Tanzania - Flue-cured Tobacco Project

RESTRICTED Report No. PA-4Za This report was prepared for uso within the Bank and its affiliated organizations. They do not accept responsibility for jts accuracy or completeness. The report may not be published nor may it be quoted as representing their views. INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT INTIERNATIONAL DEVELOPMENT ASSOCIATION FLUE-CURED TOBACCO PROJECT TANZANIA September 16, 1970 Agriculture Projects Department  CURatENCY EQUIVALENTS US$ 1 - Tan2ania Shillings (T Sh) 7.14 T Sh 1 US$ 0.14 1 Xillion T Sh - US$ 140,000 W5MHT AND MASURIES (British 9ystem) 1 acre (ac) - 0.405 hectare 1 inch (in) a 2.54 centimeters I mile (mi) - 1,609 meters 1 squazm mile (sq mi) - 2.59 square kilometers 1 pound (lb) - 0.454 kilogram 2,204.6 pounds - 1 metric ton ABBREVIATIONS NDCA: National Development Credit Agency TTB: Tanganyika Tobacco Board TTPC: Tanzania Tobacco Processing Company  TANZANIA FLUE-CURED TOBACCO PROJECT TABLE OF CONTENTS Page No. SUMMARY AND CONCLUSIONS ................................. i I. INTRODUCTION ......... ................................... I II. BACKGROUND .......... .................................... 1 A. Agriculture ......................................... 1 B. The Tobacco Sector .................................. 2 C. Tobacco Development Policy .......................... 3 III. THE PROJECT AREA .4 A. Location and Transportation. 4 B. Climate, Soils and Vegetation. 4 C. Population and Land Tenure. 4 D. Production of Flue-Cured Tobacco. 5 IV. THE PROJECT ......... .................................... 5 A. Project Definition .......................,., 5 B. Field Development . 6 C. Storage, Auction and Processing Facilities .... ...... 10 D. Cost Estimates ...................................... 12 E. Proposed Financing .................................. 13 V. ORGANIZATION AND MANAGEMENT ............................. 15 A. General ......... .................................... 15 B. Field Staff, Research and Training .... .............. 16 C. Marketing and Processing ............................ 17 D. National Development Credit Agency .... .............. 19 E. Cooperatives ......................... 22 F. Accounts and Audit .23 This report is based on the findings of an appraisal mission to Tanzania in October/November 1969, composed of Messrs. S. D. Eccles, R. J. Dewar, J. Willems (IDA) and G. R. Henderson (Consultant). 1Mr. Hassid (IDA) contributed materially concerning credit aspects.  TABLE OF CONTENTS (Continued) Page No. VI. PRODUCTION, MARKETING AND FINANCIAL RESULTS .... ......... 23 A. Yields and Production ............................... 23 B. Markets and Prices .................................. 24 C. Financial Results .................................... 25 VII. BENEFITS AND JUSTIFICATION ..27 VIII. RECOMMENDATIONS ....................... , ......... 28 iNNEXES 1. Types of Tobacco and Production Methods 2. Organization Chart 3. Project Phasing 4. Project Costs 5. Disbursement Schedule 6. Balance Sheets - Tanganyika Tobacco Board and Tanzania Tobacco Processing Company 7. The National Development Credit Agency (NDCA) 8. Marketing Prospects for Flue-Cured Tobacco 9. Farm Price Calculations 10. Farm Budget 11. Government, NDCA and Tobacco Board Cash Flow Projections 12. Economic Rate of Return Calculation and Sensitivity Analysis MAP  TANZANIA FLUE-CURED TOBACCO PROJECT SUMMARY AND CONCLUSIONS i. This report appraises a project to increase production of flue- cured tobacco by some 20 million lb by 1978, when total production is ex- pected to reach some 45 million lb. A credit of US$ 9.0 million is pro- posed to cover some 60% of project costs over the six years 1970/71-1975/ 76 which are estimated at US$ 14.7 million, of which US$ 7.4 million would be in foreign exchange. In the three years following disbursement of the credit, additional investments of some US$ 1.6 million, financed within Tanzania, would be needed to bring the project to full development. ii. Agricultural exports amount to about 80% of total exports of about US$ 225 million, with cotton, coffee and sisal the most important; fluctua- tions in the prices of these crops stress the necessity to diversify produc- tion. The project would increase the value of flue-cured tobacco exports by some US$ 11.5 million by 1978, and would yield an estimated economic rate of return of about 30%. iii. The main components of the project are: (a) an increase in the number of small farmers growing tobacco from 15,000 to 30,000 in 150 new village communities and cultivating, when settlement is complete, 30,000 additional acres of tobacco; (b) seasonal and medium term credit facilities, for farm inputs and the construction of curing barns and grading sheds; (c) a road access system, a reliable water supply for both domestic and farm use, and a baling shed for each village; (d) establishment of a project organization and an adequate extension service; and (e) introduction of a system of tobacco auctions and construction of auction floors and additional to- bacco storage and processing facilities. iv. The project would be situated in woodland areas of western Tan- zania where alternative development possibilities are presently negligible. Some 15 cooperative societies, each with about 1,000 farmers and with train- ed cooperative staff to assist in their management, would be established to act as channels for farm credit and distribution of inputs, and to super- vise baling. Extension staff would assist farmers in improving methods of cultivation and systems of tobacco curing, leading to higher yields of bet- ter quality tobacco.  v. Processing is carried out by the Tanzania Tobacco Processing Com- pany (TTPC) under the control of the Tanganyika Tobacco Board (TTB), which is responsible for marketing. The project includes a survey to determine the slze and timing of construction of the auction sales floor and addition- al storage and processing facilities, and to investigate the organizat4onal implications of the introduction of the auction system. Withdrawal of USS 1.5 million of the proposed credit for the construction of these facil- ities would be contingent on the result of this survey. vi. Funds disbursed from the proposed credit for use by farmers, co- operative societies, TTB and TTPC would be channeled through the National Development Credit Agency (NDCA), a parastatal body. NDCA's performance in the flue-cured tobacco sector to date has been satisfactory, while its overall performance has only been fair. NDCA is executing agency for IDA Credit 80-TA (US$ 5 million) of 1966, the only financial assistance aiven to the agriculture sector in Tanzania by the Bank or IDA other than IDA Credit 132-TA (US$ 1.3 million) of 1969, which was for the development of cattle ranches. NDCA will be re-formed into a Rural Development Bank in the near future; discussions between Government and IDA are in progress and the proposed re-formation should not adversely affect the project. vii. Other funds disbursed from the proposed credit would be used directly by the Ministry of Agriculture, Food and Cooperatives which has overall responsibility for the project. Project coordinating committees would be established at both national and regional levels. viii. Procurement of farm inputs, Government vehicles and most process- ing machinery and the letting of major civil works contracts would follow international competitive bidding procedures. Minor civil works would be procured following local bidding. Machinery for expansion of an existing processing line would be procured on the basis of negotiated contracts. ix. World demand for tobacco is rising at about 2% per annum and sanc- tions against Rhodesia have provided an opportunity for the smaller producers of leaf of a satisfactory quality, such as Tanzania, to expand production to enter the international market. Tanzania's share of world flue-cured tobacco exports in 1969 was about 2% (13 million lb); it is expected to rise to about 4% (42 million lb) by 1978. The project would account for almost 70% of this increase. x. The project is suitable for an IDA Credit of US$ 9.0 million. The balance of finance would be provided by Government, from budgetary sources (US$4.8 million), and by the Tanganyika Tobacco Board, from trading surpluses (US$ 0.9 million).  TANZANIA FLUE-CURED TOBACCO PROJECT I. INTRODUCTION 1.01 The Government of Tanzania has requested an IDA credit to help finance flue-cured tobacco development. The project includes the estab- lishment of some 15,000 new growers under a 5 year settlement program, the planting of some 30,000 acres of tobacco, the introduction of an auc- tion marketing system, and the extension of processing and storage facil- ities. The project was prepared by the Government in 1968/9, with assis- tance from the Bank's Permanent Mission in East Africa and the Agricultur- al Development Service. This appraisal report is based on the prepation report and the findings of a mission to Tanzania in October/November 1969, composed of Messrs. S. D. Eccles, R. J. Dewar, J. Willems (IDA) and G. R. Henderson (Consultant). Mr. M. Hassid (IDA) contributed materially con- cerning credit aspects. 1.02 Previous IDA credits to Tanzania for agriculture have been US$5 million in 1966 for a smallholder credit project (Credit 80-TA) and US$1.3 million in 1969 for the development of five government-owned beef cattle ranches (Credit 132-TA). Performance under the credit project has only been fair, although there have been recent improvements. The execu- ting agent is the National Development Credit Agency (NDCA), which would act as the credit channel for the presently proposed project and whose performance is analysed further in Chapter V. The livestock project is developing satisfactorily, although there have been delays caused by a change in the constitution of the financial channel (National Development Corporation) and in obtaining long-term leases for the land for some of the ranches. II. BACKGROUND A. Agriculture 2.01 Mainland Tanzania has a land area of 341,150 sq mi and a variety of ecological and climatic conditions. The country has a few principal regions: the coastal belt (sisal and cashew nuts); a fertile belt around Lake Victoria (cotton, tea and coffee); an area around Mount Kilimanjaro (coffee); and the Southern Highlands (tea and pyrethrum). The remainder comprises "miombo" woodland with little known economic development poten- tial apart from tobacco cultivation and cattle rangeland. Most of the po- pulation of about 13 million, which is increasing at about 2.7% per annum, live at subsistence levels with national income per capita about US$ 62.  2.02 Tanzania's economy depends mainly on agriculture: it accounts for about 50% of GDP, generates about 80% of exports totalling US$ 225 million per annum and provides a livelihood for about 90% of the population. Per- formance of the agricultural sector over the period 1960-68 has been impres- sive but uneven, in terms both of commodities and geographical location. Tobacco exports grew at about 18%, pyrethrum at 16%, cashews at 13Z, coffee and tea at 10% and cotton at 6%. Sisal production declined by 6. by weight, but its export value in 1968 was only slightly more than one-half of the 1960/62 average. Tanzanian strategy for the sector is to concentrate scarce resources, such as skilled manpower, into crops or areas of high potential; and to encourage the rural population to live in villages, as an efficient means of providing social and technical services while encouraging coopera- tive effort and self-help. The project fits in well with this overall strat- egy. B. The Tobacco Sector 2.03 Flue-cured tobacco growing was started in the early 1950's mainly by expatriates. Information on flue-cured tobacco and production methods is given in Annex 1. Until Rhodesia's unilateral declaration of independence in 1965, the crop was sold for local cigarette manufacture, but thereafter export markets opened up and production rose sharply to an anticipated 17 million lb by 1970. Exports in 1968 worth some T Sh 30 million (US$ 4 mil- lion) accounted for about 2% of the national total. Production is concen- trated in the miombo woodland of the Tabora and Mbeya Regions (see Map) - the project area - and the Iringa Region, where woodfuel availability will limit further development for some time. Government has also started a pi- lot program in the Tunduru District. 2.04 This expansion has been largely brought about by Government pro- grams to help small farmers, such as settlement and supervised seasonal credit, and about 75% of the crop is now produced by some 15,000 family farms. Seasonal credit arrangements - through cooperatives with funds ob- tained from the National Development Credit Agency, or through the now de- funct Ministry of Settlement - have been generally satisfactory, in regards to both timely delivery of inputs and loan recovery, now that the latter is effected by deductions from sales proceeds. However, lack of planning has often led to tobacco growing areas being located without adequate fuelwood reserves, and some farmers are already being forced to move to new land. The intensity and quality of technical advisory services has been very var- ied, contributing to sub-optimum yields and losses in handling, curing, grading and baling. These losses have been exacerbated by the lack of medium-term credit to assist farmers in constructing weather-proof grading sheds. 2.05 Fire-cured tobacco, details of which are given in Annex 1, is also produced in Tanzania but quantities are relatively small with 1968 produc- tion valued at about T Sh 5 million (US$ 0.7 million). Aromatic tobacco was grown in Tanzania until 1968 when production ceased due to quality and mar- keting difficulties.  2.06 Marketing of all tobacco is the monopoly of the Tanganyika To- bacco Board (TTB) described in Chapter V. Two different expatriate com- panies are currently used for export sales of flue-cured and fire-cured tobacco, respectively, while TTB itself handles sales to BAT (Tanzania) Ltd., a Darastatal body which has a monopoly of local cigarette produc- tion. C. Tobacco Development Policy 2.07 Government policy is to take advantage of the present world mar- ket situation for flue-cured tobacco, and the good quality and competitive cost of Tanzanian production, to more than double national production by the late 1970's through increased production by small farmers. In carry- ing out this policy, Government intends to take full account of the lessons learnt over the past five years. The Ministry of Agriculture, Food and Co- operatives has thus assumed sole responsibility for all tobacco development and would provide an intensified supervisory service aimed at improving methods of cultivation, grading, curing and baling. Development would be concentrated in new villages where common services can be provided at mini- mum costs and adequate fuel wood reserves can be set aside. Farmers would be provided with medium term credit to build more adequate barns and grad- ing sheds, as well as the currently available seasonal credit, but would only be assisted to develop a tobacco acreage that could be handled with- out reliance on hired labor i.e. about two acres per farmer. 2.08 The Government also intends to follow the principles of the 1967 Arusha Declaration, in which the Government party, the Tanganyika African National Union (TANU), formulated its development strategy, and develop the traditional African system of group cooperation (the "Ujamaa" or "extended family" principle), rather than concentrate on individual entrepreneurs. This involves giving priority assistance to those groups of small farmers who demonstrate self-reliance and a capacity for self-help and who engage in communal development activity such as road and school construction. The principle appears to be well understood and acceptable to the rural popula- tion, partly because of the effective decentralization of responsibility for the carrying out of development programs to the Regional Commissioners - Presidential representatives with wide powers within the framework of na- tional policy established by the central Government. 2.09 New tobacco growers would be settled in Ujamaa villages. In its most developed form, an Ujamaa village would be a complete production coop- erative. A typical intermediate form would involve cultivation of a single large block of land, but with each farmer responsible for a particular part of it. This intermediate form - which is expected to be followed by most project villages - makes it easier for the technical extension services, for transportation of inputs and crop, and for any communal work that might be agreed on, while at the same time leaving the villagers with a suitable de- gree of individual responsibility. In its least developed form, villagers would continue to cultivate individually but would get together, for example,  - 4 - to construct a school or a dispensary or to maintain the local roads. The Government is not forcing the pace, however, and encourages each village to find its own level. The Government rewards progress by arranging for water supplies, medical and school facilities and improved access roads for vil- lages which show a cooperative and self-reliant attitude. IlI. THE PROJECT AREA A. Location and Transj?ortation 3.01 The project is located in the woodland areas of western Tanzania, in the Tabora and Mpanda Districts of the Tabora Region and the Chunya Dis- trict of the Mbeya Region. The towns of Tabora and Mpanda are linked by rail to Morogoro, where the processing facilities are situated, and to the port of Dar es Salaam. The town of Chunya is linked to Morogoro and Dar es Salaam via an East African Railways road/rail service operating through Itigi on the Morogoro-Tabora rail line. Tabora, which will be the site of project headquarters, is also linked to Chunya and Mpanda by road. East African Airways operate regular services from Dar es Salaam to Tabora and Mbeya (some 50 miles from Chunya); and Morogoro, Chunya and Mpanda have airstrips used by charter flights. B. Climate, Soils and Vegetation 3.02 Rainfall in the project area ranges between 29 and 50 in per an- num, which together with light, sandy soils make ideal conditions for tobac- co production. Vegetation consists of open "miombo" woodland, characterized by Brachystegia and Julbernadia tree species, and containing few trees of value except "muninga" (Pterocarpus angolensis), an excellent furniture tim- ber. The miombo woodland is infested with tsetse fly. Cattle do not there- fore play any significant role, and development, other than for tobacco pro- duction, has been confined to exploitation of muninga and the collection of honey and beeswax. C. Population and Land Tenure 3.03 Population densities are very low ranging from about two persons/ sq mi in Mpanda, to about five in Chunya and eight in Tabora. There is no pressure on land and no problems are anticipated in this regard for settle- ment. Land in the project area, whether occupied or unoccupied, is public land and is held by its occupants under a traditional system of usufruct. Under customary law land is allocated to individual farmers by village el- ders who also have the power to cancel rights of occupancy in the event of failure to comply with any conditions pertaining to the use of land. In an  Ujumaa village settlement program the powers of elders would be assumed by the village cooperative committee. Past experience has shown that this sys- tem of control can be effective. D. Production of Flue-Cured Tobacco 3.04 In 1969 Tabora District produced 7.8 million lb of flue-cured to- bacco, Mpanda District 1.5 million lb and Chunya District 0.2 million lb. Farmers are organized into cooperatives, through which inputs are distrib- uted on credit. In each village there is a cooperative baling center, at which the Tanganyika Tobacco Board classifies and purchases the crop. 3.05 Tobacco is grown in a five-year rotation alternating with a food crop (mainly maize) and fallow. Tobacco seed is obtained from the Research Station at Tumbi, near Tabora, and sown in beds in September. Seedlings are transplanted into the field at the beginning of the rains in November. Harvesting starts in February and is completed by March. Seed beds are fu- migated and receive applications of fertilizer and pesticides; after trans- planting a compound fertilizer is applied in the field. Nitrogenous fertil- izers are commonly used on maize. 3.06 The average farmer produces a crop with a potential yield of about 1,000 lb of wet, cured leaf Der acre but an analysis of harvesting, curing and subseqtuent grading has shown that the average small farmer loses about 10% of his crop in the field due to lack of barn capacity, about 10% in the barn through poor curing techniques, and another 25% in inefficient bulking and grading operations, mainly because of leaking roofs and poor construction (or non-existence) of his grading shed. Consequently the average small farmer produces only between 550 and 600 lb/acre of saleable tobacco. These yields compare to those of larger commercial farms in the Iringa Region now averaging just under 1,000 lb/acre and the national average of 572 lb in 1967/68 and 640 lb in 1969/70. IV. THE PROJECT A. Pr 4.01 The project forms part of the Government of Tanzania's flue-cured tobacco development program, over 1970/71 - 1975/76, in the Tabora, Mpanda and Chunya Districts (see Map). It involves: - settlement of about 15,000 farmers, over 1970/71-1974/75, in about 150 Ujamaa villages, arranged in complexes of about 10 villages each, with each farmer cultivating 1 ac of tobacco in his first year and 2 ac thereafter;  - establishment of cooperatives, at the rate of about one per complex of 10 villages, to supervise the baling of cured to- bacco and to handle the distribution of credit for on-farm seasonal inputs, curing barns and grading sheds; - construction of communal infrastructure, including some 225 mi of extraction roads, some 750 mi of access tracks, vil- lage water supplies and baling centers: - establishment of a "Tobacco Project Group" within the Min- istry of Agriculture to plan and manage the project; pro- vision of extension services for farmers and cooperatives, staff for the Tumbi Research Station, a Chief Marketing Of- ficer and a Chief Accountant for the Tanganyika Tobacco Board, a Financial Analyst for the National Development Credit Agency and related infrastructure, housing, vehicles and equipment; - carrying out a tobacco marketing and handling study and con- struction of auction floors, additional storage and process- ing facilities; - aerial survey of parts of the project area. In addition, project farmers would build their own houses and establish subsistence gardens, and would engage communally in the building of dis- pensaries and schools. 4.02 The Ministry of Agriculture, Food and Cooperatives would have overall responsibility for the project and would execute its field aspects. The National Development Credit Agency would provide credit through the co- operative movement. Processing would be carried out by the Tanzania Tobac- co Processing Company, under the control of the Tanganyika Tobacco Board, which handles all marketing. These institutions, and means of coordinating their activities, are described in Chapter V 'Organization and Management". An Organization Chart appears at Annex 2. B. Field Development Phasing 4.03 Development of villages and complexes, settlement of farmers and incremental planting of tobacco would be phased approximately as follows:  - 7 - 1970/71 1971/72 1972/73 1973/74 1974/75 1975/76 1976/77 Total Complexes established / 3 6 4 2 - - - 15 Villages I established /2 18 33 33 33 33 - - 150 Farmers settled 1,800 3,300 3,300 3,300 3,300 - - 15,000 Incremental tobacco acreage - 1,800 5,100 6,600 6,600 6,600 3,300 30,000 /1 Planning of the layout of the complex of about 10 villages, construc- tion of crop extraction roads and complex headquarter buildings, staf- fing of complex headquarters and establishment of the cooperative. /2 Planning and demarcation of each village within the complex, construc- tion of access tracks, water supply, extension service house and baling center. This phasing takes into account the availability of extension service staff and the capacity of project planning staff. Settler availability is not ex- pected to be a constraint, nor is the capacity of the marketing system. De- velopment would be approximately equally divided between Tabora, Mpanda and Chunya Districts, see Annex 3. 4.04 Assurances have been obtained during negotiations that no farmer would be allowed to grow more than two acres of tobacco in any year unless the complex field officer has satisfied himself that the farmer could han- dle the increased crop; that no non-project flue-cured tobacco development, other than completion or improvement of existing schemes, would be carried out in the project area; and that the annual targets for settlement of new farmers on project schemes would not be increased unless Government can provide all the necessary additional services, with any such increases beyond 20% requiring IDA's prior approval. Complex and Village Planning 4.05 Complex and village planning would be carried out by survey and planning staff, whose equipment, vehicles and operating costs would be fi- nanced under the project. Parts of the project area have already been aerial photographed, but some additional aerial survey work would need to be done by an experienced commercial firm under contract. Planning, in- cluding aerial surveys, is expected to start prior to loan signature.  -8- 4.06 Villages would normally accommodate about 100 farmers, a conve- nient number for services and giving a manageable village size of some 8 sq mi. Each village would encompass about 50 ac per farmer, including 10 ac of arable land suitable for tobacco (i.e. 1,000 ac per village), 30 ac of mature woodland to provide fuelwood for the curing barns and an average of 10 ac of non-arable land, such as rocky outcrops. The arable land would be farmed in the rotation pattern described in para 4.12. Adequate land is available and no particular difficulty is foreseen in locating com- plexes, each of about 10 villages, within a reasonable distance of water supply and existing roads. 4.07 Present techniques of curing, and the type of barns currently in use, make inefficient use of fuelwood. At present, production from one acre of mature woodland is required to cure an acre of tobacco; the 30 ac of woodland set aside per farmer would therefore only be sufficient for 15 vears. However, it is anticipated that within five years, experimentation would lead to improvements in barn design and curing techniques, which would reduce wood fuel consumption, and to the establishment of small blocks of high-yielding fuelwood plantations. The experimentation program would be fi- nanced under the project (see para 4.16). Infrastructure 4.08 Government would provide each village with a water supply for domestic and seedbed needs, access tracks and crop extraction roads. De- pending on location, the water supply would be a shallow well, borehole or 'charco" (small dam). The cheapest adequate supply (shallow well) would be used wherever possible as these can be constructed largely by self-help means. On average, about 5 mi of access tracks for each village and about 15 mi of crop extraction roads for each complex would be needed to link up with the existing road network. An Earth Moving Unit consist- ing of a crawler tractor, motor grader, five-ton truck, personnel vehicles and two caravans would be attached to project headquarters. It would have the capacity to construct all the necessary roads and up to 30 charcos. The necessary equlinent and experience for drilling boreholes is available locally. Maintenance of roads and water supplies would be organized on a self-help basis. 4.09 Government would build houses and related buildings for its staff, make available vehicles at the village and complex level and provide each village with a simple baling shed for use by farmers under the control of the cooperative. Vehicles, both personnel and load carrying, would be pro- vided, and the latter would be available for hire by farmers through their cooperatives. As each cooperative became well established, however, it would be encouraged to purchase and operate its own vehicles (see para. 5.19). 4.10 One acre of land in a central block would be allocated to each family and, in the first year of settlement, each farmer would construct his house, his grading shed and his first curing barn. A further two curing barns would be constructed in the following two years. The house would be completed without cash expenditure, but the grading shed would  need a corrugated iron roof (with some corrugated perspex sheets for natural lighting) and a concrete floor to give the farmer a satisfactory place to sort and grade his leaf without the considerable losses through dampness and incorrect grading which are now prevalent (see para 3.06). Each curing barn would need galvanized sheeting for flue-pipes and a barn thermometer. These cash expenditure items amount to about T Sh 1,280 (US$ 179) per farm. As a result of the proposed experimentation on barn design and curing techniques (see para. 4.16), smaller numbers of larger barns might become the rule during development of the project; and some villages may elect to construct communal, rather than individual, grading sheds. 4.11 Farmers could be expected to construct a dispensary and a school room at an early stage of each village settlement, on the basis of self- help. In accordance with the policy described in para 2.09, Government would then be prepared to provide a nurse and a school teacher. No finance for these items is included in the project. Village Settlement and Farming 4.12 Most villages would probably assume an intermediate form of the Ujamaa village concept, as described in para 2.09. Each year, the coop- erative would select sites for seed beds and the 200 acre block on which the tobacco would be grown; within these each farm family would be allo- cated a seed bed and two acres for tobacco (one acre in the first year). The 200 acre block would'be selected from the 1,000 acres of arable land allocated to the village and farmed on a 5-year rotation, with a maize crop in the second (and sometimes third) year and with a period of fallow before a second tobacco crop is grown. 4.13 No problems are envisaged in obtaining adequate numbers of suit- able project participants. With the advent of tobacco growing in recent years there has been some shift of local population to suitable growing areas, and also an increasing influx of people, both as new growers and laborers - many of whom become growers - from other parts of the country. It is anticipated that this trend will continue. The project should prove attractive to people now living in overcrowded areas, such as around Tukuyu and Kilimanjaro, and to labor now unemployed due to the closure of sisal estates. No formal selection procedures would be used, it being as- sumed that persons prepared to move on their own account would have the basic qualities needed. Seasonal Production Credits 4.14 Project participants would obtain seasonal production credit of about T Sh 385 (US$ 54) per acre of tobacco. This would be distributed in kind, comprising a seed pack containing seed bed fumigants, fertilizers and insecticides; compound fertilizer and insecticides for tobacco; nitro- genous fertilizer for maize; and flue-pipe replacements for curing barns. In addition, cooperatives would receive seasonal credit of about T Sh 35  - 10 - (US$ 5) per acre of tobacco to meet baling costs. Project costs include provision only for incremental increases in the level of seasonal credits. Project Administration and Extension Staff 4.15 Experience in Tanzania and elsewhere indicates that it takes at least five years of concentrated attention by Government administrative and extension staff to ensure the success of a settlement scheme, particu- larly one based on a non-traditional crop. The costs of administrative and extension services have therefore been considered as investment costs for the initial five-year period for each settlement. Research and Training 4.16 While research results from other countries have been adequate for Tanzania's needs in the past, and would continue to be so for the im- mediate future, research programs directed at Tanzania's specific problems should now be implemented. Under the project, a Government Tobacco Re- search Station would be established at Tumbi on land owned by Government and partially developed. The Station would be staffed with five research officers and equipped with houses, laboratories and equipment. An outline research program would be submitted for IDA's approval within one year of credit signature. This program would be concerned primarily with practical investigations, including: husbandry techniques; rotations; pest control; varieties - in particular low nicotine varieties; barn design and curing techniques, especially the efficiency of different fuels; fuelwood species trials; and the testing of other cash and food crops for introduction into tobacco growing areas. Training facilities would also be developed at Tumbi under the project, including accommodation for project staff to attend short courses in tobacco production. C. Storage, Auction and Processing Facilities 4.17 The project would include a study of the storage, auction and processing facilities needed to handle the increased production that would result from Tanzania's flue-cured tobacco development program. The study is expected to start prior to loan signature. It would cover the determin- ation of the capacities and locations of needed storage, auction, and pro- cessing facilities; the production of preliminary designs and cost esti- mates for these facilities, together with schedules for their construction; and the organization, managerial, and training arrangements for their ef- fective operation. The study would be conducted by a team whose members, terms of reference and conditions of employment would be subject to IDA's approval. During negotiations, assurances were obtained that the team's findings and recommendations would be discussed promptly with IDA and would not be implemented until agreement thereon is reached with IDA.  - 11 - Storage 4.18 The Tanganyika Tobacco Board (TTB) has the use of facilities for storing 4 million lb of wet, cured tobacco in the project area, while the Tanzania Tobacco Processing Company (TTPC) can store a further 6 million lb at its processing facilities in Morogoro. It has been tentatively esti- mated that additional storage for 5 million lb may be needed both in the project area and at Morogoro. Provision of T Sh 6.0 million (US$0.8 mil- lion) has been made in project costs for its construction. Processing 4.19 Processing involves bulking, in a blend specified by the buyer, of various grades of wet, cured leaf tobacco; redrying; machine threshing to remove the lamina of the leaf from the stem; and packing. TTPC's pre- sent plant at Morogoro has a monthly capacity of 2.5 million lb of wet, cured leaf and is barely adequate to cope with current production. Pro- vision of T Sh 4.0 million (US$ 0.6 million) is made under the project for doubling the present facilities by March 1971. Preliminary estimates indicate that an expansion to 10 million lb/month would be required by 1973, and provision of a further T Sh 10.0 million (US$ 1.4 million) has been made in project costs. Auction Floor 4.20 Tanzania's expected production in 1970 of about 20 million lb of wet, cured leaf is probably approaching the maximum that can be graded, classified and marketed efficiently through an agency arrangement such as is performed by the Trans-Continental Leaf Tobacco Corporation on behalf of TTB (see para 5.06). The anticipated more than doubling of national tobacco production by 1978 makes an alternative to these arrangements es- sential. The alternative is to shift more responsibility to growers and the cooperatives for the satisfactory grading and baling of their leaf and for them to offer it for sale on their own account at an open auction at- tended by international buyers. The key to a successful auction is a suf- ficient quantity of good quality and well presented tobacco to attract in- ternational buyers. The anticipated growth in volume of the Tanzania crop and the improvements in quality expected as a result of the project would meet these conditions. The project therefore includes the construction of an auction floor at Morogoro, tentatively estimated to cost some T Sh 6.0 million (US$ 0.8 million). Since the introduction of the auction system would increase the responsibilities of the TTB an experienced Chief Mar- keting Officer and Chlef Accountant would be appointed and their salaries would be considered as a project cost for a period of five years, as would the cost of training abroad of a Tanzanian national to take over the former post. 4.21 As IDA is financing central storage, processing and auctioning facilities which would be used by all Tanzanian flue-cured tobacco, as- surances were obtained during negotiations that IDA would be consulted annually concerning flue-cured tobacco development outside the project; and that the Tanzania Government would take such actions as might be necessary to maintain the quality of all flue-cured tobacco, including the provision of adequate storage, processing and auctioning facilities.  -12- D. Cost Estimates h .22 Toftal project costs are estUim'Lted at T Sh 105.2 million (US.1 :J'-7 inil.ion), -!ith a fore- n. change coirponent of Tboutt I' Sh 53.1 nmillion (USa` 7.4 1illUioil) . Those eitiinatco a7 e suzi,.arizod in the Trablc below. IXurthcr details are giv, n in Annex 14. SUlIlARY PlOJThOT COSTS F (mr i -n L.ocat Foreig.;n Local Fo ri,n Lgn Currfency Lexchange) Tot..l Currency E7.,Wntr_ Tel 1. Co;nj eIirmt A. Credit* - * - - * - (T{,ll '^(U O ..... ........ . J~ . -. - (K.;, ,< ~, . .......... ~~,<-- A. Credit -~K II7 .( (-J . 14edium-term credit-. 2 6,650 9,930 16.630 930 1.400 2,330 60 Tncreirental seasonaal credit-/ 2,010 8f,540 10,550 280 1.,195 1,!475 81 NIDA Finaricial Analyst3/ 90 270 360 15 35 50 75 B. Settlement Infrastru>cture Rcads, waiser supply, buildings 6,245 4,050 10,295 870 570 L,)thO 39 Vehicles and equippent 230 2,030 2,260 ho 280 320 90 Extension service4/ 18,0iO 2,610 20,620 2,520 365 2,885 13 C. Project Headquarters Salaries and other costs 5,567 3.855 7>422 780 260 1,040 25 Buildings 1,020 3hi0 1,360 140 50 190 25 Vehicles and equipment 23 180 203 5 25 30 89 Aerial survey 50 450 500 5 65 70 90 D. Research and Training Buildings 940 320 1,260 130 45 175 25 Vehicles and equipment 20 220 240 5 30 35 92 Salaries and other costs 3,240 2,360 5,600 450 330 780 42 E. Storage and Hlandling Storage 3,000 3,000 6,000 420 420 840 504 Processing facilities 1,500 12,500 14,000 210 1,750 L,960 89 Auction floors 3,000 3,000 6,000 420 420 840 50 ,arketing study 100 400 500 15 55 70 80 Chief Larketing Officer and Chief Accountant 400 1,000 1,400 55 140 195 71 Total 52,095 3d05 1C 200 I 7'l 90 l7),` -1-75 0 1/ Curing barn3, baling sheds anid tools. 2/ Fertilizer, seeds, insecticides and other materials. 3/ See para 5.13. 3ince this table, and other financial projections in the report, were prepared, it has been decided to include in the project continuation of the employment of tlhe expatriate Chief Accountant for NIDCA, at a total cost of about us$ 50,000 (T Sh 360,000). 4/ During the developiin,nt period only.  - 13 - 4.23 In most cases, estimates are based on recent experience in tobac- co settlement projects in Tanzania, and include 10% contingencies. In the case of storage, auction floor and processing facility extensions - other than that required by 1971 - tentative estimates by the appraisal mission have been used which include 25% contingencies and which will need revision in the light of the results of the study described in para 4.17. E. Proposed Financing 4.24 An IDA credit of US$ 9.0 million (T Sh 64.3 million) would be made to Government, to cover about 60% of expenditures incurred on the project during the six years, 1970/71 - 1975/76. SUMMARY PROPOSED FINANCING T Sh '000 US$ '000 % IDA Credit 64,280 9,000 61 Tobacco Board 6,290 880 6 Government 34,A30 4,850 33 Total 105,200 14,730 100 IDA would finance retroactively certain preinvestment activities under the project: planning and surveying, and the marketing study. The amount is not expected to exceed US$ 50,000, all expenditures having been incurred subsequent to March 15, 1970. 4.25 US$ 7.4 million or 82% of the proposed IDA credit, would cover the estimated foreign exchange costs of the project during the disburse- ment period. The balance of US$ 1.6 million, or 18% of the credit, would finance local currency costs. IDA's share of project financing would be sufficient to ensure the satisfactory organization and execution of the project. The Government's contribution of US$ 4.8 million would come from budgetary sources; while the Tobacco Board's contribution would come from its trading surpluses, see para 5.09. In addition to eventual loan repayments, farmers - who need 100% credit to enable them to participate in the project - would also contribute labor valued at about T Sh 7.5 mil- lion (US$ 10 million). 4.26 The project will give rise to additional investments over the three years following the six-year disbursement period of the proposed IDA credit. These investments involve a continuous build-up of credit needs and project supervision and are estimated to cost some US$ 1.6 million (T Sh 11.5 million). Further details are given in Annex 4. The Government would finance US$ 1.0 million of this amount from budgetary sources, while the remaining US$ 0.6 million would be financed by' NDCA out of principal repayments by farmers, TTB and TTPC of medium and long term loans made at the beginning of the project.  - 14 - Procurement 4.27 In procuring on-farm inputs required by cooperative members, val- ued at about US$ 3.5 nmillion, N)CA would follow its standard policy of in- ternational comnetitive bidding, although Drovision has 'been made in the credit documents for local preference for a national fertilizer manufacturer which is expected to sLart prodtuction in 1971/72. This preference would be the lower of 15% or tihe appropriate customs duty (presently nil). Tihe i4in- istry of Agriculture would procure all its vehicles, including tractors and vehicles for the earth moving unit, through international competitive bid- ding procedures unless, for reasons of standardization and/or service avail- ability, IDA agrees otherwise. The value oF these would amount to about US$ 420,000. Buildings and water supply, valued at about US$ 1.3 million, would be procured by the Ministry following local competitive bidding, be- cause the amounts involved would be too small to attract foreign firms. 4.28 TTPC is procuring the machinery for the extension to its process- ing facilities to be installed prior to 7March 1971 by means of negotiated contracts, following extensive investigationi of the various alternative plant layouts available. Such extensions are estimated to cost about US$ 510,000. In view of TTPC's decision to extend these facilities by in- creasing the capacity of units in the present processing line, internation- al competitive bidding procedures would not have oeen appropriate. For civ- il works involved in constructing further extensions of these facilities, the auction floors and additional storage facilities, both TTB and TTPC would follow international competitive bidding procedures. TTPC would pro- cure machinery for such further extensions through international competi- tive bidding procedures, unless IDA agrees otherwise when the precise na- ture of those extensions becomes known. The total cost of construction and processing machinery is US$ 3.1 million. Disbursement 4.29 It is proposed that the IDA credit be disbursed against the follow- ing, apart from the unallocated contingency element of US$ 1.0 million: (a) 100% of payments out of NDCA medium-term and incremental seasonal loans to cooperatives, and long-term loans for extending processing facil- ities prior to March 1971, and for the market- ing study (US$ 4.0 million). (b) 100% of the cost of certain easily identified items with a high foreign exchange component, namely vehicles, aerial survey and related items, services of expatriate staff and train- ing abroad (US$ 1.3 million). (c) 30% of the cost of certain other easily identi- fiable items, namely operating costs of the  - 15 - earth-moving unit, excluding depreciation; wa- ter supply construction; housing, buildings and furniture; and services of field staff, other than expatriate staff (US$ 1.2 million). (d) 60% of the cost of extending processing facil- ities after March 1, 1971 and of building storage and auction floors (US$ 1.5 million). No disbursements should take place until the marketing study is available and agreement reached with the Government concerning design, cost estimates, organization aspects, procure- ment and disbursement. Disbursement applications would be accompanied by certificates of loans made by NDCA; by receipted bills; and by certificates of expenditure by the project's Senior Accountant as appropriate. Disbursements of the IDA credit are expected to be as follows: DISBURSEMENT SCHEDULE (US$ '000) 1970/71 1971/72 1972/73 1973/74 1974/75 1975/76 950 1,460 1,780 2,170 1,590 1,050 A disbursement schedule by quarters is at Annex 5. V. ORGANIZATION AND MANAGEMENT A. General 5.01 National tobacco development is the responsibility of the Minis- try of Agriculture, Food and Cooperatives in which a Tobacco Section, head- ed by a Senior Agricultural Officer (Tobacco), is to be established to co- ordinate activities. A Tobacco Project Group is to be set up in Tabora to execute the field aspects of the project. Coordination with other inter- ested entities - both in regards to the project and tobacco production gen- erally - would be achieved at the national and regional levels by the es- tablishment of Tobacco Coordinating Committees. An Organization Chart ap- pears in Annex 2.  - 16 - B. Field Staff, Research and Training Tobacco Project Group 5.02 All flue-cured tobacco development in the project area would be the responsibility of a Tobacco Project Group headed by a Tobacco Project Officer. Planning and survey of project complexes and villages would be carried out by a Section headed by a Senior Planning Officer. Another Section - not financed under the project - would be responsible for re- planning existing schemes to try and bring them up to the standards ap- plied to project schemes. Development of schemes would be the responsi- bility of three Sections each headed by an Assistant Tobacco Project Offi- cer - one for the Chunya District, one for the M4panda District and project schemes only in the Tabora District, and the third for existing schemes in the Tabora District. 1/ The accounts for tobacco development in the proj- ect area (including those for the Tumbi. Research Station) would be kept by a Section headed by a Senior Accountant, who would arrange for separate accounts to be kept for project schemes. The Tobacco Project Officer would be assisted, as required, by a Forest Officer, a Water Engineer and a Roads Engineer assigned from the appropriate Government Departments. 5.03 Assurances were obtained during negotiations that the Tobacco Project Officer and Senior Planning Officer would be appointed only fol- lowing IDA's approval of the suitability of candidates' qualifications and experience for the posts; that the Senior Accountant would be appointed only following consultation with IDA: and that the Forest Officer, Water Engineer and Roads Engineer required part time, would be assigned by name and would be made available to the project as a first priority whenever their assistance was requested by the Tobacco Project Officer. Provision nas been made in project financing to cover the appointment of expatriates to two of the full-time posts. Initial appointments to the posts of Tobac- co Project Officer and Senior Planning Officer would be conditions of ef- fectiveness of the credit. Extension Staff 5.04 Each of the proposed 15 complexes would be developed under the responsibility of a resident Field Officer (FO), of a diploma grade, sup- ported by an Assistant Field Officer (AFO), a Cooperative Inspector (CI) (seconded to the complex cooperative), and two Assistant Foresters. Each of the proposed 150 villages would have a resident AFO, assisted during 1/ Separation of project and pre-project schemes in the Chunya and Mpanda Districts would be impractical, in view of the relatively small number of pre-project schemes there. In the Tabora District, however, pre- project schemes would remain the major part of the whole throughout the project period.  - 17 - the first year of settlement by a newly qualified AFO. This rate of staf- fing would provide an appropriate ratio of one agricultural extension offi- cer to 50 farmers during initial development, and of 1 to 100 on a permanent basis. W4hen the project is fully operational by 1975, 15 FOs, 165 AFOs, 15 CIs and 30 Assistant Foresters, plus clerks and stores officers, would be employed. All FOs and AFOs have completed a training course in general agriculture at a Government training institute. On being posted to the project, these officers would attend an eight-months' course of training in tobacco production, curing and grading at the Tumbi Research Station. These numbers of staff are available, but assurances were obtained during negotia- tions that all officers shown to be necessary, together with adequate sup- port staff, would be fully trained and made available to the project. Research and Training 5.05 Tumbi Research Station would employ one Senior Research Officer and four Research Officers for research, and one Field Officer with exten- sive experience of tobacco production, for training, supported by appro- priate junior staff. The Station would receive administrative support from the Tobacco Project Group, but for all professional and research mat- ters it would come under the Director of Research and Training in the Min- istry in Dar es Salaam. Provision has been made in project financing for the employment of four expatriate research officers. Assurances were ob- tained during negotiations that the appointment of the Senior Research Of- ficer and of any Research Officer would be made only following consultation with IDA. C. Marketing and Processing Tanganyika Tobacco Board (TTB) 5.06 Tobacco marketing is the sole responsibility of TTB, a statutory body set up in 1963. Reporting to the Minister of Agriculture, it current- ly consists of a Chairman and 15 other members. The Minister nominates the Chairman and one other member, currently the Managing Director of the TTPC. Five members are the Principal Secretaries of the Ministries of Agriculture, Commerce, Finance, Rural Administration and Economic Affairs. The remain- ing eight members represent both flue-cured and fire-cured growers. TTB's executive staff is headed by a Chief Executive Officer and a Chief Accoun- tant. TTB relies on its sales agents, the Trans-Continental Leaf Tobacco Corporation of Vaduz, Liechtenstein, for classification and purchase of crop in the field, for practically all export sales and for general advice on flue-cured tobacco marketing. Local sales to the monopoly producer of Tanzanian cigarettes, the parastatal BAT (Tanzania) Ltd., are handled di- rectly by TTB. 5.07 TTB currently purchases the crop from cooperatives in the field on the basis of a grade classification. Prices paid for each grade are determined in advance by the Minister of Agriculture on the advice of TTB,  - 18 - and are related to Trans-Continental predictions of likely export prices. After purchase, the wet, cured leaf is transported to Morogoro for re-dry- ing under contract by the Tanzania Tobacco Processing Company (TTPC) (see below). The re-dried tobacco is then offered for sale to BAT and foreign buyers under direct contract. 5.08 Government has decided to adopt an auction system of marketing as soon as feasible (see para 4.20). The auction floor would be built at Morogoro. The organizational implications of the change from direct sales to an auction system are a subject of the study referred to in para 4.17. Some problems are likely to arise in the course of the change-over from direct to auction sales. First, TTB would lose the international market- ing expertise of Trans-Continental. Consequently, as a condition of ef- fectiveness of the proposed credit TTB would employ a Chief Marketing Of- ficer whose qualifications and experience for the post would be satisfac- tory to IDA. Second, TTB has no experience of managing an auction floor. Third, the auction system would require growers and their cooperatives to bale and grade tobacco to internationally acceptable standards which are not being reached at present. Fourth, the auction system would impose much greater organizational and accounting problems on the cooperatives and call for a higher degree of management than is now available to them. To ensure a smooth transition to an auction system and to allow the de- velopment of cooperative management capabilities, the following assurances were obtained during negotiations that, unless otherwise agreed with IDA: (a) TTB would continue its present system of pur- chasing tobacco (for sale across the auction floor on its own account) from those cooper- atives which, in the opinion of the Govern- ment, are not yet capable of selling their crop on their own account; (b) TTB would make arrangements, satisfactory to the Government and IDA, to manage the auction floor, to grade leaf at the cooperative lev- el for presentation on the auction floor, and to train TTB, Ministry of Agriculture and cooperative staff to take over these responsi- bilities. Assurances were also obtained that the posts of Chief Executive Officer and Chief Accountant of TTB would be filled only following consultation with IDA. Provision has been made in project financing for the employment of ex- patriates as Chief Marketing Officer and Chief Accountant, and for the train- ing abroad of a Tanzanian national to take over the former post. 5.09 Balance sheets for TTB since 1965 are summarized at Annex 6. As at December 31, &969, accumulated surpluses or losses on trading account, and the surplus funds from a levy of T Sh 0.05/lb of wet leaf charged for the purposes of meeting TTB overheads, were as follows:  - 19 - T Sh '000 US$ '000 Surplus on flue-cured trading 9,392 1,315 Losa on fire-cured trading (3,300) (462) Loss on aromatic trading (2,132) (298) Surplus on Board levy 501 70 General Fund Total 4,461 625 The surplus on flue-cured trading comes from a TTB policy of withholding T Sh 0.16/lb of wet leaf in addition to direct marketing costs. This with- holding is not a formal tax and the resultant trading surplus is not commit- ted for any particular use. In practice, it finances losses on fire-cured and aromatic tobacco trading, and investments in processing and storage fa- cilities. TTB's working capital needs are provided in the form of crop ad- vances by the National Cooperative Bank, a parastatal body. Tanzania Tobacco Processing Company (TTPC) 5.10 TTPC was set up in 1966 to re-dry flue-cured leaf and prepare it for export. Initially owned on a fifty-fifty basis by TTB and the National Development Corporation, a parastatal industrial finance company, TTPC has recently become wholly owned by TTB. TTPC currently acts as the agents for TTB on a fee basis, although its losses are underwritten by TTB. With the introduction of the auction system, TTPC would act as agents to the indi- vidual purchasers of wet leaf. Balance sheets for TTPC since its inception are shown at Annex 6. 5.11 Technical operations are currently controlled by an expatriate factory manager, seconded from Trans-Continental. Trans-Continental also has a visiting agency agreement with TTPC which terminates automatically on termination of Trans-Continental's sales agency agreement with TTB - a likely consequence of the adoption of the auction system. Assurances were obtained during negotiations that TTPC would at all times employ a factory manager, whose qualifications and experience for the post would be subject to IDA's prior approval; and visiting agents acceptable to, and on terms and conditions satisfactory to, IDA. Assurances were also obtained that the post of TTPC Managing Director would be filled only following consul- tation with IDA. D. National Development Credit Agency 5.12 Credit provided under the project to farmers, cooperatives, TTB and TTPC would be channeled through the National Development Credit Agency (NDCA), formed in 1964 to provide credit for agricultural purposes and enterprlses, including marketing, processing, storing and transporting agricultural products. Although ownership is vested in the cooperative movement, NDCA is under virtual Government control. Government provides  - 20 - the bulk of its funds and its Management Committee is composed largely of Government officials. NDCA's overall financial position is poor, with an accumulated deficit of T Sh 36 million (US$ 5.0 million) as at March 31, 1970. Overdue amounts of principal and interest were then equivalent to some 42% of NDCA's loan portfolio; overdues on loans under IDA Credit 80-TA amounted to some 16% of that portfolio. 1/ These bad debts were largely inherited from NDCA's predecessor organizations and from Government loan schemes, but are also partly due to the general conditions within which NDCA has to operate. INDCA lends mainly to agricultural cooperatives and their unions, whose management is often weak and whose members have to farm often without adequate technical advisory services. NDCA does not have adequate staff to analyze and advise on the financial situation of potential borrowers. Moreover, debt collection often suffers because of difficulties with cooperative management, delays in marketing crops pro- duced and difficulties in collecting loan repayments. Further details about NPCA are given in Annex 7. 5.13 Government has been considering ways and means of correcting the present situation and has decided to re-form NDCA into a Rural Develop- ment Bank, with a revised capital and legal structure and wider respon- sibilities. In preparation for this re-formation, Government has au- thorized NDCA to place some T Sh 26.3 million (US$3.7 million) of bad debts, with a corresponding amount of liabilities, into a suspense account and intends to transfer to the Rural Development Bank only those NDCA as- sets and liabilities which it considers 'live". To further strengthen NDCA's financial position, Government injected an additional T Sh 10.4 million (US$1.5 million) of new capital from the 1969-70 development budg- et, and a further T Sh 12 million (US$1.7 million) has been provided for this purpose in 1970-71. Furthermore, NDCA is in the process of recruit- ing an expatriate financial analyst and it is a condition of effectiveness of the proposed credit that an appointment be made, financed initially under Credit 80-TA. (Provision is made in the tobacco project, however, for continued financing of an expatriate in this post, and of the ex- patriate NDCA Chief Accountant, for about three years after the June 30, 1971 Closing Date for Credit 80-TA. Assurances were obtained during negotiations that NDCA would only appoint persons whose qualifications and experience for the posts had been approved by IDA.) The main means for improving NDCA's performance (and for ensuring the performance of the Rural Development Bank), however, lies in utilizing its credit facilities in projects such as this integrated tobacco project, which includes other essential elements besides credit, such as extension services, assistance to cooperative management and financially attractive farming operations. 5.14 NDCA's performance in the tea and flue-cured tobacco sectors to date has been generally satisfactory. Farm inputs have been delivered to 1/ It should be noted, however, that some loans under Credit 80-TA have been granted recently and are still within the grace period.  - 21 - cooperatives on time, and at very reasonable prices due to NDCA's interna- tional competitive bidding procedures. Credit recovery on tobacco loans has also been good since 1967 when loan repayments began to be deducted on NDCA's behalf by TTB before cooperatives were paid for their crops. In view of this, and the satisfactory accounting procedures within NDCA, it would be a suitable channel for the on-lending of the proposed credit. Since the Government proposes that the new Rural Development Bank should assume the NDCA staff currently involved in the lending program to the tobacco sector and NDCA's general lending policies to that sector, the proposed re-formation of NDCA should not adversely affect the project. Government is discussing with IDA its detailed proposals for this re- formation, in fulfillment of its obligations under Credit 80-TA and of those proposed under the tobacco project. On-Lending 5.15 As NDCA does not have the capital needed to carry out the project, it is proposed that the IDA credit be disbursed against 100% of NDCA loans under the project. These funds would be passed on to NDCA under a Subsi- diary Loan Agreement satisfactory to IDA, conclusion of which should be a condition of effectiveness of the proposed IDA credit. The Subsidiary Loan Agreement would provide that the first T Sh 13.0 million (US$ 1.8 million) would be made available to NDCA as a grant, with the remainder, estimated at T Sh 31.9 million (US$ 4.5 million) as a loan carrying 4% interest and repay- able over 14 years, including six years' grace. These terms are appropriate considering that NDCA would need to make seasonal credit available to project cooperatives indefinitely at an annual rate of about T Sh 13.0 million (US$ 1.8 million) after full development (see para 6.11). 5.16 The Subsidiary Loan Agreement would also provide that unless IDA agreed otherwise: (a) the interest charged on all NDCA loans - seasonal, medium-term or long-term - would be 8-1/2% per an- num, which is NDCA's present rate for seasonal credit; (b) medium-term loans to cooperatives for on-lending to farmers for barn and grading shed construction would have a term not exceeding four years, with no grace period; (c) long-term loans to TTB and TTPC would have a term not exceeding 12 years including a grace period not exceeding two years; and (d) a revolving fund would be established for the provi- sion of seasonal credits to project participants.  - 22 - E. Cooperatives 5.17 The formation and operation of cooperatives in Tanzania are governed by the Cooperative Societies Act, 1968, and the subsidiary Coop- erative Societies Rules, 1968. This legislation gives wide powers to the Registrar of Cooperatives, subject usually to appeal to the Minister of Agriculture, Food and Cooperatives, including powers to refuse to register a cooperative if it does not have a reasonable chance of being financially viable; to restrict borrowing capacity; and to intervene in management whenever necessary, including direct supervision. The day-to-day operations of the cooperatives are supervised by the Cooperative Division of the Minis- try, under the guidance of the Registrar, such supervision including monthly visits to check cash and stocks and to assist in the preparation of monthly trial balances. Unless the Registrar has agreed to the appointment of inde- pendent auditors, audits are carried out once a year by officers of the Co- operative Division. However, staff shortages and inexperience severely re- strict the standards of supervision that can be achieved in practice. 5.18 There are three well-established tobacco cooperatives in the project area: the Tabora and Urambo Cooperative Societies in Tabora District and the Mpanda Cooperative Society in Mpanda District. Farm- ers in old Ministry of Land Settlement schemes have recently been reor- ganized into six new cooperatives. The Tabora Society is solvent and has a good Management Committee and a first-class Secretary-Manager seconded from BAT (Tanzania) Ltd. The Urambo and Mpanda Societies are presently in an unsatisfactory state, with poor management and accounts out of date. The Registrar has appointed a Cooperative Supervisor to look into the Urambo Society's affairs, which are complicated by its un- economic attempt to run a tractor hire service; a similar examination will be made at Mpanda. In all cases, however, current operations of channeling credit and inputs to farmers appear to be solvent. It is intended that new cooperative societies would normally be formed for each project complex, each with about 1,000 membership producing, at full development, some 1.5 million lb of tobacco - sufficient to allow an economic operation. However, some project villages or entire complexes may wish to join one of the already existing cooperative societies; in view of this, assurances were obtained during negotiations that IDA's prior approval should be obtained in such cases. 5.19 Each new society would have a full-time officer posted from the Cooperative Division for a minimum period of five years to supervise manage- ment in the formative period. During this period, the Societies would not be allowed to purchase tractors or other vehicles (which would be supplied on a hire basis by the Ministry of Agriculture). Thereafter a Society would be allowed to purchase such vehicles provided that NDCA carries out an individual appraisal of the ability of the Society to organize, maintain and operate the vehicles economically and provided that the Society has ac- cumulated sufficient equity (reserves and share subscriptions) to be able to put up at least 25% of the capital cost involved. Assurances in these respects were obtained during negotiations.  - 23 - F. Accounts and Audits 5.20 NDCA and TTB have their accounts audited annually by Cooper Brothers & Company, while TTPC engages Rajpar, Davda & Company, both re- spected local auditing firms with international connections. Cooperative Society accounts would be audited annually by the Cooperative Division of the Ministry of Agriculture, although larger units may, with the Registrar's permission, engage external auditing firms. Ministry of Agriculture ex- penditures are subject to regular Government audit. These arrangements are satisfactory, but assurances were obtained from Government during nego- tiations that they would be continued, that separate accounts would be kept for the project, and that the annual audited accounts of major project en- tities would be submitted to IDA within four months of the close of their financial year; except that, for NDCA, a period of six months would be allowed because of the complexity of its accounts and the need for these to be consistent with those of its holding company (see Annex 7). VI. PRODUCTION, MARKETING AND FINANCIAL RESULTS A. Yields and Production 6.01 In recent years, the national average yield of flue-cured tobac- co has varied between 572 lb of wet, cured leaf/acre and 640 lb/acre. Yields in excess of 1,000 lb/acre are obtained by many growers and the po- tential under efficient husbandry is much higher. It is estimated that project growers would average 550 lb/acre from their first year's crop, ris- ing to 650 lb in the second year and to 750 lb in the third and subsequent years. This prediction is considered conservative in the light of services that would be provided to growers, and the provision of adequately construc- ted grading sheds to avoid some of the losses referred to in para 3.06. 6.02 Annual production from the project is expected to reach some 22.5 million lb of cured, marketable wet leaf by 1977/78, at which time non-project production is expected to be some 29.1 million lb. The build- up of national production is estimated as follows:  - 24 - 1970/71 1971/72 1972/73 1973/74 1974/75 1975/76 1976/77 1977/78 .................... (million lb wet leaf) . Project production - 1.0 4.2 8.8 13.8 18.7 21.8 22.5 Non-project production/l 23.0 23.0 23.0 23.0 23.0 23.0 24.8 29.1 Totals: 23.0 24.0 27.2 31.8 36.8 41.7 46.6 51.6 /1 Assumes that development will continue after completion of the project settlement program (in 1974/75) at approximately project rates; and that any improvements or expansions outside the project prior to that time would be offset by reductions in production from other areas due to shortages of wood fuel for curing. With an average estimated outturn of 86-1/2%, project production of re-dried leaf is estimated to reach 19.5 million lb by 1977/78, and total Tanzanian production about 44.6 million lb. B. Markets and Prices 6.03 World production, consumption and trade in tobacco have increased steadily during the last decade and favorable prices have encouraged both area expansion and yield improvement. A major factor resulting in high prices is the 1965 embargo on Rhodesian tobacco. While Rhodesia may again become a larger exporter, its production and exports are not likely to rise to previous levels, due to loss of significant numbers of growers to other farming activities and the emergence of competitive producers such as Ugan- dan and Tanzanian smallholder tobacco growers. Further, it is unlikely that manufacturers will wish to return to arrangements which committed them to making the bulk of their purchases from a single producer. 6.04 Increasing concern, particularly in the developed countries, over the possible hazards of tobacco smoking to health may reduce the consump- tion of tobacco. The main growth of tobacco use during the last 10 years, however, has occurred primarily in developing countries and is like- ly to continue. It is possible, therefore, that the market expansion rate, which was about 2.5% per annum during the 1956-64 period, will diminish slightly in the 1970's but will still leave the opportunity for producers to expand their sales of a quality product. More information on the market for flue- cured tobacco is given in Annex 8. 6.05 Domestic flue-cured tobacco consumption in Tanzania is currently about 2.5 million lb, dry weight. While increasing, domestic consumption is not likely to exceed about 3 million lb in the near future, so that  - 25 - some 42 million lb would need to be exported by 1977/78, compared to 1969 exports of about 13 million lb. These larger exports would comprise only about 4% of world trade and no particular difficulties are anticipated in marketing leaf of Tanzanian quality. International market prices are unlikely to be significantly affected by product production which would account for less than 2% of world demand in the 1980's (see Annex 8 para. 27). 6.06 Recent average export prices for Tanzanian flue-cured tobacco have been as follows: ............ T Shillings . Crop Year 1966/67 1967/68 1968/69 1969/70 (Forecast) Average price per lb, dry weight FOB Dar-es-Salaam 3.80 4.22 4.25 4.46 Local sales to BAT (Tanzania) Ltd are currently fixed at about 10% above export prices, but this differential is expected to disappear when the auction system is introduced. Average export prices around T Sh 4.40 - 4.50 are expected to prevail over the next few years. A long-term FOB Dar es Salaam price of T Sh 4.20 per lb dry weight has been used for economic and financial projections. This price takes into account the possible end of the embargo on Rhodesian tobacco and the impact on con- sumption of the apparent hazards to health of smoking. Prices over the range T Sh 3.50 - 4.50 have been used for the purpose of assessing the sensitivity of the economic rate of return, see para 7.01. 6.07 The FOB Dar es Salaam average price of T Sh 4.20 per lb, dry weight, is equivalent to T Sh 3.63 per lb of wet leaf. From this price deductions of about T Sh 1.17 are made to cover actual costs of trans- porting, processing and selling, with a further T Sh 0.16 withheld by TTB (see para 5.09), leaving an average of some T Sh 2.30 per lb of wet leaf for the grower. A break down of the deductions is given in Annex 9. In 1970 the average price to be paid to growers will be T Sh 2.56 per lb of wet leaf. C. Financial Results Growers 6.08 A typical farm budget for a project grower is shown in Annex 10. This budget shows that by the fifth year of production, the farmer should have paid off all his medium-term loans for barn and grading shed construc- tion and be left with a net income of over T Sh 2,200 (US$308) per annum, as well as subsistence crops. This income is substantially better than anything else available in the project area and equal to, or better than, that available from most other smallholder cash crops. 6.09 While the farm income is too low to be subject to income tax, farmers would contribute to Government revenues through indirect sales  - 26 - taxes, which average about 10% on all non-essential goods, and through amounts withheld by TTB (see para 5.09). Assuming that 30% of the farm net income above T Sh 500 (US$70) per year would be subject to sales tax, and that TTB's withholding rate remains at T Sh 0.16 per ib, farmers in the project would each contribute nearly T Sh 300 (US$42) per annum to Government revenues at full development or a total of about T Sh 4.5 mil- lion (US$630,000) per annum. Cooperatives 6.10 Until such time as project cooperatives operate their own produce transport services and discontinue the use of Ministry of Agriculture facil- ities (see para 5.19), their financial obligations would remain seasonal, except for those incurred in acting as a channel for medium-term loans to their members. Initially, project cooperatives would follow current prac- tices for charging for their services, namely: (a) a 10% mark-up on all inputs to farmers to cover the cost of distribution from the railhead to village; (b) an additional 5% mark-up on all seasonal inputs to cover the interest charged by NDCA; 1/ (c) 8-1/2% interest per annum on medium-term loans to farmers to cover the interest charged by NDCA; (d) a cess of T Sh 0.10/lb of cured leaf to cover the costs of baling, handling and packing materials; and (e) a cess of T Sh 0.15/lb of cured leaf to cover the cooperative's overhead and generate a reserve. The above charges should allow a moderately well managed cooperative to meet all costs and to build up modest reserves. Since overhead costs would be met and reserves generated by a cess on production, however, the efficient producer is to some extent disfavored by this system and there is some farm antagonism to the cess. Consideration thus would be given during implementation of the project to increasing the mark-ups on inputs and interest rates and to removing or reducing the T Sh 0.15 cess. National Development Credit Agency 6.11 The effect of the project on the cash flow position of NDCA is shown in Annex 11, Part I. On the basis of the on-lending pattern described 1./ The NDCA charges 8-1/2% per annum, but the average period that season- al loans are outstanding is about seven months and actual interest paid usually averages slightly below 5%.  - 27 - in para 5.15, the overall NDCA cash flow is improved at all stages of the project, and NDCA should be able to meet the additional credit needs of the project farmers beyond the disbursement period of the IDA credit without re- course to further borrowing. Tanganyika Tobacco Board 6.12 The impact of the project on the cash flow position of TTB is shown in Annex 11, Part II. Assuming that TTB continues to withhold T Sh 0.16/lb (see para 5.09) and that credit for construction would be provided by NDCA, TTB would run deficits in 1971/72 through 1973/74 as a result of project financed activity, but would accrue substantial surpluses in subse- quent years. These deficits could be met from withholdings from tobacco marketed from pre-project schemes. Government 6.13 The direct effect of the project on Government revenues is shown in Annex 11, Part III. Funds would accrue to Government directly from draw-down of the IDA credit; payments by farmers, through the cooperatives, for the use of Government transportation facilities; additional sales tax revenues; and repayment by NDCA of Government loans. These direct receipts would not cover direct Government project cash costs, however, and there would be a net outflow of Government revenues, of about T Sh 26.4 million (US$ 3.7 million) through 1975/76. On the other hand, Government would be in receipt of indirect revenues in the form of the surpluses accruing to NDCA and TTB. Taking these into account, there would be a cumulative nega- tive cash flow through 1973/74 of some T Sh 21.8 million (US$ 3.1 million), which would have been recovered by 1978/79. This overall cash flow is shown in Annex 11, Part IV and V. VII. BENEFITS AND JUSTIFICATION 7.01 The project's primary benefits would be increased production of flue-cured tobacco, all of which would be exported, and consequently pro- vision of an improved livelihood for some 15,000 farm families. Export revenues at full development in 1977/78 and subsequent years would reach some T Sh 82 million (US$ 11.5 million). Assuming an opportunity cost for family labor of T Sh 2.50/man-day, a negligible opportunity cost of land, and an economic life for each settlement and its tobacco farms of 15 years, the economic rate of return from investments in the project is estimated to be about 30%. The calculation is shown in Annex 12, which also indicates the sensitivity of the rate of return to various factors. The rate of re- turn varies from about 12% to about 37% over a range of export prices from T Sh 3.50 - 4.50; and from about 19% to about 54% over a range of costs from 20% higher than estimates (excluding contingencies) to about 10% lower.  - 28 - 7.02 The rate of return is particularly sensitive to labor costs. A strong case can be made that the opportunity cost of family labor is cur- rently negligible; if it is not costed the rate of return would rise to about 55%. Costing family labor at T Sh 5.00/man-day - a value near the Government official minimum wage rate - reduces the rate of return to some 2%. The value of T Sh 2.50/man-day placed on family labor for the purpose of the principal calculation of the rate of return, is the price now paid in the project area for casual labor. 7.03 In addition to the direct benefits taken into account in the above calculations, the project would have the secondary benefit of creat- ing communities where basic social services can be supplied more efficient- ly than to isolated farm holdings. It also has the socio-political bene- fits of providing an economic utilization of large areas of hitherto unproductive miombo woodland; of relieving heavy population pressure in other parts of Tanzania, particularly the Tukuyu and Kilimanjaro areas; and of providing alternative employment for displaced sisal estate workers. While not quantifiable, these benefits should be regarded as important and serve to underline the high priority of this project for a country in which the opportunities for agricultural development are limited. VIII. RECOMMENDATIONS 8.01 During credit negotiations, agreement was reached on the following principal points: (a) control of the production of project and non- project flue-cured tobacco (paras 4.04, 4.21); (b) arrangements for the conduct of the marketing survey and implementation of its recommenda- tions (para 4.17); (c) the appointment of various officers to the "Project Group", the extension services, Tum- bi Research Station, TTB, TTPC and NDCA (paras 5.03, 5.04, 5.05, 5.08, 5.11 and 5.13); (d) interim measures related to the introduction of the auction system of marketing (para 5.08); and (e) staffing, management and operations of new coop- eratives and involvement of existing cooperatives in the project (paras 5.18 and 5.19). 8.02 Conditions of effectiveness are that: (a) the Tobacco Project Officer and the Senior Plan- ning Officer have been appointed to the "Project  - 29 - Group" (para 5.03); TTB has employed a Chief Mar- keting Officer (para 5.08) and NDCA has employed a Financial Specialist (para 5.13); and (b) a Subsidiary Loan Agreement, satisfactory to IDA, has also become effective (para 5.15). 8.03 No disbursements would be permitted for construction of auction floors or storage facilities, or for the extension of existing processing capacity beyond that scheduled for completion by March 1971, pending the results of the marketing study and agreement between Government and IDA on design, cost estimates, organizational aspects, procurement and disburse- ment (para 4.29(d)). 8.04 The proposed project constitutes a suitable basis for an IDA cre- dit of US$ 9.0 million.   ANNEX 1 TANZA4IA FLUE-CURED TOBACCO PROJECT TYPES OF TOBACCO AND PRODUCTION METHODS _'Tpes_of Tobacco and Curinp Methods 1. Two types 1/ of tobacco are currently grown in Tanzania: flue- cure(d used for Virginia tobacco, and fire-cured for dark tobacco. The flue-cured Virginia tobacco is used almost entirely in the manufacture of cigarettes, while fire-cured is used in pipe srmoking mixtures, chewing to- bacco and to a minor extent in cigarettes (dark cigarettes and shag tobac- co for individually rolled cigarettes). 2. In fire curing, the green leaves are strung in tiers in a barn over a small smokey fire; the cured leaf becomes dark brown in color. Flue curing takes place in a barn of a more permanent structure heated by a system of flue pipes with an outside funace. Strict control of temper- ature is necessary and after curing the tobacco has a bright yellow color. Project Farming Methods for Flue-Cured Tobacco 3. Under the "Ujamaa" system of village settlement, each year a site for the village tobacco nursery and tobacco growing block are se- lected by the local cooperative on advice from the agricultural exten- sion officer. Within these areas each farmer is allocated his own seed bed and tobacco area of two acres (one acre in the first year of farm- ing). The cropping pattern is based on a rotation: Year 1 tobacco; Year 2 maize; Years 3, 4 and 5 grass fallow. 4. Cultivation is similar to that practised in Rhodesia, Zambia and Malawi; seed beds are fumigated, treated with fertilizer and in- secticides and sown early in September followed by transplanting at the beginning of the rain season in November. Compound fertilizers are ap- plied and insecticides (Aldrin or ])DT) used when necessary; eelworm is controlled by the use of nematoci,de. The principal tobacco varieties are K51 and white gold and seed is bulked in the Tumbi Research Station. 1/ Aromatic tobacco (somietimes known as "Turkisli" or oriental) was grown by smallholders until 1968: as a result of problems of marketing and quality it is no longer grown.  ANNEX 1 Page 2 5. Harvesting commences at the end of January and continues until March. Adequate barn accommodation is essential and three barns are re- quired for each farm of two acres. (It is estimated that under the pre- sent system of smallholder production, there is a current loss of 407 during harvesting and curing, due to lack of barn accommodation and damln and leaking barns and grading sheds). After curing the tobacco is graded and baled ready for sale. Bales have ranged in size from under 100 lb to nearly 200 lb but with the introduction of an auction sale system, re- placing the present direct sale to the Tanganyika Tobacco Board, they will be standardized at abotut 180 lb. 6. Maize would be planted in November with an anplication of sul- olhate of ammonia to supplement the residual fertilizer from the previous tobacco crop. Labor Requirements 7. Some farmers with small or young families may find it necessary to hire labor in March and November for transplanting and harvesting re- spectively; under normal circumstances, however, the farmer with his fam- ily should be able to look after twTo acres of tobacco and two acres of maize.  TANZANIA FLUE CURED TOBACCO PROJECT ORGANIZATION CHART