per ha of green leaf already exists (equivalent to about 1,100 kg of made tea). 2. Social Infrastructure 4.18 Housing. About 4,500 simple semi-permanent houses would be built at a cost of Rs 1,100 (US$69) each. These would provide adequate basic shelter to the project participants for some years. It is expected that settlers would themselves builcd more substantial structures as soon as their farms generated sufficient revenue. They would have floor space of 6x6 m, divided into living and sleeping space. In addition, a small verandah (6x1.5 m) and an outdoor cooking place would be provided. All building materials would be available locally and provided by NADSA--coconut leaves (cadjans) and bamboo or arecanut trees for roof materials and poles, clay for floor materials and timber for window frames and doors. Floors would be of clay, mixed and beaten with earth. Until these houses are built by project labor, the project participants would continue to live in their present homes. They would normally walk to project sites or if necessary use the available cheap and efficient public transportation system. 4.19 Water Supply. The project would construct about (a) 540 village wells, (b) 135 water distribution points through diversion of small streams and construction of small tanks, and (c) 45 water distribution points through construction of spring tanks. In some cases, natural springs would provide the source for potable water. Tanks with 2.5 m high and 6 m long walls would be constructed, based on rubble foundation and built with bricks, sand and cement. To safeguard against waterborne diseases assurances were obtained from GOSL that all wells would be protected by covers and that handpumps for raising water would be installed. For all water tanks assurances were obtained that GOSL would arrange for regular disinfection with common bleaching powder. 4.20 Access Roads. About 100 km of low-cost access roads to settlements would be built at a unit cost of Rs 35,000 (US$2,200) per km. The roads would be 4.5 m wide having approximately 6 culverts per km. Only levelling and grading would be done, but no metalling. In addition, about 100 km of exist- ing deteriorated estate roads would be maintained by NADSA, including (a) clearing of culverts, (b) filling and patching of holes with metal and sand, (c) minor strengthening of shoulders, (d) clearing of drains, and (c) replacement of damaged road components (e.g., culverts). Unit cost of maintenance per km is estimated at Rs 1,200 (US$75). For details on costs of the social infrastructure component including the total amount of labor requirements see Annex 1, Table 4.  - i13 - 3. Forestry 4.21 Commercial Forests. The project would plant about 1,900 ha to fast growing Eucalyptus grandis and 1,800 ha to Pinus carribea. The choice of these species is based on performance of existing forest stands in the catchment. The forests would be established on land not suited to any other form of land use and would be planted in relatively large contiguous blocks to facilitate management. The eucalyptus forests would be grown on the coppicing system and would be for fuelwood. The pine forests would be for timber. 4.22 For planning and supervising all technical aspects of the reforesta- tion, the project would provide additional staff (6 professionals plus support staff), vehicles (2 4-wheel drive, 3 tractors with trailers) plus funds for site preparation for nurseries, seed material, and incremental operating ex- penditures to the Forestry Department (FD). The actual work would be done by the prospective settlers under the supervision of staff of FD and NADSA. For cost details see Annex 1, Table 7. 4.23 Conservation Forests. In the project area, land with slopes over 75% (2,100 ha), now under jungle and patna (a fire climax plant community largely consisting of grasses), would be preserved as conservation areas to prevent erosion and to reduce the rate of rain run-off. These lands would be protected against agricultural use by locals and against fire in the driest months, particularly January to March by employing settlers as fire watchers. The forests would be owned and managed by the Forest Department. C. Implementation Schedule 4.24 The project as designed consists of two phases, a development phase (3 years) and a consolidation phase (3 years). IDA's direct involvement would only extend over the development phase plus the first year of the consolida- tion phase. All developmental works would start between July 1978 and January 1979 and are scheduled for completion within 3 years by June 1981 as shown on the Implementation Schedule (Figure 4.1). The consolidation phase would extend over another 3 years until June 1984. The development phase would be charac- terized by complete responsibility of NADSA and its staff for getting the homesteads and farms established and would end with the allotment of farms to settlers. The consolidation phase would be characterized by a gradual with- drawal of NADSA staff from the area and would end by the time (year 6) the tree crops would have come into bearing. The sequence of project activities would be: - increased production of planting material from existing nurseries starting early 1978; - Soil conservation work starting in the projiect area in July 1978 and ending in April 1979, except for the establishment of grass strips on farms which would continue in April-June 1979 and the green manuring on farms which would be a continuous process;  - 14 - Surveying of homesteads and farms in January-March 1979; Construction of access roads, houses and provision of water supply starting by the end of 1978 as soon as progress on the soil conservation works allows and ending by end of March 1979; Planting of vegetable plots, mixed tree farms and cardamom farms starting with the first rainy season in 1979; Forestry planting spread over 3 to 4 rainy seasons starting by October-December 1979; Importation of dairy heifers by late 1979 to be kept at a calf rearing station until 15 months later; and planting of pasture in the first half of 1981.  SRI LANKA Figure 4.1 TREE CROP DIVERSIFICATION PROJECT (TEA) I Implementation Schedule YEARS 1978 1979 1980 1981 QUARTERS 1st 2nd 3rd 4th 1st 2nd |3rd h 1st 2ni 3rd 4th 1st 2nd 3rd 4th ESTABLISHMENT OF NADSA - Staff recruitment * _ - Training I _ _ (a) Preqlualification of settlers * _ b) Hand-over of estates from SPC aend Janawasama INFRASTRUCTURE AND HOUSING (a) Construction of roads to clusters (b) Building of houses (ci Water supply SOIL CONSERVATION (a) Homesteads - bench terraces (b) Farms-lock and spill drains (c) Farms-graded stone walls (d) Biological (grass strips) SOIL CONDITIONING (a) Establishment of Guatamala grass strips (for grees manure) X. -Homesteads -Farms (b) Green manuring -Homesteads -Fa,rms- PLANTING (a) Nursery (Incd. grass) - (b) Homesteads __ - _ _ _ 1 _ ic) Farms - SURVEYING (a) Homesteads (b) Farms - ALLOCATION OF FARMS DAIRY COMPONENT (a) Introduction of cows )b( Planting of pasture (c) Distribution of cows to farmers FORESTRY PLANTING - - World Bank - 18369  - 15 - D. Cost Estimates and Financing '.25 Details of project costs are in Annex 1 and summarized in Table 4.3. Total project cost is estimated at US$6.5 M, of which US$2.0 M or 31% is for- eign exchange. Taxes and duties amount to US$0.1 M. Cost estimates are expressed in projected mid-1978 prices by inflating appraisal estimates (end 1977) by 5% for local costs and 3.75% for foreign costs. Works estimates are based on detailed quantity estimates provided by the Soil Conservation Divi- sion, the Forestry Department, tea estates and local contractors, familiar with the works required. Unit prices for equipment and materials are based on December 1977 quotations. ]Physical contingencies of 15% have been applied to all labor requirements and 10% to material, equipment and operating expend- itures; no physical contingencies were included for salaries. Estimated price increases over the implementation period amount to about 13% of base cost plus physical contingencies and assume the following annual rates of price infla- tion for local and foreign costs: Inflation Rate (Percent) 1978/79 1979/80 1980/81 1981/82 Local 10 8 8 7 Foreign 7.5 7 7 7  - 16 - Table 4.3: Summary Cost Estimate Base Local Foreign Total Local Foreign Total Cost - (Million Rupees)-- Wrhousand US$) /a - 7 NADSA - Capital Costs 3.2 5.1 8.3 200 319 519 10 - Operating Costs 8.1 6.7 14.8 506 419 925 17 Farm Development - Soil Conservation & Conditioning 7.6 0.6 8.2 475 38 513 10 - Surveying 0.8 - 0.8 50 - 50 - - Higher & Lower Altitude Mixed Tree Farms 11.2 2.1 13.3 700 131 831 16 - Dairy Component 3.9 11.2 15.1 244 700 944 18 - Cardamom Farms 2.8 0.1 2.9 175 6 181 3 Social Infrastructure - Roads 3.9 0.1 4.0 244 6 250 5 - Houses 4.9 0.3 5.2 306 19 325 6 - Water Supply 2.2 0.1 2.3 138 6 144 3 Forestry 7.5 2.2 9.7 469 137 606 12 Base Cost Estimate 56.1 28.5 84.6 3,507 1,781 5,288 100 Physical Contingencies 6.1 1.5 7.6 381 94 475 9 Expected Price Increases 9.8 2.0 11.8 612 125 737 14 TOTAL PROJECT COSTS 72.0 32.0 104.0 4,5,00 2,000 6,500 123 /a US$1 = Rs 16.0 Financing 4.26 The proposed IDA Credit of US$4.5 million would finance the full foreign exchange costs and about 55% of local costs, or about 70% of the total project cost, net of taxes and duties. GOSL would contribute the remaining US$2.0 M. GOSL would allocate 100% of project funds on a grant basis to NADSA in the form of annual budgetary allocations. The NADSA budget would include funds to cover project related expenditures of the Soil Conservation Division, the Forestry Department, and the PTU of the National Livestock Board (see Annex 1, Table 2, 6 and 7). GOSL and NADSA would submit to IDA, not later than November 1st each year, a detailed implementation and financial plan for the following fiscal year which coincides with the calendar year. Details of the contents of these reports would be agreed with IDA prior to the submission of the first report.  - 17 - 4.27 Proposed project financing is summarized below (in thousand US dollars): GOSL IDA Total NADSA 744 700 1,444 Farm Development 519 2,000 2,519 Social Infrastructure 219 500 719 Forestry 106 500 606 Contingencies 412 800 1,212 Total 2,000 4,500 6,500 E. Procurement and Disbursement 4.28 Equipment and Vehicles. NADSA would be responsible for all procure- ment activities. Since it would be extremely difficult to ensure adequate maintenance and spares for a small number of a variety of imported vehicles (US$0.23 M) and tractors and trailers (US$0.40 M) not already available in Sri Lanka, procurement would be through competitive bidding advertised locally and in accordance with local procedures which are acceptable to IDA. There are enough local firms and representatives of foreign firms to ensure a suf- ficient degree of competition and reasonable prices. Small tools, and survey and office equipment (US$0.18 M) would be procured by prudent shopping follow- ing local procedures which are acceptable to IDA. 4.29 In order to expedite early project implementation, GOSL has been asked to prepare tender documents for vehicles, tractors and trailers for which retroactive financing of up to US$0.5 M is recommended. 4.30 Cattle. Approximately 1,650 six to nine months old dairy heifers costing about US$0.7 million net of contingencies would be imported. The import of these heifers would be organized by the PTU. Because of the wide variations in breeds, types and adaptability of animals available, and the disease situation in supplying countries, priced proposals would be obtained from at least three countries free from foot and mouth disease, that can supply heifers of the type required, in accordance with procedures acceptable to IDA. Before accepting any offer, these animals would be inspected in the country of delivery by a duly authorized representative of the PTU who would also check that transportation arrangements were acceptable to ensure healthy arrival. 4.31 Assurances were obtained from GOSL that, for all tenders for vehicles, and dairy cattle worth over US$100,000, technical specifications and draft tender documents would be submitted to IDA before bidding documents were issued; and for individual contracts for goods worth over US$100,000, bid evaluation would be submitted to IDA for approval before making the awards.  - 18 - 4.32 Civil Works and Land Development. The bulk of project works, in- cluding construction of access roads, housing, water supply, soil conserva- tion, planting of homesteads, farms and forest, would be executed directly by NADSA using the prospective settlers as the labor force. Any skilled labor requirements would be met by contracting local craftsmen. 4.33 Planting Material. The Forestry Department would buy pines seed (US$0.03 M) directly from suppliers. Supplies are limited and only two supplier countries are known. Ginger rhizomes would be procured directly by negotiated purchases after quotations have been solicited from at least two suppliers. 4.34 Disbursements. Disbursements would be made against: (i) 100% of the foreign expenditures for directly imported goods; (ii) 60% of expenditures for goods procured locally; (iii) 100% of the ex-factory price of goods manu- facturer locally; (iv) 50% of project related expenditures for salaries of staff and operating costs of NADSA, Soil Conservation Division, Forestry Department, and PTU; 1/ (v) 60% of expenditures for all project works carried out by settler-labor and skilled labor; (vi) 100% of expenditures for consult- ants. Disbursements for salaries of staff and operating expenditures would be made against certified statements of expenditures. Disbursements for project works under force account would be made every quarter against physical works progress report and a certificate of expenditure from NADSA. The documenta- tion for this would be retained by NADSA and made available for inspection during project supervision. Estimated schedules of expenditures, disburse- ments and the proposed allocation of the proceeds of ithe loan are presented in Annex 1, Tables 8, 9 and 10. F. Accounts and Audit 4.35 Separate project accounts would be kept by NADSA for all project expenditures incurred by NADSA, the Soil Conservation Division, the Forestry Department and the PTU. 4.36 Throughout the project, NADSA would maintain summary accounts and a record of disbursement applications forwarded to IDA. Project accounts would be audited annually by independent auditors acceptable to IDA. Certified copies of the audited accounts and the auditor's reports would be submitted to IDA within six months of the end of each calendar year. The audit report would be required to include, inter alia, a statement that funds disbursed against statements of expenditure had been used for the purpose for which they were provided. 1/ Staff and operating costs for NADSA, SCD, FD and PTU are shown in Annex 1, Table 1, page 2 and 3; Table 2, page 1; Table 6, page 1 and Table 7.  - 19 - G. Environmental Effects 4.37 The thrust of the whole project is to stop the physical degradation of the environment in the projlect area and to ensure that project lands can be used for agricultural purposes in the future. Soil conservation measures would be carried out accompanied by a shift in land use from uneconomic tea to perennial tree crops, pasture and forest which provide good ground cover. These project activities, besides enabling continued agricultural use, would reduce the silt load in some of the tributaries of the Mahaweli Ganga River and the silting up of irrigation tanks fed by that river. V. ORGANIZATION AND MANAGEMENT A. Project Organization The National Agricultural Diversification and Settlement Authority (NADSA) 5.01 General. GOSL established NADSA early in 1978 under the State Agricultural Corporations Act of 1972 to implement this and similar projects on a larger scale in the future. It is an agency under the Ministry of Agri- culture and Lands (MAL) and will be financed by regular GOSL budget appro- priations. A seven-member Board of Directors will have overall responsibility for project implementation. All seven Board members have been appointed by the Minister of Agriculture and Lands. The Board would meet at least quarterly to review the status of project implementation. Full delegation of authority for day to day operations would be made to the General Manager (GM) who would be the Authority's Chief Executive. 5.02 Organization. The GM would be directly assisted by a small tech- nical unit and there would be two departments: operations and support services. The operations department would have four sub-units, three area operations divisions and one nursery unit. The support services department would com- prise units for accounting, auditing, and personnel and office administration (Figure 5.1). 5.03 Staffing. At headquarters, the GM would be assisted by a Deputy General Manager and a Chief Technical Advisor, a Progress Control Officer and a Board Secretary, all directly reporting to him. Also at headquarters, the operations department would be headed by a Chief of Operations, responsible for planning and programming all field work. He would either be a member of the planters community, having worked as estate superintendent for at least ten years, or an outstanding agricultural degree holder with at least ten years experience in successfully managing a large labor force. The Chief of Operations office staff would include two engineering assistants responsible for the infrastructure component of the project (houses, roads and water supply), one transportation and equipment officer and one nursery officer.  SRI LANKA Figure 5.1 TREE CROP DIVERSIFICATION PROJECT (TEA) I Proposed Organization Chart of NADSA Board of Directors I 7j pp.Members r General Manager | Support Services Dept. Operations Dept. -Accounting - Auditing - Personnel & Office Administration Area Operations Area Operations Area Operations Nursery Division I Division 11 Division III Unit | Field Operations | | Field Operations l Field Operations Officers j Officer Officer World Bank - 18368  - 20 - The support services departrnent at headquarters would be headed by a Chief of Support Services and would comprise officers for accounting, internal auditing, personnel and office administration. 5.04 At field level three Area Operations Officers (AOO) would be in charge of 25 Field Operations Officers (FOO) and the Nursery Unit would be under a nursery officer. The three AOO would either be former estate super- intendents with five years of experience or agricultural university graduates with five years of pertinent experience. The Nursery Officer would have at least 5 years of experience in running as a manager a government or private nursery and the Field Operations Officers would either be former estate assistant superintendents with ten years of service or diploma holders in agriculture with seven years of service. In all cases, familiarity with organizing a large labor force would be the decisive selection criteria. 5.05 The 25 FOO, one per 180 settlers, would form the backbone of NADSA. They would be in daily contact with the settlers and would live in the field. The FOO would be responsible for organizing the settler labor force for all project activities, including those requiring technical expertise from a specialized government agency (Soil Conservation Division, Forestry Department, PTU, Survey Department). Each would be assisted by one clerk who would keep records of the number of working days provided by each settler family, to provide the basis for payment and the decision on settler entitlement to a plot after three years. For details on staffing see Annex l, Table 1, Page 2. 5.06 Training of the FOO. FOO, before joining the project, would undergo refresher courses in three subjects: a one month training in soil conserva- tion provided by staff of the Soil Conservation Division, a two weeks train- ing in agronomy of spice crops provided by staff of the Research Station of the Minor Export Crops Department and a two weeks course in extension methods provided by the In-service Training Institute of the Department of Agriculture. Courses would be held in the existing facilities of these agencies and field trips would be organized. After these initial courses during project implementation, FOO would meet fortnightly with their AOO to program the activities for the next fortnight and to review experiences from the pre- vious fortnight. The Senior Technical Advisor from the GM's office would attend these meetings and Subject Matter Specialist (SMS) from the spe- cialized agencies involved in project execution would be called in as and when required. 5.07 Equipment. All NA)SA professional staff would be mobile; the AOO would have 4-wheel drives and the FOO motorcycles. The nursery unit would have 2 tractors, 4 trailers and 5 trucks. The trucks would be available at Headquarters for transport of planting materials and fertilizer. One tractor with two-way radio communication equipment and 2 trailers per FOO would be available in the field. These tractors and trailers would be used in soil conservation works for the transport of stones over short distances, in all planting activities for transporting planting materials and fertilizer, and in implementing the social infrastructure component for transporting building materials from roads to the settlements. Minor farming equipment and work tools such as hoes, rakes, shovels, baskets, etc., would be provided to all project participants. For cost details see Annex 1, Table 1, Page 1.  - 21 - 5.08 Other Agencies Involved. Coordination between NADSA and the spe- cialized government and semi-government agencies referred to in paras 4.06, 4.14 and 4.22, would be ensured because the heads of these agencies report directly to the Minister of Agriculture and Lands. Detailed annual working plans would be prepared by these agencies and approved by NADSA. For day-to- day operations, field staff of these agencies during their involvement in project implementation would be responsible to NADSA. While this staff would be working out of the offices of their parent agencies or live in the field, their parent agency would only be responsible for the technical backstopping and the quality and soundness of their work. But all decisions as to when and where soil conservation or forestry works would be carried out for example, would be taken by NADSA. Assurances have been received from GOSL that all staff provided under this project to SCD, FD and PTU would work exclusively on the project. B. Project Implementation 5.09 A legal opinion by the Attorney General confirming that project lands acquired under the Land Reform Act are owned by the government and are free of incumbrances has been received. Although th;e transfer of control of land covered by the Project Area to NADSA would be a condition of effective- ness, management of those lands will be transfered to NADSA according to a time phase transition plan under development by Government. Until such man- agement is turned over to NADSA under this plan, those agencies currently managing the lands will continue to perform that function. Settler - Selection 5.10 General. It is desirable that the labor force on the 61 tea estates to be handed over from SPC and JEDB to NADSA should be carefully prepared for the event and informed about potential effects of the project on their future. To this end the Government is preparing an estate by estate census of the citizen status of the labor force involved to determine the size of the Tamil population and to identify those scheduled for repatriation or application for citizenship. Assurances have been received that GOSL would take all measures possible to provide, either through relocation to other estates, or through accelerated repatriation to India, relief to those workers not eligible for settlement under the project. 5.11 Selection Criteria and Procedures. Selection would be a two stage process and it would be NADSA's responsibility to see that it be carried out. At the beginning of the project, about 4,500 families would be selected as participants in the project to constitute the work force of NADSA and as prospective permanent settlers. Selection would be from the following categories in order of priority: I - displaced families of citizens or of those to be granted citizenship prior to June 30, 1981 from the 61 project estates;  - 22 - II - landless families from villages in the project catchments; and III - unemployed youth from the project catchments. Settlers would be selected from a lower priority category only if there were no eligible families remaining in the preceding category. Farming experience would not be a factor except in the allocation of plots with a dairy component, these would be given whenever possible to settlers with some dairying expe- rience. This initial selection would qualify a settler to receive a plot at the end of the third year subject to satisfactory performance while employed by NADSA in project works during this initial development phase. The main criterion would be accepting, and satisfactorily completing, at least 80% of the jobs offered by the Authority. Any vacancies among the settlers would be filled from the highest priority category containing eligible settlers. Assurances have been received that GOSL would follow this settler selection procedure. Plucking of Tea 5.12 As project works progress the area under tea would decline, but meanwhile, NADSA through the FOO would organize the plucking on the residual areas, which would be done by settler family members, mostly wives. NADSA would pay the going wage and sell the green leaf tea on its own account. Over the three year period in which plucking would continue, NADSA would at least break even (see Annex 2, Table 8). Transport and processing of green leaf would be carried out by tea estates remaining in the area, managed by JEDB and State Plantations Corporation. Assurances have been received that arrangements satisfactory to IDA would be made between GOSL and NADSA to guarantee the transport and lprocessing of green leaf from the project area. 5.13 The area of vegetaltively propagated (VP) tea having high enough yields to remain under tea (180 ha) would be assigned by NADSA to settler families in plots of 0.8 ha. The first three years tea plucking would be organized as a training cum selection period. Wages would be paid out of income. In year 4, these settlers, about 200, would enter into marketing arrangements for transport of their green leaf with the nearest estate. Land Allocation 5.14 At the end of the first year settlers selected would be allocated, on a temporary lease basis, a homestead consisting of 0.2 hectares of land on which a house would have been constructed. Homesteads together with farms would be allocated on a permanent basis under the Sale of State Lands (Special Provisions) Law, No. 43 of 1973. Thereunder the settler would elect to: (i) purchase his holding and receive a "grant" (a title) to the land; or (ii) pay the price of the holding in twenty annual installments. Until the last installment is paid he would hold a "permit" for the land which would be replaced by a "grant" when payment was complete. The amount of the installments would be determined in accordance with para 8.07. In- stallment payments would begiLn on the following basis: (a) tea smallholdings  - 23 - in the fourth project year; (b) high altitude mixed tree farms in the sixth project year; (c) cardamom farms in the seventh project year; and (d) low altitude mixed tree farms in the eighth project year. Assurances have been received that GOSL would follow this land allocation procedure. Sale of Dairy Animals 5.15 As most project farmers would not be skilled cattle keepers, imported yourg heifers after about 15 months at the Haragama calf rearing unit would be sold as pregnant heifers, five month in calf, to successful dairy farmers under the present IDA-financed dairy project and not to project participants. These farmers would trade one of their local animals for one imported and finance the price difference between the two out of their own resources or through existing credit sources. Funds thus generated would be used by NADSA to pay part of the cost of keeping the imported animals at Haragama. Cows exchanged would be inspected by PTU staff to ensure that they would be suitable for sale to project farmers, and this exchange scheme would be under the responsibility of the PTU. 5.16 Local animals obtained under this cow exchange system would be sold to project farmers on credit under procedures already developed under the ongoing dairy project. Both the Bank of Ceylon and the Peoples Bank have expressed interest in participating. Banking facilities are available in the area. Farmers would apply for credit through one of the PTU farm plan- ning technicians in the project area who would help the applicant to prepare a farm development plan for submission to the lending institution. Where PTU considers project farmers have sufficient livestock experience, imported heifers from Haragama may be sold to them directly. 5.17 Presently the IDA financed Sri Lanka Dairy Development Project is being reformulated along the lines of the successful AMUL-model developed in India. Farmer owned Dairy Cooperative Societies (DCS) and Milk Unions would be established under by-laws similar to those developed in the three IDA- assisted Indian dairy projects. A Registrar of Cooperatives in the Ministry of Agriculture and Lands (MAL) would be responsible for registering DCS. The PTU would be responsible for organizing the service functions and establishing DCS. The four existing processing plants of the National Milk Board are in the process of becoming Milk Unions and equity would be transferred gradually to member DCS who would be represented on the Board of the Union. All inputs and services would ultimately be controlled from the Union processing plant and distributed through the DCS to member producers. Assurances have been received that GOSL would establish DCS in the project area with by-laws satis- factory to IDA not later than July 31, 1981 before dairy animals kept at Haragama would be sold to farmers. Consulting Services 5.18 To assist in carrying out the project and especially in variety selection and the processing and marketing of the most important spices to be grown under the project, especially ginger, pepper and cardamom, NADSA would employ experts for short-term visits to assist its own senior technical staff.  - 24 - The estimated consultant input is 12 man-months at a total cost of US$96,000 including international transportation. Assurances have been received that NADSA would employ consultants with qualifications and on terms and conditions acceptable to IDA. Project Monitoring 5.19 The Progress ControL Officer and his staff and the Chief Technical Advisor would be responsible for project monitoring. They would keep records on utilization of projects inputs, especially the use of tractors, vehicles, trucks, planting material, staff and the settler work force. They would also collect data on timely generation of project outputs in relation to the imple- mentation schedule such as construction of soil conservation structures, houses, roads, and water supply and establishment of farms and on unit costs thereof and on circumstances that are critical to the effective implementation of the project (cooperation at field level between staff of NADSA, SCD, FD and PTU). 5.20 The monitoring unit would also be responsible for ongoing evaluation. They would conduct sample surveys to determine the overall family income of the settlers, including: (a) pay from NADSA, (b) pay from employment outside the project and (c) income from the house vegetable plot and farms (built-up of yields and farmgate prices for inputs and outputs would be closely monitored). Assurances have been received that NADSA would collect these data twice a year and that, if family incomes were below Rs 3,600 in constant 1978 Rupees, GOSL would promptly take necessary measures, acceptable to IDA, to ensure that settlers have the financial ability to continue to develop their farms. To provide an independent assessment in greater depth of the economic and social impact of the project, provision has been made for the monitoring unit to commission such surveys and studies as might be necessary. All information generated by the monitoring unit would be reviewed in time to form the basis for the project design of other diversification projects that are planned. C. Agricultural Support Services 5.21 All agricultural production support including planting material, fertilizer, and extension and research would be provided initially cost free to the laborer/settlers by NADSA during the development phase (3 years). These input costs along with accrued interest would be capitalized and par- tially recovered in the sales prices of the land (paras 8.07 and 8.08). Thereafter NADSA staff would be gradually reduced from a ratio of 1 FOO to 180 settlers to 1 FOO to 450 settlers. This remaining staff would continue until the end of year 4 to provide fertilizer to the settlers. To ensure that proper husbandry practices are followed in year 5 and 6 when farmers' incomes are still low, assurances have been received that NADSA would continue to organize and remunerate selected settlers (see Table 8.4) for training of pepper vines and necessary weeding, fertilizing and pruning for all crops on settlement farms in the project area. From year 7 onwards NADSA's respon- sibility for the settlers would stop and they would depend on existing agricultural service institutions in the area. By that time settlers would have become experienced farmers.  - 25 - D. Preparation of Future Projects 5.22 Land rehabilitation, diversification and settlement of degrad-i lands is needed on a broad scale. The project would be only an initial effort, preparing the way for a larger and more comprehensive program. In March 1978, GOSL issued a statement on its diversification policy which outlines its general objectives for diversification and establishes respon- sibilities and authority for diversification efforts; between concerned ministries and institutions (Annex 5). The Minister- of Agriculture and Lands has also assigned to the Agricultural Diversification Division the task to inventory lands suitable for diversification and to prepare a second project not later than mid-1980. VI. AGRICULTURAL PRODUCTION 6.01 All crops included in the mixed tree farms and cardamom have been grown in Sri Lanka for a long time. A great deal of research and extension information is available on their planting, growing, harvesting and pro- cessing. Only ginger would be a relatively new crop to the project area. It appears that at least two local varieties of ginger are grown in Sri Lanka, of which only one produces good dry ginger acceptable to the inter- national market. Analysis of the available cultivars for their dry ginger characteristics would be carried out before being used as planting material in the project. 6.02 In the type of mixed tree farming proposed for the project three factors basically determine the value of total output: yields per individual plant, plant density, and species composition. 6.03 The following assumptions have been made: Yields (for the major crops at full development) Fertilizer Application pepper - 3.6 kg dry per vine LAMTF - 250 kg/farm/year ginger - 12.0 tons green per ha HAMTF - 900 kg/farm/year cardamom - 600 kg green per ha CF - 250 kg/farm/year cloves - 2.73 kg dry per tree coffee - 0.45 kg dry per plant  - 26 - Plant Density and Species Composition: LAMTF HAMTF Species Number of Plants per Farm Number of Plant per Farm (0.8 ha - 1.0 ha) (1.3 ha) Banana 16 32 Papaya 12 - Coconut 22 Avocado 4 - Coffee 30 400 Arecanut 40 20 Bread Fruit 4 2 Jak 8 4 Gliricidia 40 80 Albizzia 4 2 Clove 9 9 Nutmeg 4 _ Other Trees 20 10 Pepper 80 100 Ginger 0.08 ha 0.08 ha 6.04 The cropping pattern proposed has a plant density which is consid- ered to be optimum at the present levels of soil fertility. It has a high proportion of high value crops such as pepper and cloves. Additionally, ginger has been introduced. With improved planting material, cultural practices and the use of fertilizer, a 10% higher yield then in existing farms was assumed. Soil fertility on project lands will gradually improve, and plant population densities can be increased accordingly. On the basis of only 10% higher yields, the gross returns per ha on the project farms would be about Rs 12,000, or more than double gross returns on traditional farms, primarily because of the difference in species composition. 6.05 Production Output (see Table 6.1). About 70% of value of total production excluding forestry in 1990 would be from export crops, the rest from crops destined for the local market. Among the export crops, pepper at 1,300 tons and 31% of total value of project output is the single most im- portant crop. Next in importance is ginger with 600 tons and 12% of total value, cardamom with 400 tons and 11% of total value and, cloves with 106 tons and 8% of total value. The share of coffee in total value is 5% and other crops, mainly nutmeg, coconut, and arecanut account for 6% of the total value. Produce for the local market would include vegetables (8% of total value), milk (6% of total value) and other crops including bananas, papayas, avocados, and citrus fruits (13% of total value). Forestry products would include on an annual basis 35,000 cu meter of eucalyptus after Year 8 and 13,000 cu meter of pine after Year 22. 6.06 Production Inputs. The main annual inputs to sustain this level of production would be labor and fertilizer. Total annual labor requirements for all the farms are estimated at 5,500 man-years, easily obtainable from the 4,500 settler households, and at 500 man-years for forest activities, exclud- ing felling and processing. Annual project fertilizer consumption is esti- mated at 2,000 tons.  - 27 - Table 6.1: PROJECT OUTPUT IN 1990 Total Price Value of Total Pro- Species Production per ton duction at Farmgate Price % (in tons) (Const. (Thousand Rs) (Thousand uS$) /a 1978 Rs) Crops for Export 1. Pepper 1,285 13,200 16,962 1,060 31% 2. Ginger 592 11,000 6,512 407 12% 3. Cardamom 392 15,000 5,880 368 11% 4. Clove 106 44,000 4,664 292 8% 5. Other /b - - 3,172 198 6% 6. Coffee 287 9,900 2,841 178 5% Subtotal 2,503 Crops for Local Market Vegetables 3,582 1,300 4,657 291 8% Others /c - - 7,104 444 13% Milk 2,127 1,700 3,616 226 6% Subtotal 961 Total 3,464 100% Forestry /d Eucalyptus 34,650 cu meter 22.2 Rs/cu 770 48 meter Pine 13,050 cu meter 398.0 Rs/cu 5,194 325 meter /a 1US$ = Rs 16.0. /b Including nutmeg, coconut, and arecanut. 77 Including banana, papaya, avocado, breadfruit, jak, and citrus trees. 7d Computed average annual production for eucalyptus after Year 8 and for pinus after Year 22. One sixth of eucalyptus area would be harvested every year. Pinus would be harvested over six years in 22-year cycles. VII. PRICES, MARKET PROSPECTS AND MA1RKETING 7.01 More than 70% of the marketable surplus of crops from the project area would be for export. Of these, pepper would be the most important. Sri Lanka has a long tradition of producing and exporting its pepper which is well known internationally for high (and favorable) oleoresin content; in spite of which Sri Lanka has had only a small share (0.5%) of the international trade of around 90,000 tons per year. Pepper exports have declined (now less than 25% of the 400 ton 1957-76 average) because internal consumption has increased  - 28 - since Government stopped import of chillies, for which pepper is a substitute. Project production would allow Sri Lanka to reverse this declining trend. By 1990 the expected project area production of about 1,300 tons would be only about 5% of the expected increase in international pepper trade. Sri Lanka would be able to export this quantity without major difficulties or adverse effect on the world market prices. 7.02 The project would produce an exportable surplus of other spices such as ginger, cardamom and cloves and a small quantity of coffee. The market outlook for these commodities is good, and as the project surplus would form less than 6% of international trade in any of the spices and the coffee pro- duction would be less than 1% of international trade, no major marketing prob- lem is expected. Prices of these commodities are currently high and lower long-term average prices have been projected. The ratio of projected farmgate prices to the current farmgate prices varies between 55 and 80 percent (see Annex 4). Traditional marketing channels exist in the project area and are efficient in all cases excepting cardamom. Assurance have been received from GOSL that cardamom growers would be registered and that it would ensure that government-owned cardamom estates in the project area would provide a process- ing and marketing channel for the cardamom producers. 7.03 Project milk surplus would amount to about 1,500 tons per year, which would displace about 170 tons of milk powder imports annually. In the catchment areas milk collection is currently carried out by multi-purpose cooperative societies in an inefficient manner. Assurances have been received from GOSL that dairy cooperatives along the lines being suggested under Credit- 504 would be formed in the catchment areas (see para 5.17). Milk collected by cooperatives would be sold to the National Milk Board (NMB), which could process the milk at the nearest milk processing plant at Pallekelle. This plant is currently operating at 30% capacity and NMB intends to rehabilitate it to bring it up to capacity. 7.04 Non-industrial demand for fuelwood in the catchment areas is esti- mated at 85,000 cu m annually. Eucalyptus from project lands would be very competitive with the existing sources of fuelwood, which are of poor quality and dwindling in supply. No marketing problem is anticipated. Fuelwood price has increased by more than 100% in the last eighteen months and a further 30% increase, in real terms, is anticipated by 1985. The Project would also produce pine for lumber. Market prospects are excellent. GOSL recently had to ban cutting of trees in state forests as a reaction to overcutting of forests. This has severely constrained supply and project output should be able to relieve this problem. Price of pine for lumber is conservatively projected at about $40 per cu in, which is the current price for inferior quality lumber. Further details of pricing and market prospects and marketing are given in Annex 4.  - 29 - VIII. FINANCIAL ANALYSIS Farm Budgets 8.01 Detailed budgets for the various settlement components and sub- components are in Annex 2 and summarized in Table 8.1. These budgets reveal one of the problems associated with smallholder tree crop cultivation. Many of the crops, though highly profitable, mature slowly. Hence, revenue buildup is slow. This places a heavy financial burdens on settlers in the initial years. Recognizing this financial problem, the project would try to overcome it in three ways. First, revenues would be provided in initial years by in- cluding some fast growing crops such as vegetables ancd ginger. Second, inputs would be provided by NADSA directly and recovered as part of farmers rent pay- ments. Third, NADSA would employ prospective farmers as paid laborers during 6 years (see para 5.21). This would guarantee employment opportunities for 2.4 persons per family in year 1, 1.7 persons in year 2, 1.01 persons in year 3, and 3 month per settler in year 4-6. For cardamom farmers about 6 months annually of paid work opportunities would have to be provided during years 4-6 to provide them with an adequate subsistence income. This approach would also ensure that (a) inputs are used directly for developing farms, (b) farms are not neglected because of frustration of farmers over slow develop- ment, and (c) farmers are trained in proper agricultural practices. 8.02 The time period for which NADSA would have to contribute financially to the development of farms would vary according to plant composition of the different types of farms (Table 8.2).  - 30 - Table ,8.1: Summary Farm Budgets (Rupees per farm) Years Lower Altitude 1 2 3 4 5 6 7 12 Mixed Farm (2,550 farms) Revenue /1 - 780 2,810 3,420 3,620 4,340 5,260 10,580 Cash Expenses - (320)/2 (340) (340) (390) 390 390 390 Net Revenue - 780 2,810 3,420 3,620 3,950 4,870 10,190 Higher Altitude Mixed Farm (1,300 farms) Revenue /1 - 780 2,710 5,080 6,970 8,020 8,650 13,300 Cash Expenses - (510) (750) 830 2,520 1,970 2,440 1,590 Net Revenue - 780 2,710 4,250 4,450 6,050 6,210 11,710 Cardamom Farms (390 farms) Revenue /1 - 780 910 1,040 1,580 2,660 7,760 16,400 Cash Expenses - (310) (310) (310) (310) (490) 490 490 Net Revenue - 780 910 1,040 1,580 2,660 7,270 15,910 /4 Smallholder Tea Farms /a (230 farmsT Revenue /1 - 780 6,910 7,040 7,040 7,040 7,040 7,040 Cash Expenses - (770) 770 770 770 770 770 770 Net Revenue - 780 6,140 6,270 6,270 6,270 6,270 6,270 /a Subdivision of existing higher yielding clonal tea. Note: 1. Revenues are estimates of value of annual production, including home consumption at farmgate prices. 2. Expenses within brackets represent value of free inputs provided by NADSA and are capitalized as developmental costs. They do not represent a cash outflow for the farm. 3. All monetary values rounded to nearest 10 and number of farms rounded to nearest 5. 4. Long term average annual net revenue from cardamom farms is estimated to be Rs 12,000.  - 31 - Table 8.2: Period of Financial Contribution from NADSA Farm Model Time Period (years) LAMTF 5 HAMTF 4 CF 6 Tea Farm 2 8.03 The Summary Farm Budgets (Table 8.1) show no cash expense figures in the initial years. This reflects full financing of investment by NADSA. For HAMT farms, in year 4, only partial financing from NADSA would be required. In later years, current revenues would be able to cover cash expenses and leave sufficient surplus to cover basic needs of the settler's family. Cash expenses for most of the farms are very low because tree crops usually require very small cash inputs (a little fertilizer). No outside labor would have to be hired because of small farm size, extended harvesting period and absence of any significant annual replanting. Family labor requirements are presented in Table 20 of Annex 3 and are summarized below for year 12. A socio-economic survey of the catchment area shows available labor participation of 2.2 persons per family. Considering 285 available days per year, labor availability would be 627 mandays per farm. Even considering some seasonality in harvesting, the family labor would suffice. In view of the low cash expenses for all farms (excepting HAMT farms), high and well distributed revenues, very little working capital would be required. For HAMT farms, expenses appear high because of debt service on dairy loans. Dairy cooperatives would deduct the necessary amount from each day's milk sales. Therefore no cash reserve for debt service will have to be maintained. Working capital required would vary between 2-4 percent of revenue and would be internally financed. Table 8.3 Labor Requirements (mandays per farm in year 12) Farm Model Labor LAMT farm 212 HAMT farm 338 Cardamom 225 Tea 523 Farmers' Income 8.04 After the crucial consolidation phase, most farmers' net revenue would rise rapidly to average annual levels between Rs 6,000 and Rs 12,000 ensuring sufficient surplus to allow farmers to pay back most of the project costs. See Table 8.4.  Table 8.4 Farmers' Income -/ (Rupees per farm) - -- -- -- -- -- -- -- -- -Y E A R S - - - - - - - - - - - - - - - - - - - - - - 1 2 3 4 5 6 7 8 12 LAMT Farms Farm Income - 780 2,810 3,420 3,620 3,950 4,870 5,690 10,190 +Wages from NADSA 6,200 4,360 2,540 810 780 - - - -Annual Charges 2/ - - - - - - - (1.540) (1.540) Total Net Income 6,200 5,140 5,350 4,230 4,400 3,990 4.870 4,150 8,650 HAMT Farms Farm Income - 780 2,710 4,250 4,450 6,050 6,210 7s780 11,710 +Wages from NADSA 6,200 4,360 2,620 270 - -Annual Charges 2/ - -- - - (1(430) (1.430) (1.430) .1L430 Total Net Income 6,200 5,140 5,330 4,520 4,450 4,620 4,780 6,350 10,280 > Cardamom Farms Farm Income - 780 910 1,040 1,580 2,660 7,270 9,400 15,910 +Wages from NADSA 6,200 4,360 3,130 3,070 2,670 1,060 - - - -Annual Charges 2/ - - - - - - (2,264D (2.260) (2,26 0) Total Net Income 6,200 5,140 4,040 4,110 4,250 3,720 5,010 7,140 13,650 / Tea Farms Farm Income - 780 6,140 6,270 6,270 6,270 6,270 6,270 6,270 +Wages from NADSA 5,060 4,200 - - - -Annual Charges 2/ - - - (620) (620) (620) (620) (620) (620) Total Net Income 5,060 4,980 6,140 5,650 5,650 5,650 5,650 5,650 5,650 l/ Figures rounded to nearest ten. 2/ See cost recovery, Annex 2, Table 12. 3/ Long term average annual net income from cardamom farms after payments of annual installments is estimated. to be Rs 10,000.  - 32 - Project Rent 8.05 To evaluate settlers' capacity to repay project costs, project rent for the four farm models has been estimated and presented in Annex 2, Table 12. The "rent" accruing to a settler family indicates the maximum amount the family can pay and still benefit from the project. All the above indices have been computed in real terms and in present value terms. The discount rate used is 10%, representing the estimated economic opportunity cost of capital. From the present value of the net farm revenue, opportunity costs of family labor, allowance for risk and imputed value of management services are deducted. On a per farm basis the project rent is estimated at Rs 19,500. This implies that an average settler could be expected to pay at most Rs 2,300 per year for 20 years. Project Cost Recovery 8.06 Total project costs are valued at the end of the developmental phase (in year 4) at market prices using a 10% interest rate! and are estimated at about Rs 86 M. The value excludes the opportunity cost of uprooting tea, because that is done as a part of national objective of tea diversification and farmers in the project area should not be made to pay for the cost of that policy. It also excludes the costs for the forestry component. On a per farm basis the cost works out to be Rs 19,000. 1/ This is below the total project rent and therefore full rent recovery would result in full project cost re- covery. Considering that the project area farmers' income would be close to the critical consumption level of about Rs 3,600 for the first several years and would not rise above the national average family income of Rs 12,000 at full development, full rent recovery is not advisable and a 66% project rent recovery index would be aimed at. The HAMT farms would be paying an additional fixed charge of Rs 660 per year for five years to repay full cost of heifer purchase along with accrued interest at 10%. 8.07 The Sale of State Lands Law provides for cost recovery in equal installments over normally 20 years. Annual installments based on a 10% discount rate together with the corresponding sales prices and first years of repayments are shown in Table 8.5. 1/ At Rs 19,000 (US$1,190) the cost per family are in the lower range of World Bank supported settlement projects. See Annex 2, Table 11.  - 33 - Table 8.5 Cost Recovery Annual Charges Commencement of Sales Price for 20 years Repayments (Rs per farm) (Rs per farm) (Prcject Year) LAMT 20,900 1,540 8 HAMT 19,400 1,430 6 Cardamom Farm 30,700 2,260 7 Tea Farm 8,400 620 4 Since these figures are estimated in constant 1978 rupees, before contract signing, these values would have to be appropriately modified to reflect then current values. In the absence of inflation, the annual installment payments together with government revenues from forest lands would represent a cost recovery index of about 80 percent. If there is a 7 percent annual inflation the cost recovery index in real terms would decline to about 55 percent. During negotiations cost recovery methods have been discussed and an assurance has been received that GOSL would adopt a cost recovery scheme satisfactory to IDA. Fiscal Impact 8.08 The GOSL contribution to the project would be Rs 104 M in direct project costs and Rs 27 M in lost revenues and taxes from uprooted tea in the project area. The project cost would amount to about Rs 30,000 (US$1,815) per family or Rs 5,000 (US$315) per capita. The costs would be spread over six years and would average about Rs 22 M per year, which is 0.5 percent of GOSL's capital budget. As the proceeds from the sale of State lands would recover Rs 58 M, the net cost of the project would be Rs 73 M, or Rs 2,700 (US$170) per capita only. 8.09 Incremental operations and maintenance cost of the project would be Rs 0.7 M per year and would be more than offset by direct revenues amounting to Rs 6.0 M per year from Government-owned commercial forests to be created by the project. The project would also create incremental employment for 1,200 families and thereby move them out of Government's welfare program. Assuming that rice subsidies continue at the present rate, the project would save Rs 1.3 M per year in subsidies. Thus the net long-term impact of the project on the Budget would be positive and would amount to about to about Rs 6.6 M per year, excluding payments for capital cost recovery. IX. BENEFITS, ECONOMIC JUSTIFICATION AND RISKS Benefits Production Benefits 9.01 The main benefits wculd be from export crops such as pepper, ginger, cardamom, cloves, and nutmeg (para 6.05). The farmgate value of the  - 34 - incremental production of these export crops would be Rs 40 M (US$2.5 M) per year by 1990. Other crops, milk and forestry products (excluding pines) 1/ worth Rs 14 M (US$0.9 M) per year for the local market would also be produced. Additional income opportunities amounting to Rs 14 M (US$0.9) per year would be created for processing and marketing of the farm produce. Employment Effects 9.02 Owing to declining employment opportunities on tea estates and in- creasing population, the catchment areas suffer from severe unemployment. 2/ Without project, another 4,800 ha of tea land would be abandoned by 1985, putting another 3,300 households (9% of catchment area population) out of work. The project, by giving direct employment to about 4,500 households, would reverse this trend and reduce migration to the already overcrowded cities. This employment would be generated at a cost of only about Rs 12,000 per person. This compares with Rs 44,000 of fixed capital per person employed for public sector enterprises in 1973. income Distribution 9.03 All the direct beneficiaries of the project would be landless farm families or unemployed youth. Average per capita income in these groups is presently less than US$60 per annum. The project would help to more than double their per capita income to about US$125 per annum by 1990. The land- less farm workers would receive individual title to about 5,700 ha of land from the state sector. This is consistent with the policy of GOSL to make the population less dependent on the state sector. Foreign Exchange Impact 9.04 Sri Lanka would earn Rs 48 M (US$3 M) per year in foreign exchange from export of spices and coffee and would save Rs 7 M (US$0.5 M) per year by 1990 in import of milk products and fuel oil. The annual cost of importing fertilizer would be about Rs 9 M (US$0.6 M), giving a net foreign exchange benefit of Rs 46 M (US$2.9 M) per year by 1990. Other Benefits 9.05 The benefits from the social infrastructure component are largely non-quantifiable, but would respond to the basic needs of the settlers by providing improved shelter, safe and adequate water supply and better access to the area's basic road network and nearby services. Upgrading of water supply would have a beneficial impact on the health standards of the popu- lation by reducing water-borne diseases. The planned community development 1/ Pines would not be harvested until year 2000. 2/ A Government survey of the catchment areas, conducted in 1976, esti- mated unemployment at 30% of the workforce.  - 35 - by concentrating housing into clusters and improving accessibility of serv- ices would also constitute a major step in government's policy to increase viability and attractiveness of settlement schemes. 9.06 GOSL's strategy in the tea sector is to increase productivity of tea lands in better tea areas through replanting. To prevent any increase in production which might create marketing problems, GOSL's strategy also involves diversifying out of tea in the least desirable tea areas. The proposed project, by creating a diversification implementation capability, would assist GOSL to implement its strategy in the tea sector. 9.07 Partly due to the lower profitability of tea in the mid-country, the level of investment on tea estates in the project area was low. This contri- buted to high rates of soil erosion, resulting in heavy silt load in some of the tributaries of the Mahaweli Ganga River and silting up of irrigation tanks fed by that river. Every year landslides occur on tea estates resulting in property damage. The project would shift land use to pasture, forestry and those tree crops which provide good perpetual canopy. This would gradually reduce soil erosion and help re!habilitate the soil in the project area. Economic Analysis 9.08 The economic rate of return (ERR) of the project would be about 16% and is based on cost and benefit streams shown in Table 9.1.  - 36 - Table 9.1 Economic Costs and Benfits ('000 Rs) Less: Less: Net Incre- Benefits /a Less: Infra- Less: Benefits mental (with Capital structure Operating (without Net Year project) Costs Costs _b Costs /c project) /d Benefits 1 - 19,856 9,148 - 3,521 -32,525 2 3,124 39,258 161 270 5,444 -42,009 3 10,919 27,357 635 347 8,325 -25,746 4 17,656 1,239 161 11,696 5,772 -1,212 5 18,578 496 161 11,910 3,072 2,939 6 22,914 721 161 12,579 950 8,503 7 28,866 290 161 13,504 - 14,911 8 33,499 290 161 14,095 - 18,953 9 40,069 290 161 14,865 - 24,753 10 48,407 290 161 15,463 - 32,493 11 53,704 290 161 15,858 - 37,395 12 56,410 290 161 16,148 - 39,811 13 56,769 290 161 16,148 - 40,170 14 57,180 290 161 16,148 - 40,581 15-22 55,545 290 161 16,148 - 38,946 23 75,290 290 161 16,148 - 58,691 24 77,265 290 161 16,148 - 60,666 25-26 79,436 290 161 16,148 - 62,837 27 81,832 290 161 16,148 - 65,233 28 84,454 290 161 16,148 - 67,855 29-40 55,545 150 161 16,148 - 39,086 Economic Rate of Return - base 17% - with physical contingency 16% /a Benefits from export crops, fruits and vegetables (CF=1); fuelwood (CF=1.6); milk (CF=1.083) and lumber (CF=1.037). Does not include bene- fit from tea. See Annex 3, Table 11-14 for calculations of conversion factors (CF). CF is the ratio of the world market (border) price to the domestic market price, where the domestic price is expressed in dollars converted at the market exchange rate. lb 56% of the costs are for unskilled labor (CF=0.67). /c 43% of the costs are for unskilled labor (CF=0.67), 42% for fertilizer (CF=3.24-3.7) and 9% for dairy cattle feed (CF=1.31-1.69). /d Income from tea foregone due to project. Income foregone is low in the beginning because of phased uprooting of tea. See Annex 3, Table 1, 5, 6, 18 and 19.  - 3 7 - 9.09 In calculating the economic rate of return, foreign exchange is valued at the current exchange rate of US$1 = Rs 16 to which the Sri Lankan rupee was devalued in November 1977. 1/ Output prices used in the analysis have been discussed in Chapter 7 and Annex 3. For the most part economic prices of project output are very close to their financial prices. On the cost side, fertilizer has been valued at its projected c.i.f. price plus internal processing and distribution cost. Owing to heavy subsidization, this cost is almost three times the present financial cost to farmers and is expected to increase to almost four times that value by 1985. Opportunity cost of unskilled labor has been valued at two-thirds of the Government established minimum wage rate, because of the high unemployment rate in and around the project area (para 9.02). Opportunity cost of feeding dairy cattle has been estimated based partly on the f.o.b. value of exporting coconut meal. By 1985 the economic cost is expected to be almost 70% higher than the finan- cial cost. Some project land is currently under tea. Economic benefits foregone due to uprooting of tea are deducted from project benefits. The deductions are lower in the first two years of the project because tea would be uprooted slowly and in a phased manner. After year six of the project no benefits are deducted because even without project, tea would have been abandoned because of lower yields, lower tea prices and higher unit costs of production. 9.10 Economic analysis of project components and subcomponents is pre- sented in Annex 3, and summarized below. 1/ With high import duties (up, to 100%+) the equilibrium rate is probably nearer 18 than 16 in the long-term. Had a shadow exchange rate of 18 been used, the ERR would be slightly higher.  - 38 - Table 9.2 Economic Analysis of Project Components Settlement Components Economic Rate of Return (percent) Lower Altitude Mixed Farm 18 Higher Altitude Mixed Farm /a 15 Dairy Farm Subcomponent of HAMT Farm 12 Cardamom Farm 20 All Settlement Components 16 Forestry Components Eucalyptus (Fuelwood) 16 Pine (Lumber) 14 All Forestry Components 15 Total Project 16 /a Excluding the Dairy Farm subcomponent, which is analyzed separately. 9.12 The table shows that ERR of project components and subcomponents range between 12% for the dairy subcomponent and 20% for the cardamom farms component, with the majority of components clustered between 14% and 18%. Though the rate of return of the dairy subcomponent is relatively low, it is still justifiable since the opportunity cost of capital is estimated at about 10%. Dairy pastures would occupy some of the poorest lands in the project area; more than 33% of which would be suitable for no other productive use. Though the cardamom farms have a high rate of return, the component cannot be increased in size as cardamom requires high elevation, good shade and heavy rainfall and no additional areas are available in the three catchments. Sensitivity Analysis and Project Risks 9.13 The project has been designed to minimize risks, both to the economy and the farmer, through diversified production. None of the commodities would contribute more than 30% to project's economic benefilt nor more than 35% to any settler's income. This reduces risks inherent in commodity price fluc- tuations. In addition, risks of decline in yield due to plant diseases and pest attack are also reduced. This aspect of the proposed project is revealed by a sensitivity analysis of prices, yields and cost of production of the different commodities.  - 39 - Table 9.3 Sensitivity Analysis of Project Economic Rate of Return Economic Rate Sensitivity /a of Return (%) Base ERR - without contingency 17 - with physical contingency 17 25% higher capital cost 15 No infrastructure cost 18 25% higher infrastructure costs 16 50% higher labor costs (equivalent to no shadow pricing) 14 50% higher net benefits from without project situation 16 25% higher operating cost for dairy farms 16 25% lower milk price 16 25% lower yield from pepper 15 25% lower price for pepper 15 25% lower yield from forestry 16 25% lower prices for forestry products 16 1-year delay in yields 14 2-year delay in yields 13 /a All sensitivities are done on base ERR (with physical contingency) figure. 9.14 Though most of the economic risks associated with changes in indi- vidual commodity price levels and yields have been reduced through diversified project design, the risk due to delay in realization of project benefit remains. 1/ If there is a one--year delay in realization of project benefits due to slower than anticipated rate of soil rehabilitation, the ERR would be 14%. This level of ERR is still reasonable and therefore the risk is accep- table. 9.15 The project also attempts to reduce social risks. All of the set- tlers would come either from estates or villages in the project area. There- fore, the potentially disruptive psychological and social impact of resettle- ment would be minimized. However, a different type of social risk exists. The project, by hastening the removal of marginal tea would also hasten the unemployment of tea estate workers. Though the project would provide land to a larger number of families than those losing estate employment (para 9.02), 1/ If project implementation is delayed, so that both costs and benefits are delayed, the ERR would not change. However, the ERR would suffer if only benefits are delayed due to, for instance, slower than antici- pated pace of soil rehabilitations even though project execution may have been correct and timely.  - 40 - not all tea estate workers would be eligible to become settlers since under Sri Lankan law, only citizens would be eligible and a substantial proportion of estate workers include those whose citizenship status is as yet undetermined. Since they also happen to belong to the cohesive Tamil minority, there is a risk of social friction. Assurances have been receivetd that GOSL would make arrangements satisfactory to IDA for those workers not. eligible for settle- ment under the project (see para 5.10). X. AGREEMENTS REACHED AND RECOMMENDATIONS 10.01 During negotiations the following major assurances were obtained: (a) To safeguard against waterborne diseases assurances were obtained from GOSL that all wells would be pro- tected by covers and that handpumps for raising water would be installed. For all water tanks assurances were obtained that GOSL would arrange for regular disinfection with common bleaching powder (4.19); (b) All staff provided under this project to SCD, FD and PTU would work exclusively on the project (5.08); (c) GOSL would take all measures possible to provide, either through relocation to other estates, or through acceler- ated repatriation, relief to those workers not eligible for settlement under the project (5.10); (d) GOSL would use a two-stage settler selection procedure acceptable to IDA (5.11); (e) Arrangements would be made between GOSL and. NADSA to guarantee the transport and processing of g;reen leaf from the project area (5.12); (f) GOSL would adopt land allocation procedures satisfac- tory to IDA (5.14); (g) GOSL would establish DCS with by-laws satisfactory to IDA not later than July 31, 1981, before dairy animals would be sold to farmers (5.17); (h) NADSA would employ consultants with qualifications and on terms and conditions acceptable to IDA (5.18); (i) NADSA would collect income data of settlers at least twice a year and that if family incomes were below Rs 3,600 in constant 1978 Rupees, GOSL would promptly take necessary measures, acceptable to IDA., to ensure that settlers have the financial ability to continue to develop their farms (5.20);  - 41 - (j) In project year 5 and 6 NADSA would continue to organize and remunerate settlers for training of pepper vines and necessary weeding, fertilizing and pruning for all crops or settlement farms in the project area (5.21); (k) GOSL would start a registration scheme for cardamom growers in the project area and that GOSL would ensure that Government-owned cardamom estates in the project area would provide a processing and marketing channel for the cardamom producers (7.02); (1) GOSL would adopt cost recovery proposals satisfactory to IDA (8.07). 10.02 During negotiations it was agreed that transfer of exclusive con- trol of land covered by the project area to NADSA would be a condition of credit effectiveness (5.09). 10.03 With the indicated assurances, the project is suitable for an IDA Credit of US$4.5 M to the Government of Sri Lanka.  ANNEX 1 SRI LANKA TREE CROP DIVERSIFICATION PROJECT (TEA) I Cost Tables, Expenditure and Disbursement Schedules, and Allocation of Credit Proceeds Table 1, Pages 1-3 NADSA - Capital and Operating Costs Table 2, Pages 1-2 Soil Conservation Costs Table 3 Surveying Costs Table 4 Social Infrastructure Costs Table 5, Page 1 Soil Conditioning Costs " EPage 2 Farm Development Costs (HAMTF and LAMTF) to Page 3 Cardamom Farm Development Costs Table 6, Page 1 Dairy Component Costs of Page 2 Pasture Development Costs Table 7 - Forestry Component Costs Table 8 - Estimated Schedule of Expenditures Table 9 - Estimated Schedule of Disbursements Table 10 - Proposed Allocation of Credit Proceeds  SRI LANKA TREE CROP DIVERSIFICATION PROJECT (TEA) I NADSA - Capital and Operating Costs (hTousand Rupees) - - - - - - - - - - - Y E A R S - - - - - - - - - - Unit - - 1 - - -- 2 -- - - - - - 4 - - Total Foreign Cost No. Cost No. Cost No. Cost No. Cost Exch.(%) I. CAPITAL COSTS A. Office Equipment & Furniture 1/ Headquarters 150 1 150 - - - - - - 150 - Field Offices 21 3 63 - - - - - - 63 - B. Vehicles (i) Headquarters Car 110 1 110 - _ - _ _ 110 55 4-Wheel Drive 150 3 450 - - - - - - 450 65 Truck 110 5 550 - - - - - - 550 90 Motorcycle 15 2 30 - - - - - - 30 60 (ii) Field Level a) 2 Area Operations Divisions 4-Wheel Drive 150 2 300 - - - - - - 300 65 Motorcycle 15 25 375 - - - - - - 375 60 Tractor 120 25 3,000 - - - - - - 3,000 95 Trailer 23 50 1,150 - _ _ _ _ - 1,150 35 b) Nursery Unit Tractor 120 2 240 - - - - - - 240 95 Trailer 23 4 92 - - - - _ _ 92 35 Motorcycle 15 2 30 - - - - - - q0 C. Minor Equipment 2/ (Tools for Settler-Labor)- 0.4 4478 1,791 - - - - 1,791 20 Sub-Total(A+B+C) 8,331 - - - 8,331 62 Physical Contingencies on 19 A+B+C (10%) 833 - _ _ 833 Price Contingencies 458 - - _ 458 62 TOTAL 9,622 - - - 9,622 58 1/ Includes typewriters, furniture, duplicating machlnes, calculators, etc. 2/ Rs 400 per settler family for farming equipment (spades, shovels, baskets, X X rakes, knives, wheelbarrows, sickles, secateurs, hoes, buckets, mattocks),