Brazil - Bahia Rural Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 2009a-BE STAFF APPRAISAL REPORT NORTHEAST BRAZIL BAHIA RURAL DEVELOPMENT PROJECT - PARAGUACU May 19, 1978 Projects Department Latin America and Caribbean Regional Office 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 EOUIVALENTS Currency Unit Brazilian Cruzeiro (Cr$) US$1 Cr$17.27(Selling Rate,jIay 1978)* Crr$1= US$0.0579 Cr$1 million US$57,904 WEIGHTS AND MEASURES 1 meter (m) = 3.28 feet 1 kilometer (km) 0.62 mills 1 hectare (ha) 2 10,000 m = 2.47 acres 1 square kilometsr (km ) 100 ha = 247.1 acres = 0.386 sq.miles 1 cubic meter (m ) 1.31 cubic yards = 264.2 US gallons 1 kilogram (Kg) 2.2 pounds 1 ton 1,000 Kg = 2,205 pounds GLOSSARY OF ABBREVIATIONS (see next page) GOVERNENT OF BRAZIL FISCAL YEAR - January 1 to December 31 POLONORDESTE PROGRLM FISCAL YEAR - April 1 to March 31 * Average Exchange Rates (Selling): Oct. 1975 1976 1977 US$ = Cr$ 8.128 10.786 15.000  FOR OFFICIAL USE ONLY GAiJSSARY OF ABBREVIATIONS BB = Banco do Brasil (Bank of Brazil) BNB = Banco do Nordeste do Brasil (Bank of the Northeast of Brazil) BNCC - Banco Nacional de Credito Cooperativo (National Cooperative Credit Bank) CAMAB = Companhia de Adubos e Materiais Agricolas da Bahia (Bahia Fertilizers and Agricultural Materials Company) CASEB = Companhia de Armazens Gerais e Silos do Estado da Bahia (Bahia State General Warehouses and Silos Company) CEPA-BA = Comissao Estadual de Planejamento Agricola-Bahia (Bahia State Agricultural Planning Commission) CERB = Companhia de Engenharia Rural da Bahia (Rural Engineering Company of Bahia) CFP = Comissao de Financiamento da Produ,cao (Production Financing Commission) COBAL Companhia Brasileira de Alimentos (Brazilian Food Company) CRIBA = Consorcio Rodoviario Intermunicipal da Bahia S.A. (Inter- municipal Road Consortium of Bahia) DCOOP = Divisao de Cooperativismo (Division of Cooperatives of Bahia Secretariat of Agriculture) DERBA = Departamento de Estradas de Rodagem da Bahia (Department of Roads of Bahia) EMATER-BA Empresa de Assistencia Tecnica e Extensao Rural - Bahia (Bahia Technical Assistance and Rural Extension Company) EMBRAPA = Empresa Brasileira de Pesquisa Agropecuaria (Brazilian Agri- cultural Research Company) EPABA = Empresa de Pesquisa Agropecuaria da Bahia (Bahian Agricultural Research Company) FUNRURAL Fundo de Assistencia ao Trabalhador Rural (Assistance Fund for Rural Workers) FUSEB = Funda9ao de Saude do Estado da Bahia (Bahia State Health Foundation) IBGE = Instituto Brasileiro de Geografla e Estatistica (Brazilian Institute of Geography and Statistics) INCRA = Instituto Nagional de Coloniza;Zo e Reforma Agraria (National Institute of Colonization and Agrarian Reform) INTERBA = Instituto de Terras da Bahia (Bahian Land Institute) PIASS = Programa de Interioriza,ao das Acoes de Saude e Saneamento (Program for the Interiorization of Health and Sanitation Activities) PIDERP = Projeto Integrado de Desenvolvimento da Bacia do Rio Paraguagu (Integrated Rural Development Project for the Paraguacu River Basin) POLONORDESTE = Programa de Desenvolvimento de Areas Integradas do Nordeste (Development Program for Integrated Areas in the Northeast) PROAGRO = Programa de Garantia da Atividade Agropecuaria (Guaranty Program for the Crop-Livestock Sector) SEC = Secretaria da Educagao e Cultura do Estado da Bahia (Education and Culture Secretariat of Bahia) SENAC = Servi9o Nagional de Aprendizado Comercial (National Commercial Apprenticeship Service) SENAI = Servigo Nacional de Aprendizado Industrial (National Industrial Apprenticeship Service) SENAR = Servigo Nacional de Forma,cao Profissional Rural (National Rural Training Service) SESAB = Secretaria de Saude da Bahia (Secretariat of Health of Bahia) SUCAM = Superintendencia de Campanhas Medicinais (Superintendency of Public Health Campaigns) SUDENE = Superintendencia de Desenvolvimento do Nordeste (Northeast Development Superintendency) 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.   NORTHEAST BRAZIL BAHIA RURAL DEVELOPMENT PROJECT - PARAGUACU STAFF APPRAISAL REPORT Table of Contents Page No. I. THE AGRICULTURAL SECTOR ..... .......................1 The Sector in the Economy ..... .....................1 Agriculture in Northeast Brazil .................... 3 Government Rural Development Strategy in the Northeast ................. 4 II. THE STATE OF BAHIA AND THE PROJECT AREA ............ 5 The State .......................................... 5 The Project Area ........ ........................... 6 Location and Physical Features ..... .............. 6 Socioeconomic Conditions ......................... 7 Agriculture ........ .............................. 9 Physical Infrastructure .......................... 10 Social Services .................................. 10 Agricultural Services ............................ 10 III. THE EXECUTING AGENCIES ............................. 11 State Framework for Executing the POLONORDESTE Program ........ .................................. 11 Institutions Engaged in Rural Development in the Project Area .................... ................. 12 IV. THE PROJECT ........ ................................ 12 Origins, Objectives and Approach ..... .............. 12 Brief Description .................................. 15 Detailed Features .................................. 16 Extension ......... ............................... 16 Agricultural Research and Seed Production ........ 17 Credit ........... ................................ 18 Land Purchase Credit ...... .......................21 The report is based on the findings of an appraisal mission in October 1977 composed of Messrs. E. Senner, G. Ashkenazi, M. Baxter, N. Hughes, 0. Lafourcade, P. Morris and A. Tobelem (Bank) and Messrs. E. Calderon and J. Mora and Ms. J. Tendler (consultants) and Mr. F. Mortier (FAO/CP).  Table of Contents (Cont'd) Page No. Land Surverying and Titling Services .......... 22 Cooperative Society Support . . .................... 23 Mechanization/Input Supply .................... ... 24 Irrigation .......... 25 Multi-purpose Dams . . ................. 26 Crop Storage ..................................... 27 Feeder Roads .... ................................. 28 Health Services ........ .......................... 29 Village Water Supply ...... ....................... 30 Education and Vocational Training ................ 32 Project Administration, Monitoring and Evaluation 33 V. PROJECT COSTS AND FINANCING .... .................... 34 Cost Estimates ...................................... 34 Financing .......................................... 36 Procurement ........................................... 36 Disbursements ....... ............................... 37 Accounts and Auditing .... ........................ . 37 VI. PROJECT IMPLEMENTATION .............. .. ............. 38 Organization and Management ............ ............ 38 Operating Agreements and Procedures . ............... 41 Implementation Schedule .............. .............. 42 Evaluation ...................... ................... 42 VII. PROJECT PRODUCTION, DEMAND AND MARKETING .... ....... 43 Production ......................................... 43 Demand, Markets and Prices ............ .. ........... 44 Marketing .......................................... 46 VIII. FINANCIAL ANALYSIS ...... ........................... 48 On-Farm Operations ...... ........................... 48 Producer Income .................................... 49 IX. ECONOMIC BENEFITS AND EMPLOYMENT . . . .51 Economic Rate of Return... 51 Employment ....54 Project Risks . ...54 Fiscal Impact ....54 Environmental Impact ....55 X. SUMMARY OF AGREEMENTS REACHED AND RECOMMENDATIONS .. 55  Table of Contents (Cont'd) ANNEXES: 1. Design Standards for Feeder Roads 2. List of Collector Roads Planned Parallel to Project 3. Annual Phasing of Project Costs 4. Estimated Schedule of Bank Disbursements 5. List of Executing Agencies 6. Organization Chart of Project (PIDERP) Management Unit 7. Indicators for Monitoring and Evaluation 8. Implementation Schedule 9. Technical Coefficients and Specifications and Illustrative Farm Models 10. Cost and Benefit Streams for Economic Analysis 11. Related Documents Available in Project File MAPS: POLONORDESTE Priority Areas in State of Bahia (IBRD-13284) Project Area (IBRD-13285) Climatic Zones and Rainfall (IBRD-13286)   NORTHEAST BRAZIL BAHIA RURAL DEVELOPMENT PROJECT - PARAGUACU STAFF APPRAISAL REPORT I. THE AGRICULTURAL SECTOR The Sector in the Economy 1.01 Although agricultural production in Brazil has been growing at a lower rate than the GDP as a whole and now accounts for only 11% of the GDP compared to 20% in 1967, agriculture continues to be an important sector in the economy. The national agricultural workforce numbers 12.8 million or 39.5% of the economically active population, and 43.1 million hectares (out of a national territory of 8.5 million km2) are in agriculture. 1.02 About 56% of merchandise exports originate in the agricultural sector. Coffee, soybeans and sugar alone generated 42% of the country's export earnings in 1976. Brazil is a major world supplier of coffee, soybeans, sugar, cotton, peanuts, cocoa, orange juice, castor oil and sisal. It is still the largest exporter of coffee in the world, and is now the largest producer of cane sugar in the world. The production of soybeans has increased from less than one million tons in 1968 to over 11 million tons in 1976 and an estimated 12 million tons in 1977, making Brazil the third largest producer in the world. This expansion has been instrumental in restructuring agriculture in southern Brazil. 1.03 The most important crops for the domestic market include manioc (cassava), maize, rice, wheat, beans, potatoes, vegetables, bananas, citrus fruits, dairy products, poultry, beef and swine; coffee, sugar, soybeans, cotton, and cocoa are important for both the export and domestic markets. With the exception of wheat, Brazil is largely self-sufficient in basic food production, although some staples (such as garlic, potatoes and even beans) are imported seasonally. Half of the wheat consumed in Brazil is still imported. 1.04 Over 1967-74, the agricultural sector performed well in supplying domestic demand for food and fiber as well as a rising volume of export commodities, with output growing at 5.7% annually despite adverse weather conditions in three of the seven years. Farm output increased 4.2% over 1975-76, an increase due more to productivity gains with livestock (12.2%) rather than crops (0.4%, though 9% if coffee is excluded). Productivity of soybeans, corn, rice, wheat and sugar increased, while productivity of cotton, beans, castor beans, cocoa, sisal and tobacco declined. Recently, the value of output has been increasing steadily but it is due more to an expansion of production area rather than improved productivity. However, despite increased attention to productivity (including advocacy of mechanized production) and increasing market integration, subsistence and semi-commercial production and rudimentary technology continue to be widespread, particularly in the Northeast, North and Center West.  -2- 1.05 The agricultural development strategy of Brazil's Second National Development Plan (1975-79) aims at more efficient exploitation of the country's agricultural potential by increasing agricultural output by 40% in real terms during the 1975-79 period. This goal reflects the Government's intention to provide an increasing share of Brazil's growing domestic requirements for food and fiber, as well as to increase agricultural exports. In addition, the Government has acknowledged the gravity of the country's income distribu- tion problems. Major measures to increase agricultural output include: (i) utilization of substantial public resources to incorporate new areas into production; (ii) significant expansion in the supply of agricultural credit; (iii) reorganization of federal research and extension services; and (iv) special regional programs, such as POLONORDESTE, that focus on poorer regions and smaller scale farmers. The POLONORDESTE program represents an important shift in Brazilian agricultural policy, to give emphasis to the development of programs to raise the productivity, incomes and quality of life of the rural poor. 1.06 One important aspect of the Government's agricultural policy is its practice of encouraging a large volume of agricultural credit at subsidized interest rates. For most of Brazil these rates now vary between 13% and 21% per annum without monetary correction, which in Brazil's inflationary environ- ment means that they are highly negative in real terms. In the POLONORDESTE program, the standard rate for small loans is now 10% without monetary correction. Credit for fertilizer purchase is interest free. The subsidized interest rate system was originally introduced in Brazil some years ago as a means of compensating the agricultural sector for adverse terms of trade resulting from high prices for fertilizer, machinery and other inputs. The subsidized rates were also intended to provide an incentive to expand agricultural production and, in the case of the Northeast, to give special help to a less developed part of the country. The main institutions extending agricultural credit, the Banco do Brasil and the Banco do Nordeste do Brasil, are able to extend this credit on subsidized terms by virtue of large interest-free deposits which public agencies place with them, so that the financial burden of the subsidies is borne by the public sector as a whole, rather than by individual financial institutions. However, the system has severe disadvantages, including possible leakages from the agricultural sector into other, lower priority sectors, the inequity of granting large farmers access to subsidized credit, and, in the case of larger agricultural enterprises, the encouragement of unduly capital intensive techniques. The possible adverse effects of such subsidized credit are much less in the case of credit to very small farmers, such as those being assisted under the POLONORDESTE program, because many of the beneficiaries are below the poverty level and those programs are being intensively supervised in a manner to ensure that the credit is used for the intended purposes. Bank staff have for many years conducted a dialogue with the Government on this matter, and the Government is aware of the serious disadvantages of the system. In the past year a few modifications have been made. Larger farmers are now required to finance at least 25% of on-farm investments with their own resources, and a sliding scale of rates has been established charging larger farmers somewhat more than smaller farmers. The basic POLONORDESTE rate in the Northeast has been increased from 7% to 10%. The Government has felt that it was politically impossible to move further at this time, and, therefore, most of the undesirable features of the program still remain.  1.07 The Bank has so far made 13 loans, totalling US$480.4 million, for agriculture and rural development in Brazil. These include two, amounting to US$60.5 million, for livestock development; two, amounting to US$137 mil- lion, for agro-industries; one, of US$18.2 million, for grain storage; one, of US$40 million, for an agricultural research program for the Northeast, North and Center West of Brazil; one, of US$100 million, for the ._velopment of agricultural extension services; and six, totalling US$124.7 million, for various settlement, irrigation and rural development projects, five of which are located in the Northeast. Bank loans for a nutrition project and a secondary and feeder road project are also providing important benefits to the rural population. In addition to other agricultural and rural development projects, a rural education program is also under consideration for possible future Bank financing. Agriculture in Northeast Brazil 1/ 1.08 The nine states of Northeast Brazil have a population of about 30 mil'lion, two-thirds of whom live in rural areas. Much of the region has a semi-arid climate with periodic serious droughts 2/ and generally poor soils. Nonetheless, the Northeast is a significant agricultural region, usually accounting for 20% of Brazil's agricultural production and a major share of the national production of cocoa, cotton, manioc, beans and sugar. In 1977, 26.9% of the area harvested in Brazil was in the Northeast. Throughout the Northeast, agricultural productivity is low and rural per capita incomes are estimated to be only one-quarter the national average (about US$1,140 in 1976). In much of the Northeast, production of both food and industrial crops is small scale and with simple technology. There is little mechanization of production and very limited use of modern inputs. Annual crops are grown largely on a shifting basis. While technology is unsophisticated and yields poor, farming systems are adjusted to the environment: lack of credit for working capital or investment, limited agricultural extension services, market isolation and weak linkages with the regional economy, insecurity of land tenure and, in areas, a difficult physical environment. 1.09 The environmental problems faced in reducing poverty and improving productivity are compounded by a highly skewed distribution of land owner- ship. Less than 1% of the total number of farm establishments account for 40% of the area, and in some areas up to 80% of the rural labor force owns no land at all, though this proportion varies considerably by region. Among Northeastern states, the proportion of landholders who have under 10 hectares ranges from 49% in Ceara to 87.6% in Maranhao. 1/ Readers are directed to "Rural Development Issues and Options in Northeast Brazil," (Report No. 665a-BR dated June 23, 1975). 2/ The 1976 drought, for example, was primarily responsible for a 7% decline in Northeast agricultural production. Especially hit were maize, beans, cotton, castor beans and sisal.  -4- Go vrnment Rural Development Strategy in the Northeast 1.10 Numerous Government programs have been directed at speeding economic deeoent in the Northeast. The Government operates a special development agency (SUDENE) and a special development bank for the Northeast (Bank of the Northeast: of Brazil, BNB), and substantial fiscal and monetary incentives have been made available over the past decade for both industry and agricul- ture. Per capita income in the region is still, however, only about half the national average and rural per capita income is only one-quarter the national average. Programs to generate jobs in the industrial sector have not kept pace with population increases, and the gradual emigration from rural to urban areas has been coupled with considerable continued emigration to the more industrialized Center South of Brazil. Past Government programs directed to the rural population have often been in the areas of public works, particularly substantial efforts in drought-prone areas to develop dams and ponds. in comparison, the current Government has moved toward the reorientation and strengthening of the institutions offering the services required to develop 'r7e productive labor-intensive small scale farming. 1.11 The POLONORD2STE program, established in 1974, is a major new rural development strategy. it is helping to finance: (i) preparation and initial implementation expendcL.-ores for new integrated rural development projects in subregions selected for their development potential, and (ii) the continuation or completion of projects (particularly land settlement and irrigation) begun or planned before the creation of POLONORDESTE. By the end of 1977 around US$584 million had been allocated to these various efforts. The establishment of programs such as POLONORDESTE has helped mobilize the energies of various Government bodies toward impro;iing agricultural productivity and the standard of living of the rural poor. Thi& has been especially apparent in the areas of rural extension, agricultural research, education and health. Nevertheless, at all levels of government, the institutional capacity to plan, coordinate and execute the program still requires strengthening. 1.12 At the state level, in some cases, initial failure to carefully define target groups, to analyse the relationship between the needs of these groups and proposed investment alternatives, and to think through the long-run administrative and financial implications of project execution has resulted in very uneven performance, within and among projects. At the regional and federal levels, there have been frequent delays in funding, reflecting complicated internal budget, disbursement and reporting systems involving numerous (not always well-prepared or staffed) agencies. Progress has recently been made in facilitating the flow of financing through: the Government establishment of a revolving fund for advancing some of the financing for the rural development projects in the Northeast which are subject to World Bank loans; the earlier annual definition of budget fund availability for the entire program, thereby facilitating local planning; and the establishment of a tighter schedule of deadlines for review and approval of annual project operating plans. However, the SUDENE group responsible for assisting states in the preparation of proj- ects and for reviewing quarterly and annual reports and plans for the numerous POLONORDESTE projects is still very weak. Though the Bank is discussing with the Government new means of collaborating (e.g., through special technical  - 5 - assistance or training efforts) in reducing the administrative constraints in the program overall, progress in that area has been slow. Fortunately, particularly in several states, the POLONORDESTE program has had a dramatic impact in stimulating the development of local groups of highly motivated and dedicated techn4cians working on rural development project planning and execution, a factor signalling a likely continued importance of the program. II. THE STATE OF BAHIA AND THE PROJECT AREA The State 2.01 Bahia is both the most southern and largest (560,000 km2, 35% of the Northeast) of the nine Northeast states. As elsewhere in the Northeast, three distinct ecological zones are distinguishable in Bahia, although in the more humid south and the central mountain range (the Chapada Diamantina), the traditional zone classification is not so applicable. A humid coastal region, extending 50-100 km inland, once covered with dense tropical forest (hence the name, zona da mata) but now largely cleared for settlement, has an annual average precipitation of over 1,000 mm. The interior sertao is an arid and semi-arid zone with xerophytic and semixerophytic vegetation (caatinga and cerrado) and annual precipitation under 600 mm. This area contains the 56% of the state located in the "Drought Polygon"of the Northeast. An intermediate temperate zone (agreste) has a transitional climate and vegetation. 2.02 Coastal Bahia was one of the first parts of Brazil to be settled by European immigrants. The area, like the rest of the state, has experienced cyclical phases of economic activity: sugar, cocoa, precious and semi-precious stones, cattle and light industry have all had periods of particular emphasis in the economy. Agriculture for long was the key sector. Even today approxi- mately 59% of the state's population (which totals over 7.5 million) lives in the rural areas. Over the past decade, Bahian agricultural production has accounted for 28% of Northeast and 8% of national production. The Bahian cattle herd is 40% of the Northeast herd. Cocoa, castor beans, sugar, sisal and tobacco are the chief export crops; maize, beans and manioc, the chief products for the internal market. While the area under cultivation continues to increase (24% during 1960-70), however, the relative importance of the agricultural sector has declined. From 43% of the Bahian gross product during the 1950s, the agricultural sector now contributes about 30%, about half of which comes from cattle. Agricultural productivity is improving in some areas, but overall agriculture in Bahia is characterized by low levels of technical skill, little development of infrastructure or service support (including marketing and storage support and facilities), very little local product processing, resource distribution inequalities, and a long history of rural emigration. 2.03 The current state government established in its five-year plan a target annual growth rate of 10% over 1975-79, including 12% in industry and 7% in agriculture (mainly livestock). Means to this end include:  -6- (i) integrated development of the state's chief river basins (including the predominant Sao Francisco as well as Paraguacu) and expansion of the area cultivated; (ii) modernization and commercial integration of agriculture; (iii) exploitation of mineral resources, particularly copper; (iv) develop- ment of agro-industry and the processing of export agricultural products; (v) boosting of urban economies through support to industry, including small-scale industries in the interior and tourism; and (vi) execution of localized intensive development efforts, including through the POLONORDESTE program. The emphasis on industry has seen a shift in attention from the numerous small traditional industries (food, clothing) I/ supported in the mid-1960s to basic industries such as petrochemicals, textiles and electro- mechanical engineering, all primarily in the greater Salvador and Feira de Santana areas. The new direction of development is suggested by the 36 projects approved in Bahia by SUDENE for Government support in 1976: 28 were in industry, one in telecommunications and seven in agriculture. An indication of the results of the strategy is the 56% increase in receipts of industrial production tax (IPI) over 1975-76. Agriculture has taken a lesser role in development but still is important in overall growth, the chief new directions being an emphasis on particular areas (as in the POLONORDESTE program), an improved physical and service infrastructure to serve the sector, and newly- introduced crops like coffee and irrigated wheat. The Project Area 2.04 Location and Physical Features. The 60,764 km2 project area (which is larger than the entire state of Paraiba, for part of which Bank financing of another rural development project was recently approved) comprises the greater part of the Paraguacu River basin (Map No. 13284). The state capital, Salvador, is on the coast some 120 km from the project area's eastern limits. Centered at approximately 400 W and 120 S and comprising 49 municipalities, the project area extends some 225 km north-south and 375 km east-west between the humid litoral zone to the east and the semi-arid interior sertao to the west. The high and relatively humid Chapada Diamantina in the west of the project area comprises some one-quarter of the area; much of the remainder is a low, relatively flat, semi-arid crystalline basin. 2.05 Aside from the Chapada Diamantina and the eastern periphery of the region, the project area is semi-arid and part of the Northeast "Drought Polygon." Mean monthly temperatures range from 190 to 260. Rainfall has marked temporal and spatial variations: it is concentrated in the summer (November to March) and ranges from over 1,400 mm to below 500 mm annually (Map IBRD-13286). Areas with less than 800 mm precipitation (about half the project area) have irregular rainfall; some central regions are currently in the third consecutive year of markedly below-normal rainfall. The Paraguacu 1/ In 1970, 76.3% of industrial establishments in the state contributed only 4% of the value of industrial production.  - 7 - River, its larger tributaries, and a number of other perennial streams help relieve local precipitation problems. Ground water resources are limited and difficult to tap economically because of the crystalline substrata encountered in two-thirds of the project area. The altitude of the project area rises gradually from about 200 meters in the east to over 1,000 meters in part of the Chapada Diamantina. Aside from the escarpments of the Chapada, relief is generally undulating and in the east is flat or rolling, with limited hilly zones. Soils are mainly latosols, podsols and planosols of medium-to-low natural fertility. 2.06 On the basis of the physical environment and land use and distri- bution, five discrete sub-areas within the Paraguacu Basin Project area were delineated for purposes of project planning and implementation (see Map IBRD-13285). The sub-areas, and some of their key characteristics, are as follows: Population (1970) Area Munici- palities Location within Sub-area '000 % rural km2 No. Project Area I Planalto 101.8 82.7 2,865 4 Southeast II Cristalino Sul 109.5 78.4 12,279 8 Center-South III Chapada Diamantina 168.3 80.0 19,754 15 West IV Cristalino Centro 300.1 79.1 21,757 13 Center-North V Feira de Santana 303.9 47.9 4,109 9 Northeast Project Area 983.6 69.9 60,764 49 2.07 Socioeconomic Conditions. Precise data on income are unavailable (although most of the rural population is below Brazil's relative poverty level of about $380 per capita), but an impression of the socioeconomic condition of the population is available from other sources. As would be expected with a region the size of the project area, regional differentiation is apparent. The eastern area is densely settled, well integrated with markets and is heavily influenced by the urban-industrial center of Feira de Santana. To the west, settlement is more sparse, economic opportunities fewer, and physical and service infrastructure more scarce. Particularly away from the central valley road artery, areas in the west are poorly linked to the commercial system and have few services. 2.08 The project area currently has a population of slightly more than one million (984,000 in 1970, the most recent census). Average settlement density is 16.2 persons per km2, ranging from 74.0 in the east to 8.5 in the west.  - 8 - Thirty percent of the population is "urban," living in the area's 100 desig- 'na-e. urban centers. In fact, only 20% of the population lives in the eight centers with more than 5,000 inhabitants. Feira de Santana with 127,000 *.nhabIants in 1970 (13% of the project area population) dominates the area's urban structure and will become even more dominant: by 1980, it is estimated that 20% of project area population will live there. The project area has long been a source of emigrants to other parts of Brazil (to the Center West and later to the Center South and South). However, while rural population has been increasing slower than the population as a whole, this seems due more to movement to urban centers within the valley than migration to external areas. Over 1960-70, urban population increased 57%, and rural population 10%. The Chapada Diamantina had only a 4% increase over the decade. Population is youthful, with 45.7% under 15 years of age and 23.5% over 34. 2.09 Agriculture is the chief economic activity of the region. Agricul- ture accounts for 71.8% of the economically active population (50% of the population over 9 years of age), industry for 8.3% and services for 18.1%. (Respective figures for Bahia are 62.5%, 11.1% and 23.7%.) Within the project area, there is considerable variation in the composition of employment, the Feira de Santana area having markedly higher shares of the workforce in industry and services. Distribution of agricultural land by farm size is similar to elsewhere ir the Northeast. Eighty percent of owner-operated farms are under 50 ha, but these together account for only 16% of the area farmed. Data on non-owner operated farms are incomplete but it appears that these are worked more by squatters than by sharecroppers and are most common in the southern and western parts of the project area. The following table summarizes the size distribution of all farms in the project area: Sub-area No. of Farms /1 % Farms by Farm Size /2 % Farm Area by Farm Size /2 Owner Non-Owner <10 11-50 >50 ha <10 11-50 >50 ha I 5,870 2,570 58.3 29.3 12.4 6.4 17.4 76.2 II 4,770 1,760 26.6 25.8 47.6 0.5 5.4 94.1 III 14,510 3,860 56.2 31.2 12.6 5.8 17.9 76.3 IV 19,900 1,640 25.1 44.3 30.6 1.5 11.7 86.8 v 15,040 1_870 59.4 31.4 9.2 9.9 27.7 62.4 Project Area 60,090 11,700 43.8 35.8 20.4 3.0 13.1 83.9 /1 Few "owners" in the project area actually have complete land title docu- mentation. Figures for noniowners are very approximate and include renters, sharecroppers and squatters, but not landless rural laborers. Source: CEPA-BA calculations from IBGE (1970) and INCRA (1972) data. /2 Data are for owner-operated farms only: comparable data for non-owner operations are unavailable. Source: CEPA-BA calculations from IBGE (1970) and INCRA (1972) data.  - 9 - 2.10 Agriculture. Related to either its share of the area (11%) or the population (13%) of Bahia, the project area generally accounts for a dis- proportionately high share of state production of industrial crops and a near proportionate share of food crops and cattle. Much of food crop production is directed to on-farm consumption or urban markets within the project area. the Paraguacu valley has not played a significant role in producing food for external markets. There are about four million hectares of farm land in the project area, 1.1% of which is in permanent crops, 4.5% in temporary crops, 51.3% in pasture (including 23.1% in improved pasture), 21.2% in forest/scrub, and 16.5% fallow. A very small area (0.1%) is irrigated. In reflection of the diverse ecotypes of the area, there is considerable spatial variation in production, but basic crops and crop combinations throug- hout the valley are similar. The chief food crops are manioc, mulatinho and macassar beans, and maize; the main industrial crops are castor beans, tobacco and sisal. Small quantities of rice, tomatoes, onions, garlic and a variety of other fruits and vegetables are also grown. Beans and maize are frequently interplanted, as are castor beans with beans or manioc. In the drier northeast, sisal is an important crop. Tobacco is grown in small plots in more humid areas throughout the valley. Castor beans are common in the drier central crystalline region and in the western Chapada Diamantina. The estimated 691,000 cattle of the project area are also concentrated in the drier areas. 2.11 The one-fifth of project area farms (excluding an estimated 11,700 establishments operated by non-owners) greater than 50 hectares account for 84% of the valley's farm area; these properties are devoted primarily to cattle and other livestock rather than to agriculture. In terms of crops, technology, scale and commercial/subsistence orientation of production, agriculture on different sizes of farms is similar; differences that do occur are most likely to be of scale. All farmers, including cattle ranchers, grow some or all of the basic food crops, usually producing a surplus for sale. Few producers do not depend on farm production for home consumption. Estimated project area production in 1976 was as follows: Value Quantity Crop Cr$ mil. '000 tons % Bahia prod. Manioc (root) 252.8 515.9 9.3 Beans 151.2 19.9 14.2 Maize 63.3 41.1 15.5 Rice 3.1 1.0 2.6 Tomatoes 3.2 1.4 3.3 Onions 4.7 1.5 15.0 Garlic 2.7 0.2 14.0 Castor beans 95.1 38.2 24.8 Sisal 77.7 26.8 11.0 Tobacco (leaf) 107.5 12.5 36.8 Tobacco (string) n.a. n.a. n.a. Total 761.3 658.5  - 10 - 2.12 The commercial significance of crops differs from their total production. More manioc, for instance, is probably consumed at home than enters commercial trade. Production per farm is not high. Manioc accounts for about one-half of production by weight in all farm types. Most farms actually have a mixed-economy. Even small-scale farmers often have a couple of cattle or, in drier areas, sheep or goats, although ownership of all types of livestock is concentrated on the larger properties. Among the expected 17,000 families participating in the project (see para 4.04), beans (11,722 ha) and maize (10,051 ha) are the most widely planted crops, followed by manioc (7,839 ha) and castor beans (4,339 ha). Almost all maize (98.6% by area), 80% of beans and 30% of castor beans are interplanted with other crops. The estimated average value of production per hectare cultivated is Cr$4,160 (about US$280). 2.13 Physical Infrastructure. Aside from the eastern part of the project area and, particularly, the larger urban areas, infrastructural and service development is weak. The road network totals 12,300 km but many of the 9,000 km of local access roads are seasonally impassable. There are 800 km of high voltage transmission lines in the project area, but even with a current expansion project these will serve only a small fraction of the region's many rural communities. Many of the 100 "urban" centers are still without even locally-generated electricity. 2.14 Social Services. Health and education services are poorly developed and those which do exist are mainly accessible chiefly to urban populations. Approximately 47% of children aged 7 to 14 attend school (compared to 86% for all Brazil and 57% for Bahia); 38% of the population over 14 years of age is literate. Three federal, two state and eleven non-profit public and private organizations, as well as municipalities, are responsible for health care. In the entire project area, there are 74 physicians and 19 dentists. While there are 28 hospitals with 1,318 beds (one bed for every 1,000 inhabitants), half the beds are in Feira de Santana. There are 79 health centers in the project area. Reliable health data are unavailable, but public health statistics indicate high rates of meningitis, tetanus and childhood diseases like measles and whooping cough, which suggest child malnutriton. Schistosomiasis, Chagas disease (South American trypanosomiasis) and leishmaniosis are common. Only 56 of the project area's 350 cities, towns and villages have public or communal water supply systems, and only 17 of these have house connections. With the exception of one in Feira de Santana, there are no solid waste disposal systems. 2.15 Agricultural Services. While infrastructure and social services serve primarily some urban centers, specifically rural services favor large- and medium-scale farmers. Over 1974-76 for instance, more than 90% of rural credit (by value) granted in the area by the Bank of Brazil and the Bank of the Northeast of Brazil was for livestock (tending to be on large scale farms) rather than for crops. Agricultural extension services have had a similar livestock/large- and medium-farmer bias, and many of the eight cooperative societies in the area were organized by larger farmers to develop their interests. The large/medium scale farmer bias of credit and extension services is not the product of an anti-small farmer policy but rather reflects the larger-scale farmer's usually better education, broader technical and  - 11 - commercial experience, and social and political empathy with bankers and extensionists, as well as the previously predominant system of relating financing of the extension system to the volume of credit supervised (hence, usually to larger farms). 2.16 State-operated storage, mechanization and input supply services are available through a small number of outlets, but their influence has been slight. Aside from the purchase of sisal at a minimum price by the federal Production Financing Commission, there is no Government organized marketing service in the project area. III. THE EXECUTING AGENCIES State Framework for Executing the POLONORDESTE Program 3.01 The POLONORDESTE program in Bahia is coordinated by a Management Council (Conselho Diretor) created in February 1976 through state decree No. 25,136. The Management Council provides policy guidance and general coordination for the execution of POLONORDESTE-financed projects. The Council is officially headed by the Secretary of Agriculture but in practice the Governor of the State presides at most Council meetings. Other members of the Council include the State Secretaries of Planning, Health, Sanitation and Urban Development, Mines and Energy, Transport, and Education, representa- tives from the Federal Ministries of Agriculture and Interior (the latter through SUDENE), and representatives of Bank of Brazil and the Bank of the Northeast of Brazil. The Council is responsible for: setting general directives for POLONORDESTE activities in Bahia; defining overall priorities and strategy; ensuring appropriate participation in the program by executing agencies; approving the annual operating plans for specific projects and the annual plans of participating executing agencies; reviewing the quarterly execution reports of projects; and, coordinating and mediating between execu- ting agencies and resolving problems affecting project implementation. 3.02 There are four POLONORDESTE projects currently in preparation or implementation in Bahia. 1/ Day-to-day management and supervision of the projects is shared by the Secretariats of Agriculture and Planning. Coordi- nation and administration of the Paraguacu and one other project is by Project Management or Technical Units (Unidades Tecnicas) within the State Agricultural Planning Commission (CEPA-BA, Comissao Estadual de Planejamento Agricola-Bahia). These two full-time management units are headed by coordinators who answer through the chief executive of CEPA-BA to the Secretary of Agriculture. The other two projects are administered by technical units within the Secretariat of Planning. Though division of responsibility for the state's POLONORDESTE projects complicates efforts to ensure consistency in approach and implementa- tion, the arrangement results in active interest and support of the POLONORDESTE 1/ In addition to the Paraguacu project, there are the Irece, Tabuleiros Costeiros Sul ("Southern Coastal Plateau") and Alem Sao Francisco ("Beyond the Sao Francisco") projects.  - 12 - prograim by the key Agricultural and Planning Secretariats. The program also be-sfits through the close involvement of the state Governor via the Management Council. 3.03 The Paraguacu Project Management Unit (see paras 6.02-6.09) would be the executor of project management and evaluation. Other project components would be executed by existing line agencies under the general guidance and coordination of the Project Management Unit. Fifteen federal, state and private agencies and 49 municipal governments would be involved as direct executors or major collaborators in execution. Where implementation/management skills of executing agencies are weak, they would be boosted by the project, the institutional strengthening so having significance beyond the project. Each executing agency would have a project liaison officer responsible for project matters and relations with the project management unit. Institutions Engaged in Rural Development in the Project Area 3.04 The chief agencies involved in rural development in the project area are the Government-owned Bank of Brazil (BB) and the Bank of the Northeast of Brazil (BNB); the state's extension service (EMATER-BA), agri- cultural research company (EPABA), input supply and mechanization company (CAMAB), and land surveying and titling institute (INTERBA); and a number of cooperative societies and their institutional support (e.g., the Secretariat of Agriculture's Division of Cooperativism, DCOOP). Further support comes from a wide range of agencies with some involvement in rural development. Most important among these are state Secretariats of Education and Culture and of Health, the state roads department (DERBA), a roadworks consortium (CRIBA), the state rural engineering com-pany (CERB) which builds village water supply systems, small dams and some irrigation works, the state warehousing company (CASEB) and the federal Production Financing Commission (CFP). While numerous agencies are involved in rural development, their effectiveness has been hampered by both limited resources and the fact that only recently has the orientation towards integrated rural development emerged in the state. The expected roles of these agencies in the proposed Paraguacu project are described in paras. 4.08 to 4.49. IV. THE PROJECT Origins, Objectives and Approach 4.01 The POLONORDESTE program was established to raise productivity and incomes, broaden economic opportunities, and generally improve living condi- tions of small farmers in the Northeast. The program is based on the formula- tion and execution of integrated rural development projects in areas selected for both their relative underdevelopment and development potential, such as the Paraguacu River Basin. For much of this century, the Paraguacu Basin has suffered relative economic stagnation: the mining activities that were the basis of the region's economy during the latter half of the nineteenth century declined early in this century; agriculture and livestock production has been largely subsistent or directed to the internal valley market and has suffered  - 13 - from an absence of programs to improve productivity and of supporting infra- structure; and, the region has been a source of considerable out-migration by people in search of greater economic opportunity, both in agriculture in the Center West and in industrial and urban activities in the South and Center South. 4.02 In spite of its economic experience and the generally difficult environment, the Paraguacu Basin does have considerable development potential. The eastern edge is 120 km from the major consumption center of Salvador, the state capital (current population of 1.4 million), with which the eastern and central parts of the Basin are well-connected. Connections with other parts of the Basin would be improved during the project. Moreover, the Basin includes areas with both reasonable agricultural potential (like the southeastern plateau and parts of the western mountain and plateau zone) and considerable numbers of small land holders and a tradition of small-scale agriculture. The Paraguacu Basin also has growing markets and employment in its urban areas. In addition, while agriculture presently uses largely rudimentary technology and is characterized by low productivity, there is a realistic probability for upgrading production, including through small-scale irrigation schemes. In general terms, a fundamental handicap to development of the Paraguacu Basin is the absence of an efficient physical or service infrastructure. The proposed project aims to alleviate some of the most critical of these problems, so helping form the basis for self-generating development. 4.03 The proposed project originated in 1973-74 when the state and SUDENE, with the assistance of consultants, prepared a general plan for improved utilization of the region's water resources. This study, together with the seriousness of the economic conditions in the area and the creation of the POLONORDESTE program in late 1974, encouraged the state to proceed with the preparation of a special program of investments to help facilitate produc- tivity increases for the region's agriculture. The state contacted the Bank office in Recife in 1975, requesting Bank collaboration in project preparation and financing. Several Bank preparation missions followed. The initial state proposals focussed on isolated irrigation, livestock and fruit plantation programs and a small program for small-scale producers in isolated areas of the 69 municipalities all or partly in the Paraguacu Basin. With further study, and state interest in broadening project impact and in sharpening its focus on small-scale farmers, as well as Bank assistance in preparation, the project area became the core 49 municipalities of the Basin. The project is based on state proposals for this more focussed target. Original state pro- posals for project interventions in cooperative societies, mechanization, marketing and storage services were tailored and/or somewhat reduced to make them more in line with the small farmer targets and with institutional condi- tions and farmer traditions in the project area. The proposed project would seek to build on recent development activities in the Basin, including the construction and staffing of some schools and health mini-posts, some road and electrification work, and improvements to agricultural services and cooperative societies begun with POLONORDESTE funding from late 1976.  - 14 - 4.04 Because of the size (60,764 km2) of the project area, uneven dis- tribution of small farmers within the area, environmental variation, the still weak institutional base for assisting small farmers, and the need to ensure that benefits actually reach intended beneficiaries, the project would focus on sub-areas where there are concentrations of small-scale producers. Com- munities of small farmers throughout the area have been selected as primary targets. Approximately 153 communities, each having at least 100-200 small farmers, would be gradually incorporated into project activities. The distri- bution of expected direct project beneficiaries by farm size, main type of production and project sub-area is as follows: Agriculture Livestock Sub-area 0-10 ha 11-50 ha 51-100 ha Total 51-100 ha 101-200 ha Total I 1,150 740 110 2,000 - - - II 495 970 345 1,810 - 100 100 III 2,985 1,745 230 4,960 /1 - - - IV 1,320 2,260 315 3,895 360 215 575 V 2,120 1,350 190 3,660 - - - Project Area 8,070 /2 7,065 1,190 16,325 360 315 675 /1 Including up to 635 beneficiaries using irrigation, some 500 of whom would benefit from new or improved irrigation under the project. /2 Includes an estimated 2,125 sharecroppers and renters. 4.05 The proposed project is multi-sectorial and, though built primarily around agricultural development, would include significant efforts to improve production support services and physical and social infrastructure, as well as to strengthen project execution capacity. While each of the project com- ponents would have specific pre-defined objectives, to some extent these would be flexible so that they could be tailored as needed in response to local developments. In terms of target groups, project activities are in three categories: those that serve chiefly small farmer target communities; activi- ties that serve primarily the direct small farmer beneficiaries but which are accessible to a larger rural population; and components with a broader, less- specific beneficiary group. Agricultural extension, credit, land surveying- titling and irrigation would be in the first category. Benefits of agricul- tural research and seed production, input supply, storage-marketing and cooperative society activity would reach a larger group; and social infrastruc- ture and road development a still larger population. Moreover, with implemen- tation of agricultural and social extension activities and improved access to services directed initially to the 17,000 direct beneficiaries in the selected communities, project activities would have a strong demonstration effect throughout the Paraguacu valley.  - 15 - 4.06 Overall, the project is intended to increase productivity and incomes, broaden economic opportunities and generally improve the standard of living of 17,000 small farmer direct beneficiaries and their families; raise the contri- bution of the Paraguacu Basin agricultural sector to local and other Bahian food demands, and also to export earnings; and, develop further technical and administrative capacity of institutions involved in agriculture and rural development in Bahia (specifically in the Paraguacu Basin), and encourage the development of a distinct rural development/small farmer orientation within these institutions. Brief Description 4.07 The project would include: (i) intensified development of 16,325 small-farm (up to 100 ha) and 675 small- and medium-size (50-200 ha) livestock operations and provision of credit for investment and incremental working capital; (ii) expansion and strengthening of agricultural support services (including agricultural extension, agricultural research and seed production, input supply and mechanization, storage- marketing and assistance to cooperative societies), with particular attention to small farmers and problems encountered in the project area; (iii) improved access to land and to definitive land title among small-scale farmers by providing land purchase credit for approximately 820 small-scale operators (owners and non- owners) and by strengthening the surveying and titling services of the state land institute; (iv) development of basic physical infrastructure through the construction and upgrading of about 1,140 km of access roads, establishment or rehabilitation of a number of traditional small-scale irrigation schemes, building of 12 small multi- purpose dams, and construction of about 70 village water supply systems; (v) expansion and improvement of health and education services and facilities (including construction or improvement of 50 village health posts, 27 health centers and 100 rural primary schools); and (vi) provision of organizational support for project administration, including monitoring and evaluation. The project would include activities and investments undertaken during a 5-year period beginning in April 1978.  - 16 - Detailed Features 4.08 Extension. Rural extension would be a keystone of the proposed proj zct. Through an expanded and strengthened network of agricultural and social extension services, EMATER-BA (the state technical assistance and rural extension company) would promote increased small farmer use of improved agri- cultural technology and of support services such as credit, and the formation of community and farmer organizations aimed at providing or facilitating access to social and production services. The project would finance equipment, vehicles and some minor installations for the expanded extension systems; the salary, social security and insurance costs during the project period for incremental staff; incremental operating costs (materials and supplies, rents) during the project period; the participant subsistence and travel costs, materials and instructor salary costs for the pre- and in-service training of extensionists and of short-course training for farmers; and the production input costs (excepting farmer labor) of establishing a series of demonstration plots on the land of selected participants. These investments would result in the expansion of the extension system in the project area from about 77 technical staff members (equivalent to one for every 900 farms, though in practice a much smaller fraction of the farmers are actually served) to just under 300 technical staff. 4.09 As important as the increase in size would be changes in the struc- ture and methodology of extension activities, both requiring important training and retraining efforts. Rather than the principally credit-supervision approach usually adopted by the extension service in the past for its predominantly large farmer clientele, a specialization of function would be developed. It would start with middle level (high school or technical school graduate) field extensionists trained to provide, on a fixed schedule of periodic visits, agronomic and production advice to groups of up to 20 small-scale farmers organized around contact farmers and demonstration plots. Through these visits and demonstrations, field extensionists would promote adoption of improved production techniques ("technical packages") -- including recommenda- tions on soil preparation, timing of sowing, plant spacing, seed selection, fertilizer use and plant protection -- developed by the extension and research services (see Annex 9). The recommendations would be adapted to the varying local conditions within the project area and include proposals for the variety of crops and crop combinations found in different parts of the area, but including mainly beans, manioc, maize, castor beans, tobacco, fruits and vegetables. The field extensionists working with small and medium scale livestock producers in the sparsely populated central zone would work mainly with individual producers, but would take advantage of a fixed visit schedule to accommodate as many farmers as possible in a particular region. The field extensionists would be backed up by specialized credit, cooperative societies and subject matter extension staff. Credit extensionists would assist farmers to prepare and process credit proposals and help the banks supervise execution of financed farm development. Cooperative extension specialists would work in conjunction with DCOOP (see para 4.26) to strengthen the cooperative movement as a vehicle for small farmer development. Subject matter specialists would supervise and coordinate the diffusion of new production recommendations  - 1 7 - developed in conjunction with project research and experimentation programs. On the social assistance side, EMATER-BA social extensionists (usually women) would work with contact families in close collaboration with field agricultural extensionists, but would focus on providing home economics and nutrition education, developing family gardening projects and youth organizations, and promoting village health societies for project health and water supply activi- ties (see paras 4.41 to 4.46). The close collaboration between the social and field extensionists should help assure the vital local participation of both men and women in the various production and social aspects of the project. 4.10 The extension component would be managed through 4 regional offices and 22 local offices (a net increase of 7 during the project). Each agricul- tural field extensionist would eventually work with 8-10 contact farmers, through whom he would contact about 200 farmers. Each livestock field exten- sionist would work with up to 40 producers per year; and each social and credit extensionist with about 300 families. 4.11 Agricultural Research and Seed Production. In the project area there are several small research stations, and the southeast of the area is close to one of Bahia's major research centers (now the National Manioc and Horticulture Center of the Brazilian Enterprise for Agricultural Research, EMBRAPA). The project area, however, has suffered such traditional research shortcomings as the inapplicability of many research results to local condi- tions (for example, lack of focus on production systems involving interplanted crops), inadequate facilities to multiply and distribute improved seeds, weak linkage between research and extension services and between the latter and most farmers and, particularly until the creation of EMBRAPA by the Government in 1974, discontinuity in research funding and administration. Initial efforts by EMBRAPA and its state affiliate in Bahia, EPABA, to consolidate and coordinate research activities in the state and to establish stronger collabo- ration with local producers and the extension service have demonstrated that there already exists a fair amount of relevant expertise on locally applicable "improved technology" for many crops. However, there remains considerable scope to improve production through adaptive research and experimentation. 4.12 The proposed project would include: the upgrading or development of adaptive and controlled agriculture and livestock research activities at 4 small research stations in the project area (at Feira de Santana, Itaberaba, Utinga and Iraquara) already owned by EPABA but some without on-going research programs; the development of several commercial-scale observation sites managed by the EPABA stations; and the development by EPABA of improved seed production facilities (on the research stations and observation sites and through some contracts with farmers), based on a small pilot seed production program begun in 1977 with good initial results. The project costs to be financed include equipment, materials, vehicle and minor installation costs for the research stations; incremental personnel and operating costs for the stations and observation sites; and the phased start-up costs (personnel, production inputs, miscellaneous equipment, storage and handling) of seed production activities. Project disbursements for the latter would be almost entirely in the first year of the project with disbursements in the second and  - 18 - third years limited to phased expansions. Each seed production operation is intended to be self-financing after the first year as a result of EPABA seed sales to CAMAB (the state input supply company, see para 4.28), cooperatives, private dealers, or farmers. 4.13 Under EMBRAPA guidance, EPABA would be responsible for component execution. Experimentation priorities would be determined and research progress reviewed, however, through regular consultation with the Project Management Unit, the extension service and selected project area farmers. Research findings from the component would be gradually introduced on the demonstration plots of "contact farms" (see para 4.09). Research would focus initially on such subjects as improved varieties, plant spacing and protection, and nutrient application for major crops including beans, corn, manioc, and castor beans. Special emphasis would also be given to means to improve the mixed farming systems that are common in the area and in which interplanted crops and crop/livestock combinations play mutually supporting and risk-averting roles. 4.14 Credit. The proposed credit component is intended to expand signif- icantly small farmer use of the institutionalized credit system. The increased credit access would facilitate the adoption of improved farming techniques, the diversification of production and the expansion of areas cultivated, so increasing agricultural production and incomes. The component would include: (i) agricultural investment credit for about 16,325 small-scale farmers with less than 100 ha (including about 2,000 share- croppers and renters), the credit to cover mainly the costs of land clearing, on-farm storage facilities, small equipment and tools, establishment of permanent crops, and, for about 500 of the beneficiaries, the costs of on-farm and/or group off-farm irrigation works (see paras 4.30-4.32); (ii) incremental seasonal crop production credit, including credit to help cover the costs of fertilizers, seeds, pesticides, additional labor and other seasonal costs; and (iii) livestock investment credit for about 675 small- and medium- scale livestock operations of up to 200 ha in Sub-areas II and IV (the drier central zone), covering mainly stock and pasture improvement, fencing and watering points. It is expected that the great majority of the likely 17,000 benefitting families, almost 90% of whom operate farms of less than 50 ha, would be receiving institutional credit for the first time. Virtually all participants are estimated to have current agriculture incomes below the relative poverty line in Brazil (about US$380 per capita or around US$2,000 per family, see paras 8.03-8.04). 4.15 Project credit would be channelled to farmers or cooperatives through 20 existing and 3 planned branches of Banco do Brasil (BB) and Banco do Nordeste do Brasil (BNB) in or very near the project area. These federal  - 19 - banks handle virtually all of the agricultural credit in the project area and are authorized by the government to be financial intermediaries for the POLONORDESTE program. Project credit terms and conditions would be those established by the Government for the POLONORDESTE program, namely interest to the farmer at 10%, unindexed, for both investment credit and seasonal production credit (with the exception of credit for fertilizer purchase which is non-interest bearing), with repayment terms for investment credit of up to 12 years including up to 6 years' grace. The Central Bank would normally maintain credit lines with the participating banks to make available the necessary funds to cover 100% of sub-loan value. Participating banks would be required to repay the Central Bank in line with repayment of the sub-loans made, but retaining a 5% remuneration. Where lending is channelled through cooperatives, the cooperative would receive 2%, leaving the bank 3%. 4.16 Given Brazil's inflation rate, the interest rates adopted by the Government for POLONORDESTE projects are negative in real terms. Interest rates were marginally increased in early 1978 (from 7% to 10% for smaller loans and from 12% to 14% for larger loans), but a considerable subsidy remains. The Government, at least for the time being, is very reluctant to reduce further the subsidy for small farmer programs such as POLONORDESTE (see para 1.06). Since the beneficiaries of the proposed project are mainly in the lower income classes, and the possibility of misallocations of resources is greatly reduced by the proposed project extension and monitoring activities (the credit extensionist having to report at least once a year to the local bank on the progress of the individual financed projects and on the technical assistance rendered, and the project monitoring work including annual surveys of sample farms which would help provide another basis for evaluating the effectiveness of the credit intervention), the Bank should be prepared to acquiesce to the Government's position. Hopefully, the expected positive experience with small farmer repayment capacity in projects such as the one proposed will help convince the Government of the viability of adopting an interest rate policy more closely reflecting real credit costs. 4.17 During negotiations, assurances were received from the Govern- ment that it would maintain agreements with the participating banks to ensure their continued active collaboration in execution of the project credit component; that project credit would be extended to the small-scale farmers at POLONORDESTE terms and conditions, and that the Bank would be promptly informed of changes in POLONORDESTE terms and conditions. Also, to help assure that the project credit is available to as large a number of farmers as possible and that the focus on smaller scale operations is achieved, Bank disbursements would be limited to sub-loans to farmers whose total debt (investment and working capital) outstanding under the program is less than 100 MVR 1/ (or about US$5,500). Sub-loans to farm operations with income levels requiring financing of greater than 100 MVR would not be included in this project. Since livestock lending already absorbs the bulk of rural 1/ The "Maximum Reference Value" or MVR is a periodically adjusted measure equivalent to about US$50-60.  - 20 - credit resources in the project area, project livestock credit would be 'i-ited to small- and medium-scale producers in Sub-areas II and IV (the central zone) who, unlike larger-scale livestock producers, have had limited Lcc-_ss t3 credit and technical assistance and whose incomes are low. The aggregate of Bank disbursements for livestock investment credit would be limited to 15% of total Bank disbursements for the credit component. With respect to seasonal crop production credit, since repayments of short-term sub-loans would, on average, be repaid by the farmer in less than one year, the Government would make these funds available each subsequent year for reapplication under the project credit component. Bank disbursements for seasonal crop credit in POLONORDESTE fiscal year 1978 (April 1978-March 1979) would begin only after aggregate project working capital credit exceeded the comparable amount extended under POLONORDESTE in the project area in fiscal 1977. In each following year, disbursements for seasonal crop credit would be made only against credit in excess of the amount extended by the project during the previous year. 4.18 Under the POLONORDESTE program, several steps have been taken which should help expand credit coverage and help alleviate some of the constraints to small farm lending. The integrated project approach (together with comple- mentary national research and extension programs) gives special emphasis to development of improved technical packages and intensive agricultural exten- sion for small-scale operations, and to the reduction of physical infrastruc- tural bottlenecks to increased production. Also, a variety of steps have been taken within the credit system itself to improve coverage of small farmers. The most significant of these are: (i) participating banks have been instructed by the Government to give priority to small-scale and landless farmers in utilizing POLONORDESTE funds; (ii) application procedures for loans of under 50 MVR have been simplified so that banks can make such loans on the signature guarantee of the borrower, which should reduce loan process- ing cost and time for both banks and farmers; (iii) Government funding, independent of the size of individual loans, has been made available to the extension service to provide assistance to small farmers in preparing credit applications; and (iv) BB and BNB have instructed their agencies in the Northeast to initiate periodic satellite banking activities ("mobile credit" in the case of BB and pontos de credito -- "credit points" -- in the case of BNB) for small loans in municipalities without permanent bank branches, giving priority to such activities in POLONORDESTE project areas. These changes would be pursued in the proposed Bahia project, mobile banking having already been recently initiated by at least three of the project area BB branches and credit points by one BNB branch. 4.19 An analysis of the capacity of the existing banking network in the project area has pinpointed several localities (mainly in the west of the project area) where project related growth in credit demand may seriously strain the capabilities of existing bank branches. The expected demand increase is such that it may be desirable for BB to reconsider its current plans to build only 3 new branches in the project area over the next few years or for BNB to consider opening new branches. A number of aspects of the credit system infrastructure would be closely monitored under the project,  - 21 - including: the extent to which branch managers actually adopt allowable simplifications in credit procedures; the use of temporary personnel during periods of peak credit demand; the costs of processing credit; the adequacy of the interest rate spread to participating banks; the costs and benefits of the periodic or mobile banking services; the time involved and costs of credit access by farmers; the adequacy of the financing received to cover a reasonable share of total production and investment costs; and, the functioning of the PROAGRO crop/credit guarantee scheme that beneficiaries would be encouraged to utilize (at a cost of 1% of the loan amount). 4.20 Land Purchase Credit. In the project area, as in much of the North- east, ownership of sufficient land and security of tenure are important factors in determining the farmer's ability and willingness to take the often risky steps (on-farm investment, use of new farming techniques) and to get access to the support services (credit, extension and others) needed to improve productivity and expand production. This importance is reflected in several of the proposed project components, one being land purchase credit (others including titling services and assistance to cooperatives, including in land purchase and sub-division projects). Currently, Government land credit is made available only when, with the lot to be purchased, the farm is at least one "module". This is the amount of land (about 35 ha for most of the project area) defined by the National Institute for Colonization and Agrarian Reform, INCRA, to be the minimum necessary to generate in a parti- cular region or micro-region a net.farm income of four times the annual minimum wage (around US$3,000 total). While possibly justified as a way to prevent creation of excessively small farms, the "module" represents a farm size and income level well above the current situation for most farmers in the project area. The change implied by the acquisition of at least one "module" (and perceived financial responsibility) is usually so large that it precludes participation of many of the potential beneficiaries. Though INCRA can authorize exceptions to the one module minimum for land credit, exceptions have usually been granted only in cases of irrigation or highly organized colonization programs and only after very considerable administrative delay. 4.21 The state has prepared a land purchase credit component with a more flexible lower limit on the size of farms (10-23 hectares, depending on the sub-area) that might benefit from the credit, thereby allowing for a more realistic and gradual increasing of income (usually still more than doubling over about 6-8 years). Some 820 sharecroppers, tenants, and current owners of very small plots are expected to benefit, a target equivalent to 10% of the estimated project beneficiaries in the up-to-10 ha size stratum (which includes non-owner operators). The credit would be channelled through local branches of BB and BNB and would have standard POLONORDESTE land credit terms, namely, repayment in up to 20 years (including up to 6 years of grace) for up to 100% of the purchase price of land and interest at 12% unindexed. Land prices in the project area are currently in the US$150-375/ha range. The component would be financed solely with Government resources. 4.22 Execution of the component would include three important features. Firstly, short-term (up to 2 years) financing would be made available by the Government to cooperative societies to purchase larger parcels of land for  - 22 - subdivision and resale (and re-financing at the standard longer term) to new or existing small farmer members. Secondly, the state would establish a rotating land purchase fund (utilizing POLONORDESTE funding not included as a project cost) to purchase potentially usable lands where cooperative inter- vention is not possible, or to expropriate unutilized estates, both for the purpose of preparation of simple sub-division projects and resale to individual farmers who would receive project land purchase credit. Thirdly, the component would benefit from the collaboration of the cooperative assistance, extension and land titling agencies of the state to help: identify potential beneficiaries in the target group for land purchase credit and also land with productive potential available for sale; facilitate farmers' access to credit and compliance with various requirements (credit proposals, titling); and prepare state or cooperative land purchase, subdivision and resale projects. As the component would affect a relatively small part of the very large project area and a very modest number of farms, it should not, by itself, result in significant land speculation. Possible speculation effects should also be minimized by the state's effort to achieve most of the component target through a relatively small number of large land purchases by the state or cooperatives. One of the key elements in assuring that component targets are achieved and that this relatively modest initial effort can be expanded later to benefit a larger share of the state's numerous non-landowner farmers or currently very small-scale owner-operators will be flexibility in the application of the "one-module-minimum" criterion (para 4.20). During negotiations, the Government agreed to prepare and furnish to the Bank not later than December 31, 1978 new guidelines permitting financing of smaller plots of land so long as the plots by themselves or together with any land owned by the farmer concerned are estimated to yield for each adult family labor unit to be absorbed at least the minimum regional annual wage. Account would also be taken of the managerial and financial capacity of the farmer concerned and the socio-economic situation in the area where the plot is located. The Government would concurrently prepare an administrative plan to expedite more efficiently land purchases by small farmers in the project area and to speed the processing of applications for small land purchase loans, the plan to be be implemented after consultation with the Bank. 4.23 Land Surveying and Titling Services. Prior to the state's passage of a new land law in 1972 and creation of the Bahia Land Institute (Instituto de Terras da Bahia, INTERBA) in 1975, the administration of lands and security of land ownership in Bahia suffered serious weaknesses. Speculators in public lands were able to amass considerable holdings and smallholders were frequently dispossessed of their land by forced removal, intimidation or destruction of crops and property. Though the 1972 law defined the legal procedure for granting titles and disposing of public lands, it is estimated that still less than 5% of the farmers in the project area who consider them- selves "owners" of their land actually have title. The percentage is even less (2%) for farmers with under 100 ha. The lack of definitive title is a serious constraint to farmers in developing permanent increases in their incomes because the farmer without title remains subject to dispossession (a phenomenon very difficult to control). For all farmers, but especially small farmers, who have little political or economic power, security of tenure reduces "uncertainty" while also increasing the incentive to invest- ment and facilitating access to production services, particularly credit.  - 23 - 4.24 The project would significantly expand the capacity of INTERBA to survey the lands of, and grant titles to, small-scale farmers in the project area. A total of 15,000 titles (roughly equivalent to the number of "owner- operators" expected to participate in the project) would be granted. The project would finance the costs of office and surveying equipment, vehicles and materials required for a small headquarters unit in Salvador "ad 6 field offices serving the project area (4 established and 2 being established in early 1978); and the staffing, operating and maintenance costs of the system during the five-year project period. Field operations would include promotion work, initial title request processing and property surveys, with final legal clearances and processing done in Salvador. Since current cadastral data, a key input in the titling process, are not available in much of the project area (land disputes recently have been particularly common in parts of the area), and to help avoid the virtual stoppage of titling activity that could result from localized disputes or litigation, INTERBA would carry out full cadastral reviews in some localities during the first two years of project execution, in addition to its regular surveying-titling work. Because of the importance of clarifying and resolving these problems quickly, INTERBA has already established a cadastral review unit at its Utinga agency. Agreement was reached with the Government during negotiations that the establishment of 2 additional cadastral review units (at Seabra and Maracas) be a condition of disbursement for the component. Overall, the intensive field work by INTERBA should also put the agency in a good position to help advise the Project Management Unit on problems encountered by the area's non-landowner farmers in their rental or sharecropping arrangements and possible improvements which the project might subsequently promote, as well as to help identify potential beneficiaries for the land purchase credit component (para 4.22). 4.25 Taking into account all expected baseline costs of the component, the cost per title would average around US$350 equivalent, over half attri- butable to surveying costs. As the service is considered a prerequisite to developing small-scale agriculture in this area, the state would charge the farmer only a nominal fee, usually amounting to less than 5% of the cost. 4.26 Cooperative Society Support. With one exception, the 8 cooperative societies currently operating in the project area are composed mainly of livestock ranchers and larger farmers. The exception, the Serrinha coopera- tive, processes sisal. Otherwise, the principal activity of the cooperatives is to offer input supply services to members. Cooperatives receive (as groups and individuals) agricultural technical assistance and cooperative promotion advice from cooperative specialists of EMATER-BA, the state exten- sion service, which also collaborates with the Cooperatives Division (DCOOP) of the Secretariat of Agriculture in providing managerial and bookkeeping advice. As part of the project, the state intends to reinforce this support, and through this help increase the small farmer participation in some of the cooperative societies. Given, however, the possibly conflicting interests of the different types of producers in some of the existing cooperatives, the project may also see the gradual development of new cooperatives or "pre- cooperatives" from the farmer groups stemming out of project extension activi- ties. Whether with existing or new cooperatives, pre-cooperatives or informal  - 24 - farmer groups, the final objectives of the project support would be to develop re-ter financial and managerial self-sufficiency of the cooperatives or groups as a mechanism for improving services to small farmers and, where cooperative society strength and demand warrant, to expand the range of their services from input supply to certain marketing activities, credit channeling, and the preparation of land purchase and redistribution projects. 4.27 The project would finance salaries and travel costs during the project period for an expanded DCOOP headquarters and field staff working with project area cooperatives (some of this staff would be seconded to the cooperative society and gradually paid for, according to a pre-arranged schedule, by the society concerned); minor equipment and vehicle purchases; costs of pre- and in-service training for DCOOP and cooperative society staff; and miscellaneous DCOOP operating and materials costs. Execution of the component would be accompanied by intensified coordination by the Secretariat of Agriculture of the activities of the various agencies currently associated with cooperative development, including (in addition to DCOOP and EMATER-BA) the National Cooperative Credit Bank, the Bahia Organization of Cooperatives, the National Institute of Colonization and Agrarian Reform, and others. 4.28 Mechanization/Input Supply. The bulk of planned on-farm development (mainly land clearing, permanent crop installation and initial soil preparation) expected among project beneficiaries would be labor intensive. Certain develop- ments, however, such as some land leveling and drainage works for irrigation and construction of dams for the small and medium sized livestock producers would require mechanized support. Few of the expected participants have motorized farm machinery, though in some areas local merchants or large-scale farmers and/or CAMAB (the state mechanization service and agricultural supplies company) hire out machinery. CAMAB's service network does not cover many parts of the project area. CAMAB does not monopolize (nor is it likely to) 'the provision of farm inputs and mechanization services but its presence is viewed by the state as an important means to assure that at least the inputs (includ- ing, for example, the improved seeds produced under the project experimentation and seed production component) and mechanization services required for state supported development efforts are available when and where needed and that, where there are few other suppliers, prices remain competitive. 4.29 The state policy is for CAMAB to operate insofar as possible as a self-financing entity. CAMAB is accustomed to arranging local bank financ- ing for its commercial activities and setting its prices to ensure that costs and debt service are covered. Parallel to the project (without project financing), CAMAB would establish around 20 new retail input supply outlets, mainly using small rented buildings. Cooperative societies might eventually take over operation of several of the input supply points and CAMAB concen- trate on wholesale supply of inputs in the project area. With respect to mechanization services, a small amount of project financing would be required to allow CAMAB to relocate a part of its machinery pool to the currently poorly served western part of the project area. This is also where mechani- zation services will most likely be required for irrigation related invest- ment. Project costs would be limited to installation of a new equipment storage depot and workshop in Iraquara, and purchasing of several vehicles and  - 25 - office equipment. The tractors and farm equipment to be used by CAMAB would come initially from CAMAB's existing pool, with any later necessary purchases financed (parallel to the project) by local banks as is CAMAB's current prac- tice. During negotiations, assurances were received from the state that it would cause CAMAB to take all the steps necessary to establish or maintain input supply and mechanization services as required to support project farm development. The signing of a satisfactory project participation agreement between the state and CAMAB would be a condition of disbursement for the component (para 6.11). 4.30 Irrigation. Though soils in parts of the Paraguacu Basin would permit an economic development of irrigation, very little of the project area is currently irrigated, for lack of both systematic technical assistance and the physical and service infrastructure required for any significant agricultural development. The project would seek to introduce and foster improved irrigation and would include the studies and technical assistance for, and installation of, several small irrigation and drainage schemes. These would comprise mainly the rehabilitation or development by groups of farmers of small irrigation projects using water from local rivers, wells and proposed multi-purpose dams (see para. 4.33). An estimated 500 farmers, irrigating a total of 1,500 ha, are expected to benefit. 4.31 The potential works would be identified by EMATER-BA extension agents and the necessary preparation and design work would be carried out by CERB, the state's rural engineering company, which will also be executing the multi-purpose dam and village water supply (para 4.44) components. Rehabilitation and expansion works would be carried out on about 5 existing small schemes (3 already pre-identified) that would eventually irrigate a total of about 400 ha. New schemes would be built around several existing wells, most of the proposed small multi-purpose dams and at several sites along the area's permanent rivers (e.g., the Santo Antonio, Una, Utinga and Paraguacu itself), to irrigate eventually a total of about 1,100 ha. Usually less than 30 farmers would be involved in any one scheme. Works would be carried out by existing or new farmer groups and water user associations under CERB supervision and, as necessary, contracting the state mechanization service company, CAMAB, or private companies for construction. The study, training and technical assistance costs (personnel, materials, contracted services) of CERB would be financed by the project on a grant basis. The off-farm or communal irrigation works and equipment would be financed through group loans, under the project agriculture investment credit. On-farm works would also be financed under the project credit component, but through either individual or group loans, depending on the situation. 4.32 It is expected that some 500 farmers would benefit under the irrigation component, with about 3 ha irrigated each, at an average total cost per farmer of some US$4,200, excluding studies and farmer training. During the first year of the project, the project unit (with the collaboration of EMATER-BA, CERB and the participating banks) would prepare a brief manual outlining the sub-project selection, preparation and evaluation criteria (e.g., rate of return, technical feasibility, organization) to be utilized.  - 26 - During negotiations, agreements were be reached that: (i) a condition of disbursement for this subcomponent would be the preparation of a draft manual on sub-project preparation and evaluation criteria satisfactory to the Bank; and (ii) the detailed plans for and evaluation of any one small-scale irriga- tion scheme requiring more than a total of US$250,000 in off- and on-farm investment would be provided to the Bank for agreement prior to inclusion in the project. Details of the other minor schemes financed under the project would be made available to the Bank for review during the course of project supervision. 4.33 Multi-purpose Dams. The central part of the Paraguacu Basin has traditionally suffered from irregular and low rainfall and the fact that ground water of adequate quality and quantity can be tapped only at very deep levels in most of the area. The situation results in considerable social hardship as well as difficult-to-quantify costs associated with labor produc- tivity losses caused by use of contaminated water and losses in agricultural and livestock production for lack of water. Not unlike other parts of the drought-prone Northeast, this situation in the past has led to various pro- posals to undertake massive public works programs in dam construction. Several such programs have been executed (mainly in other Northeast states) and some of the dams have been very beneficial in helping communities avoid long-distance water transport costs, animal deaths and severe hardship during drought. Relatively few, however, have been fully exploited for continued use for village water supply, fisheries or irrigation. The Ministry of Agriculture did construct a small number of dams in the central part of the Paraguacu Basin during the 1960's and an effort is currently being made by the state to review this experience and to try to adapt it to the needs and targets of the proposed rural develoment activities, especially taking into account the virtual certainty that small dams would be the least-cost village water supply solution for a fair number of the target communities. As part of this effort, the project would include the feasibility studies for the design and construc- tion of 12 small earth dams. 4.34 It is expected that the dams (with average earth volumes of 12,000 m3 and water volumes of up to 250,000 m3) would be located and sited to serve village water supply as well as small irrigation schemes. Construction of the dams (for which design and construction costs are expected to average roughly US$85,000 each) would utilize local labor to the extent possible. Once constructed, the dams would be subject to the general operation and maintenance supervision of specially assigned CERB technicians (one for every three dams) with day-to-day operation of water supply systems the responsibility of the community health societies (paras 4.44-4.46) and of irrigation the local water user association or irrigation group (paras 4.30-4.32). Some dams could also serve small fisheries projects and provide emergency watering points for livestock during droughts. Use for electricity generation is not foreseen. The component would be executed by CERB, the state's rural engineering company. Assurances were received during negotiations that: (i) by December 31, 1978, feasibility studies and preliminary plans, satisfactory to the Bank, would be completed for the 12 dams; and (ii) the state would cause CERB to provide appropriate supervision and periodic maintenance of the dam structures, once they are completed.  - 27 - 4.35 Crop Storage. The project would include the construction and equipping of six relatively small (about 700 ton capacity) storage units by the state storage company, CASEB, for sack storage of mainly beans and maize. The component is based on a conservative estimate of storage demand (to handle less than total potential storage needs for these crops in the project area), partly because of the past very low level of small farmer use of uff-farm storage. This situation stems from small farmers' reluctance to use distant units and incur relatively high transport costs for the small amount of storable production per individual farmer, as well as a tradition of on-farm sack and even small silo storage. Improved on-farm storage facilities for crops such as beans, corn and manioc flour are among the on-farm investments expected to be financed under the project's agricultural credit component. At the same time, however, the state wants to test in several selected sub-areas (so taking advantage of project access road improvements) the functioning of two complementary mechanisms that might provide important long term improve- ments in the market and production conditions encountered by the small producer. 4.36 One mechanism would be the Government minimum price scheme of the Production Finance Commission (CFP), under which a farmer can receive short term Bank of Brazil market financing, the amount financed being based on the minimum price valuation of stored production. Except for sisal, the minimum price scheme has not operated in the project area for lack of storage facili- ties, limited CFP interest in other than export crops and BB (as the CFP agent) and CFP reluctance to deal with a large number of individual farmers. The second mechanism would be the advance purchase scheme (CAP, "compra anticipada de producao") being tested under the supervision of the Brazilian Food Company (COBAL) in several Northeastern localities. Under this scheme, very small- scale producers organized in small groups can get limited advances against promised production of certain crops and so have a guaranteed buyer for at least the minimum price. Advances are a percentage of individual sales contracts, the total value of the latter never exceeding US$800-900 equivalent. The functioning of both mechanisms will require local bulk storage capacity, as well as the promotional support of project extension activities and the development of farmer groups and cooperative societies planned under the project. 4.37 The project-financed storage units (which, if the above-mentioned tests prove successful, the state hopes to replicate in 12 other locations in the project area) would have a modular design and could eventually be expanded to 3,000 ton units. They would be constructed along the lines of a standard model with cement floors, walls of local concrete building blocks and roofs of sheet metal or local tile. Construction would be contracted by CASEB to local construction firms. The units would be located in significant commercial centers not already served by CASEB and near Bank of Brazil agencies. Probable locations are Castro Alves, Iacu, Iraquara, Barra da Estiva, Ipira, and Lamarao. CASEB would apply its periodically adjusted standard storage tariffs. With a unit annually having about 3,600 ton-months of produce storage and 900 ton-months of input storage by the state input supply company (CAMAB), cooperative societies or local merchants, as well as fumigation services, the tariffs would cover expected minimal storage opera- ting and maintenance costs and capital cost recovery over about 25 years.  - 28 - During negotiations agreement was reached that the signing of project partic- -pation agreements with COBAL (to operate and arrange recurrent financing for advance purchase scheme activities), CFP (to broaden operations of the minimum price scheme in the project area) and CASEB (to build, operate and maintain the storage units) would be a condition of disbursement for the storage compo- nent. 4.38 Feeder Roads. In the Paraguacu area there are some 9,000 km of unpaved municipal roads linking rural settlements with municipal centers and the state and federal highway systems. Most municipal roads are no more than one-lane earth tracks impassible for much of the rainy season; those in the less densely populated western highland part of the project area are often dangerous, having steep gradients and tortuous bends. In order to help assure a greater use of cultivable land, farm inputs and agricultural and social services and greater access to markets in the selected project micro- regions, the project would include engineering design and construction or upgrading of about 1,140 km of low-cost feeder and access roads. The roads, classified mainly according to projected traffic, would include: (i) about 404 km of improved one-lane (4.5 m wide) earth roads, the construction works costing on average US$10,700/km; (ii) 517 km of gravel surfaced all-weather one-lane roads, costing an average of US$12,700/km; and (iii) 218 km of gravel surfaced two-lane (6m wide) roads, costing an average of US$21,000/km. The design standards for the three road classes (see Annex 1) were agreed during negotiations. Execution of the works would be the responsibility of CRIBA (the Intermunicipal Road Consortium of Bahia), a largely state-owned company created in 1966 to assist municipalities study, engineer and construct local roads. CRIBA has a small but competent staff. It contracts most of its engineering design and construction works to local firms. It would contract consultants on terms and conditions satisfactory to the Bank to assist in the design of project roads. 4.39 The priority road sections (a total of 116 sections averaging about 10 km) were identified in an extensive road inventory and feasibility study carried out by the state in light of agricultural development plans. That study also identified some 700 km (14 sections) of higher standard federal and state collector roads that will be necessary in the Paraguacu area to support the agricultural development activities as well as other expected transport and economic developments (see Annex 2). The state is planning to submit these latter roads to BNDE (the National Economic Development Bank) for financing under the secondary and feeder road program partly financed by Bank loan 1207-BR. Agreement was reached with the state during negotiations that it would construct the 14 collector roads by March 31, 1982 (the end of the fourth year of the project). 4.40 Maintenance of municipal roads is the responsibility of each municipality. Although the state highway department (DERBA) assists by hiring its motorgraders to municipalities and charging only for operating costs, the municipalities are usually short of maintenance funds and, consequently, municipal roads are often neglected. The state of Bahia is currently preparing (with a view to requesting Bank financing as part of a future highway project) a program to strengthen and expand its capacity to  - 29 - maintain state and federal highways. The state is also trying to define suitable arrangements to provide regular maintenance of municipal roads (in which labor-intensive activities such as grass cutting, ditch and culvert cleaning and pot-hole repairing would remain the responsibility of the muni- cipalities). Assurances were received from the state during negotiations that by December 31, 1978 it would provide to the Bank for comment its program for the strengthening and expanding of DERBA's maintenance capacity in the project area, including the definition of required equipment purchases, training plans for both DERBA and municipality staff, and sources and amounts of funding planned to assure execution of the program. Assurances were also received that the state would maintain the project roads and that, prior to starting construction of project road works in a particular municipality, the state would secure the municipality's agreement to allocate the necessary funding and manpower to assist DERBA in maintaining appropriately the roads constructed or upgraded under the project. 4.41 Health Services. The project would develop an expanded low-cost rural health delivery system that would give primary emphasis to maternal and child groups and rural workers and at full development would serve directly some 110,000 people. The health component would be administered by the State Secretariat of Health (SESAB) and its operational arm FUSEB (Bahia State Health Foundation) and would include: (i) the construction and equipping of 50 village health posts ("miniposts"); (ii) the construction, upgrading or equipping of 27 health centers that would offer various higher level health services and partly act as referral points for project-financed miniposts as well as for some 40 similar village health posts constructed in 1976 and 1977 by POLONORDESTE; (iii) the development of an improved building and equipment maintenance system and provision of spare parts for health posts and centers; (iv) the training of health workers; and (v) the strengthening of the supervisory and project execution capacity of SESAB/FUSEB. The project would finance construction and equipment costs; instructor salary, participant travel and boarding, and materials costs of the training program; and salary complements, technical assistance, equipment and operating costs for health component administration in SESAB/FUSEB. 4.42 Execution of the health component would build upon a significant community organization and promotion effort by the social extensionists of EMATER-BA (see para. 4.09), who would assist in the formation of village health societies. These societies would help define community health service priorities, contribute materials and/or labor to health post con- struction, offset a small part of health service costs through modest health pre-payment fees, contribute to the financing and managing of village water supply systems (see para 4.46), and in some cases provide a basis for greater community involvement in other project development activities. Each local health post would be staffed with a SESAB-paid health attendant. Local health services would be very simple, focussing on the provision of pre-natal care, vaccinations, health and nutrition education, and selected curative treatments. The health post would also provide a valuable mechanism for executing parallel Government endemic disease and nutrition programs. Similarly, the project health centers would also serve as channels for food supplementation activities financed by the National Food and Nutrition Institute, INAN, for undernourished  - 30 - referred patients. Project training would be directed toward the rural health attendants, traditional mid-wives, regional health service supervisors, health center attendants and laboratory auxiliaries. 4.43 The health component is of a very modest size for a variety of reasons, including current administrative weaknesses in the regional offices of SESAB/FUSEB and inadequacies in previous training and supervision efforts and in the coordination of health services. Also, preparation of a special endemic disease control sub-component (schistosomiasis, Chagas disease and leishmaniasis are prevalent in the project area) could not be undertaken as SUCAM, the Government's Superintendency for Public Health Campaigns, has recently had to allocate virtually all of its attention in Bahia to combatting the development of the mosquito vector of yellow fever in the Greater Salvador area. However, development of endemic disease control programs in the project area will be necessary: (i) to help assure that project-financed improvements in health services are sustained and health and water supply component objec- tives fully achieved (endemic diseases being among the important health problems, and the sanitation and pit latrine programs likely to be included in a schistosomiasis control program being key complements to the water supply program); and (ii) to help ensure that health conditions not be inadvertently worsened by other project activities (e.g., increasing schistosomiasis in irrigation areas or importing malaria as a result of greater population mobil- ity). Agreement was reached during negotiations that: (i) the Government would prepare endemic disease control programs for the project area, covering at least schistosomiasis and malaria, provide the program plans to the Bank by December 31, 1978 for comment, and execute the programs promptly thereafter; (ii) by December 31, 1978, the Bank would be provided, for comment, the details of training models, training objectives and techniques, course outlines and training evaluation criteria for the rural health attendant training courses; (iii) consultants would be contracted, on terms and conditions satis- factory to the Bank, to assist in the preparation of a health service manage- ment information system; and (iv) that conditions of Bank disbursements for the health component would be (a) the completion of supplementary project staffing by SESAB/FUSEB of the regional offices to be in charge of component execution and (b) the signing of project participation agreements between SESAB/FUSEB and EMATER-BA (regarding the latter's role in community organiza- tion work and collaboration in supervision), SESAB/FUSEB and the municipal governments (regarding their commitment to donate land for and help finance operations of health or sanitation facilities) and SESAB/FUSEB and FUNRURAL (also regarding contributions to the financing of health costs). 4.44 Village Water Supply. In conjunction with the proposed health activities, the project would provide for the study, design and construction of about 70 village water supply systems providing potable water to some 35,000 people; the strengthening of state and local capacity to maintain such systems; and the training of village water system operators. As with other health-related services, one of the key initial project activities would be the EMATER-BA community organization effort, which would need to involve both men and women. The community health societies formed to help support local health miniposts would also be responsible for (i) contributing land, funds,  - 31 - labor and/or local materials to help finance construction of the water system; (ii) operating and managing the systems through locally hired sanitation auxiliaries; and (iii) collecting user charges to offset operating costs. In some cases where, because of the spatial distribution of the population, a fairly large number of the health society members are not able to benefit directly from the water services and therefore need not be expectcd to contri- bute directly or as fully to water system operation and maintenance, it may be desirable to form a separate society (or sub-group within the health society) to manage the water system. The water systems would range in complexity from very simple systems (deep well, pump, storage tank, public hydrants) in villages with up to 500 inhabitants to systems including elevated storage tanks, more extended distribution and some house connections in towns with up to 2,000 people. Water system design and construction would be carried out by CERB, the state's rural engineering company, which would also be responsible for providing, through a new regional office in the project area, supervision and maintenance backstopping. CERB has considerable experience in well-drilling and small-scale water supply system construction. Community contributions to installation would be reimbursed by CERB in the form of the establishment of special community funds usable only for asset renewal or major maintenance. 4.45 Construction of several already identified systems could begin in the first year of the project. However, because in places adequate water sources are difficult to tap, and because the success of the component will hinge on the adequacy of community organization and maintenance capacity, CERB will need to carry out several special studies. The project would finance the costs of these studies as well as the equipment, materials and construc- tion costs of the water systems and the cost of regional maintenance equipment and a spare parts pool to be administered by CERB. One study would be a global feasibility study to identify water sources, locate and test wells, define technical standards and other criteria (e.g., benefitting the largest possible number of the low-income part of the local population at the lowest feasible cost) for various types of systems, detail investment and operating cost estimates, and define the remainder of the phased investment plan. A second series of studies would define a training program for local sanitation auxiliaries and prepare simplified manuals for local operation and maintenance, and also would detail the personnel and equipment require- ments and operating plans for CERB to offer regional supervision and main- tenance. A third series of studies, conducted in collaboration with EMATER-BA, would develop detailed strategies and work plans for the pro- motion of the proposed community health societies. These studies would also define the user charge collection mechanism that would ensure communities assume (gradually, if necessary, in smaller or poorer communities) at least the full operating and maintenance costs of the system and, where immediate community coverage of operating and maintenance costs is not possible, the definition of specific funding arrangements and agreements for phasing out the subsidies. 4.46 During negotiations agreement was reached with the Government that: (i) by December 31, 1978, the Bank would be provided, for comment, the results of the pre-investment, maintenance and organization studies (including draft  - 32 - operating manuals and training plans); (ii) completion of a project participa- -s-in agreement, satisfactory to the Bank, between SESAB, CERB and EMATER-BA outlining their respective responsibilities on the component would be a con- ditIon of disbursements for the component; and (iii) prior to starting con- struction of a particular system, the recipient community would have already established a health society or water user association, agreed formally to contribute a specified amount in kind or cash to the construction of the system, and also agreed to define and collect the user charges necessary to finance the system's operating and maintenance costs, with the exception that, in communities of under 500 population, user charges could initially cover as little as 50% of operating and maintenance costs but only if the community had the prior formal agreement of the municipality to provide the necessary supplements to locally collected user charges during a specified period of not more than 5 years, after which the community would be fully responsible. 4.47 Education and Vocational Training. The education component would expand education and training opportunities in the project area as well as substantially improve the quality of rural primary education. Execution of the component would be managed mainly by specially assigned staff in the Secretariat of Education (SEC), and would include: (i) the construction, furnishing and equipping of 100 rural primary schools for a total of 14,000 student places, double shift; (ii) the preparation and introduction of a new curriculum and new teaching materials for grades 1-4 in project area rural schools; (iii) the training and/or upgrading of 1,200 project area rural teachers and 259 education and training supervisors; (iv) the provision of occupational training, through the National Services for Industrial and Commercial Apprenticeships--SENAI and SENAC, respectively--for about 3,000 youths and adults in the subjects of carpentry, drawing, electricity, masonry, mechanics, accounting, secretarial skills and commerce; (v) the carrying out, by the overall Project Management Unit, under agreements with the Federal University of Bahia, of several special studies on training needs (of both men and women) and employment projections; and (vi) the strengthening of the capacity of SEC to administer and supervise the component. The project would finance design, civil works, equipment and furnishings for the primary schools; consultancy and printing costs for development of the new teaching materials to be tested; consultancy fees, participant transport and boarding costs, teaching materials and trainer salaries for the teacher and supervisor upgrad- ing program; instructor salary, subsistence and travel costs and materials and operating costs for the vocational training services contracted with SENAI and SENAC; the fees, materials and operating costs of the studies; and salary complements for full-time formal and non-formal component coordinating and supervision units in SEC and full salaries for staff to reinforce the construc- tion supervision staff of SEC during the respective sub-component execution periods, and the equipment, materials and operating costs of component admin- istration. 4.48 New rural primary schools would be constructed according to proto- type plans, which have been reviewed and found satisfactory, in sites where the population is at least 1,500 within a catchment. area of about 28 km2; where insufficient or inadequate facilities exist and students need to walk more  - 33 - than 3 km to reach the school; and where year-round vehicle access exists. Agreement to this effect was reached during negotiations as was agreement that: (i) a detailed site plan, satisfactory to the Bank, for schools to be built under the project during any given POLONORDESTE fiscal year would be provided to the Bank by November 30 the preceeding year; and (ii) disbursements against construction costs would be conditioned upon provision to the Bank of satisfactory evidence of acquisition of the respective sites. 4.49 The teacher/supervisor upgrading, curriculum development, special studies and occupational training sub-components are all facets of an effort to improve the quality of rural education in the project area and to adapt educational and training opportunities to local needs. A further facet of this effort, to be carried out parallel to the Bank-financed activities, would be the improvement of rural teachers' salaries (previously the respon- sibility solely of the often fiscally weak municipalities), through the Government program already in motion to increase gradually federal revenue sharing to the municipalities (part of which should be available for rural education financing) as well as increased state contributions to rural educa- tion financing. To help assure the effectiveness of the package of "quality improvement" measures, assurances were received during negotiations that: (i) prior to starting project school construction in a particular municipality, the state would secure the municipality's agreement to give priority to the continued employment of the trained teachers following their successful completion of the training course and pay at least the minimum salaries guaranteed by SEC for teachers on the state's payroll; (ii) the state would give priority to the continued employment of project-trained education super- visors following their successful completion of the training course; (iii) by December 31, 1978 the state would complete a final draft, satisfactory to the Bank, of the revised curriculum and would initiate use of the new curriculum no later than the beginning of the 1980 school year; (iv) by June 30, 1979 the state would provide the Bank for comment the results of the first of the several planned socioeconomic and long-term training need surveys and of the employment projection study and appropriately incorporate these results in the detailed project vocational training plans for subsequent years; and (v) the state would provide adequate funding to maintain appropriately the education facilities and services developed under the project. A condition of disburse- ment for the component would be SEC's entering into satisfactory project agreements with SENAI and SENAC. 4.50 Project Administration, Monitoring and Evaluation. Though the project would mainly be executed by existing institutions, a special Project Management Unit ("Unidade Tecnica") would be maintained in CEPA-BA to coordi- nate the planning, implementation, monitoring and control of finances of the project (see paras 6.02-6.09). This unit would include a headquarters staff in Salvador as well as 3 small regional field offices within the project area. The state would also establish a separate evaluation group (see paras 6.13-6.14) answering to the President of the State's POLONORDESTE Management Council. Project investment costs include the costs of equipping, staffing and providing necessary materials and contracted services to maintain these groups during the five-year project period. Included would be the equivalent of about 224  - 34 - man-months of local consultancy services (at an average cost per man-month of US$3,000 equivalent) for various studies on either special technical or eval- uation issues, as well as data processing services. During negotiations, assurances were received from the Government that the qualifications, experience and terms and conditions of employment of the consultants would be satisfactory to the Bank. V. PROJECT COSTS AND FINANCING Cost Estimates 5.01 The total project cost over the 5 year project period (April 1978 - March 1983) is estimated at US$106.6 million, which includes physical contin- gencies 1/ equivalent to 6% of the baseline cost, the price contingencies 2/ equivalent to 26% of baseline cost plus physical contingencies. Baseline costs have been estimated at appraisal prices, namely, late 1977 Cruzeiros converted to US$ at rate of US$1.00 to Cr$15.00. The estimated foreign exchange compo- nent is US$17.5 million or about 16% of total project costs. The project cost summary is given overleaf and the yearly phasing is given in Annex 3. 1/ Physical contingencies of 7% were added to the extension, experimenta- tion, seed production, land titling, cooperative society support, mecha- nization/input supply and project administration components; 10% to the small-scale irrigation, multi-purpose dam, feeder road, storage and education components; and 15% to the health and sanitation component. The physical contingencies were established to reflect, by component, the extent to which adjustments may prove necessary in physical quantities. This, in part, reflected the extent to which baseline costs were derived from final engineering design or component detail and, where possible, experience with recent comparable investments. On the road component, for example, though detailed engineering is not yet available, the baseline cost estimates were derived from a study of each of the road sections expected to be built or upgraded, taking into account quantities, costs and typical adjustments during the final design stage of recent comparable works. The 10% physical contingency reflects possible further quantity increases which may prove necessary during construction of the roads. 2/ Price contingencies were calculated in US$ terms as follows: for civil works, 9% in 1977, 8% in 1978, 7.5% in 1979 and 7% per year thereafter; for equipment, 9% in 1977, 7% in 1978, 6.5% in 1979 and 6% per year thereafter; for other materials and supplies, credit, salaries, etc., 8% in 1977 and 7% per year thereafter.  -35- NORTHEAST BRAZIL BAHIA RURAL DEVELOPMENT PROJECT - PARAGUACU Total Project Costs (US$ millions) % of Total Credit Local Foreign Total Baseline Costs X Foreign Exchange Investment Credit: Agriculture 10.42 1.16 11.58 14.5 10 Irrigation (On-Farm) 0.54 0.14 0.68 0.8 20 Irrigation (Group, Off-Farm) 1.06 0.35 1.41 1.8 25 Livestock 2.92 0.32 3.24 4.1 10 Increment4l Worziua Capital Credit 8.00 1.03 9.03 11.4 15 Land Purchase Credit 1.85 0.00 1.85 2.3 0 344.9 Production Support Agricultural Services: Extension and Demonstration 9.51 1.27 10.78 13.6 12 Experimentation/Seed Production 2.31 0.28 2.59 3.2 11 Cooperative Society Support 0.53 0.06 0.59 0.7 10 Mechanization/Tnput Supply 0.06 0.02 0.08 0.1 28 Land Titling and Surveying 4.83 0.58 5.41 6.8 11 Irrigation Technical Asst./Studies 0.48 0.08 0.56 0.7 15 25.1 Infrastructure: Pilot Multi-Purpose Dams 1.03 0.40 1.43 1.8 28 Off-farm Storage 0.29 0.07 0.36 0.4 20 Feeder Roads 11.59 5.41 17.00 21.4 32 23.6 Social Infrastructure Health and Village Water Supply 3.14 0.70 3.84 4.8 18 Education and Vocational Training 4.67 0.86 5.53 6.9 16 Project Administration Project Management/Monitoring 2.78 0.31 3.09 3.9 10 Evaluation 0.55 0.06 0.61 0.8 10 TOTAL BASELINE COSTS 66.56 13.10 79.66 100.0 16 Physical Contingencies 4.09 0.80 4.89 6.1 16 Price Contingencies 18.45 3.62 22.07 26.1 1/ 16 TOTAL PROJECT COSTS 89.10 17-52 106.62 133.9 16 1/ Price contingencies as percentage of baseline costs plus physical contingencies.  - 36 - Financing 5.02 The proposed Bank loan of US$37 million to the Federative Repu- blic of Brazil would finance 35% of total project costs, the balance being financed by the Federal Government under its POLONORDESTE program. The loan would be for 15 years including 3 years of grace. It would cover the project's full foreign exchange costs of US$17.5 million equivalent, as well as US$19.5 million or 22% of the local costs. 5.03 In a country like Brazil, which is making a vigorous effort to mobilize domestic resources, it would be appropriate for the Bank to give assistance in financing some of the local currency expenditures on projects like the proposed, which have very high priority and yet have a relatively low foreign exchange content (in this case, 16%). The situation typically arises in agriculture and rural development projects in Brazil in view of the high capability of the Brazilian economy in the production of the inputs required for investments of this type. If the Bank is to be effective in supporting a project of this kind, it seems reasonable for it to finance at least 35% of the costs even though this entails financing some local currency expenditures. During negotiations, assurances were received from the Govern- ment that adequate funds would be made available promptly for the purpose of effective and timely execution of the project. The Government has also assured that it would operate and maintain appropriately the facilities and services developed under the project and arrange adequate, timely funding for that purpose. The state has assured that it would arrange for the timely acquisition of any additional land or rights-of-way which prove necessary for project works. 5.04 To help assure a prompt start of the project, various activities such as the recruitment and training of staff for the project unit and extension service, the engineering design of feeder roads and hydrological surveys to locate water supply sources, the initiation of on-farm investments and provision of credit will have begun during the first half of 1978. For this reason, retroactive financing of up to US$1,000,000 equivalent is proposed to help cover eligible expenditures after March 31, 1978, but before the loan is signed. Procurement 5.05 Farm inputs would be procured by individual farmers through local trade channels. The equipment and vehicles required (in relatively small amounts) over the project execution period by the numerous agencies respon- sible for the different project components are locally produced and readily available. Such items (amounting to a total of roughly US$2.0 million without contingencies) would be procured in accordance with local procurement and bidding procedures, which are acceptable. Construction works (totalling US$15.5 million without contingencies) for the feeder roads would be divided into 19 construction lots (all executed during project years II-IV). The Bank would review the standard construction bidding and contract documents before the first round of bidding. Contracts for these works and for the  - 37 - civil works required for the pilot multi-purpose dams, storage units, extension facilities and rural primary schools (an additional US$5.1 million) would be awarded to prequalified bidders on the basis of competitive bidding advertised locally and in accordance with satisfactory procedures. Brazil has a competi- tive local construction industry capable of carrying out the project works. Since the individual construction works are relatively small and quite dis- persed, foreign contractors are not expected to be interested, though they would not be excluded from bidding. All of the other construction works included in the project (for health posts, water supply, mechanization service installations and the like, and totalling some US$2.6 million) would be of very minor size and varied design as well as geographically dispersed. Consequently, these would be carried out through a combination of self-help, force account and local bidding. Consultancy services, all of which are expected to be local, would include about 540 man-months at a total cost of approximately US$2.25 million equivalent (an average of about US$4,160 per man-month, with an expected range from US$3,000 to around US$5,000 depending on the type of consultancy). About two-thirds of the consultancy services would be for engineering on feeder roads, just under one-third for various technical assistance on project administration and evaluation, and the small remainder for technical assistance on the education and health components. Disbursements 5.06 The proposed loan, which would finance 35% of total project costs, would be disbursed against all project expenditures except land purchase credit. Bank disbursements would, therefore, be equivalent to 36% of expen- ditures for each component other than land purchase. Disbursements would be made to the Central Bank of Brazil, the fiscal agent for the POLONORDESTE program, against withdrawal applications covering statements of expenditures initiated by the various implementing agencies under the project and certi- fied by the Project Unit in CEPA-BA. Supporting documentation for credit, salaries, administrative expenses and minor construction under force account would not be submitted to the Bank but would be retained by the Project Unit and made available for inspection by the Bank during the course of project suprvision missions. Standard documentation covering civil works, vehi- cles and equipment and technical assistance would be submitted to the Bank. Details of the documents required should be reviewed during negotiations. Disbursements are expected to occur over about five years. The estimated disbursement schedule is given in Annex 4. Accounts and Auditing 5.07 Each of the participating agencies would maintain accounts of its project expenditures, which would be audited annually by Government auditors (Inspetoria Geral das Financas, IGF) according to standard Govern- ment practice for POLONORDESTE projects. Both Banco do Brasil and Banco do Nordeste do Brasil are audited by the Central Bank's auditing unit (Divisao de Auditoria da Contadoria Geral) and BNB is also audited by satisfactory independent auditors. The Project Management Unit would maintain accounts of its own project expenditures as well as statements of expenditures by each  - 38 - participating agency. Copies of the audited statements of project accounts of the various participating agencies and of the audited statements of the participating banks and of their lending under the project would be provided to the Bank through the Project Management Unit within six months of the end of their fiscal year. VI. PROJECT IMPLEMENTATION Organization and Management 6.01 General. The various components of the proposed project (see paras 4.08 - 4.49), with the exception of overall project administration, would be executed by the existing agencies traditionally responsible for the respective activity -- e.g., EMATER-BA for extension, the State Secretariat of Education for education, and so on. A detailed list of these agencies is given, by project component, in Annex 5. To coordinate and guide development activities under all its POLONORDESTE projects, the state of Bahia has already created a POLONORDESTE Management Council (Conselho Diretor, see para 3.01) composed of representatives of the principal executive agencies and headed by the State Secretary of Agriculture. The Management Council would therefore guide and coordinate all activities under the Paraguacu project, but day-to-day coordi- nation, administration and monitoring of the project would be undertaken by a Project Management Unit. 6.02 The Project Management Unit (PMU). A Project Management or Technical Unit (Unidade Tecnica) was established in 1976 within the State Agricultural Planning Commission (CEPA-BA, Comissao Estadual de Planejamento Agricola-Bahia) to complete preparation of the Paraguacu Basin rural development project. While administratively part of CEPA-BA, the unit is already housed separately and does not take part in CEPA-BA's other agricultural planning activities. The unit was restructured in 1977 to gear it mainly to coordination of project execution. Several further adjustments are planned and should make the unit well suited to this executive function. A separate evaluation group would be established for on-going and post-project evaluation (see para 6.13). 6.03 PMU staff would comprise: an executive coordinator, an average of eleven technical specialists, five part-time student interns (estagiarios) and an administrative support staff (administrative assistants, a budget clerk, typists, drivers) in its Salvador headquarters; and a "regional coordinator" of technical specialist standing and minimal support staff in each of three field offices. An organization chart of the PMU is provided in Annex 6. PMU staff would be slightly larger than that of other recently approved Bank- financed rural development projects (e.g., Rio Grande do Norte, Loan 1195-BR; and Ceara-Ibiapaba, Loan 1488-BR), reflecting the greater complexity and larger size of the Paraguacu project and project area. The proposed staff size, allocation of staff responsibilities and management organization struc- ture are intended to help avoid a situation sometimes encountered in projects of this type in which headquarters technical staff become overburdened with  - 39 - administrative tasks rather than being involved in more profitable trouble- shooting, decision-making and innovative functions. Moreover, experience of other projects suggests field staff play vital coordination and supervision roles. PMU staff, aside from administrative support and field staff, would be divided in two main sections: operations and monitoring/planning 6.04 PMU Operations Staff. Operations staff would coordinate implemen- tation activities of the numerous executing agencies. To achieve this, there would be four functional operational groups within the Project Management Unit: Agricultural Development (agricultural extension, agricultural research and seed production, and mechanization services and input supply), Production Support (on-farm and land purchase credit, land surveying and titling services, cooperative society support, and storage and marketing), Social Infrastructure (education and training, health and water supply) and Physical Infrastructure (roads, irrigation works and multi-purpose dams). Under the guidance of an operations coordinator, on average six operations staff would be assigned to the four operational groups as required by the complexity of implementation of the components. The staff of each group would assist the relevant executing agencies prepare detailed annual project operating plans (paras 6.10-6.11) and would act as PMU liaisons with the agencies in coordinating project activities and supervising the progress of each component. While Project Management Unit operations staff would be based in Salvador, each would be expected to spend at least 90 days annually in the field. Short-term implementation planning of each component would take place mainly within executing agencies with the exception of agricultural development, which would involve a relatively more active leadership and coordination role by the project unit operations staff (since CEPA-BA is itself the state's agricultural planning agency). 6.05 PMU Monitoring/Planning Staff. The section of the Project Manage- ment Unit dealing with monitoring, long-term planning and special studies would be headed by a monitoring and planning coordinator. The section would include initially three technical specialists (including the coordinator), five student interns, and administrative auxiliaries; other technical staff could later be transferred from the operations group. The monitoring sub-group would have two main tasks. First, it would monitor the output or results of each component and advise the project executive coordinator of short-falls in project achievements relative to expectations, particularly by comparing achievements to itemized objectives and indicators by component (see Annex 7). Standardized monitoring forms would be completed quarterly by executing agencies and forwarded to the monitoring group for synthesis into quarterly project reports that would be forwarded to project management, the state POLONORDESTE Management Council, Regional and Federal POLONORDESTE officials and the Bank. The monitoring forms would be pre-coded for computer proces- sing to accelerate and broaden access to project data. Once operating, monitoring by the project management unit would be chiefly a repetitive task of ensuring quarterly reports were received from executing agencies and of processing and analyzing returns. Agreement was reached during nego- tiations that: (i) the Project Management Unit would contract a consultant data analyst, under terms and conditions satisfactory to the Bank, to help standardize data requirements among participating agencies and establish  - 40 - coding and analysis procedures (with a view to extending monitoring formats, arnalysis and data retrieval to other POLONORDESTE projects) and provide a report on the recommended procedures to the Bank for review and comment prior to their implementation; and (ii) the quarterly project reports would be provided to the Bank promptly upon their completion. 6.06 The second main task of the monitoring sub-group would be to organize, administer, process, and initially analyze an annual farm pro- ductivity survey. The handling of this survey by the PMU rather than the evaluation group (para 6.13) should help isolate repetitive data collection and processing from the more qualitative and reflective approach required in evaluation. In addition, the farm survey operation should have a positive influence on project implementation by enhancing the PMU staff's awareness of realities in the project area, especially among the small farmer target groups, and of development issues in general. Project technical specialist staff (and student interns) would be involved in survey data collection for about three weeks annually. This involvement would ensure their intensive contact with project beneficiaries, an experience few might otherwise gain. The contact would provide direct feedback between beneficiary and planner and, hopefully, a better sense of the potential uses as well as shortcomings of the survey-derived data. It is envisaged that the farm survey would comprise a small sample (probably less than 700) of participant and non-participant farmers in the project area. It would be conducted with a fixed format, pre- coded questionnaire centered on production technology, farm inputs, institu- tional contact and support, and produce marketing. If greatest possible value is to be derived from the survey program as a management tool and as a basis for evaluation and planning, the first year the survey will surely be the most critical: the questionnaire would need to be created, tested, and pre-coded; farmers approached, their cooperation sought and the sample selected; enumerators trained; and, the first year's survey conducted, processed and analyzed. The PMU would require initial assistance in this work. It was therefore agreed during negotiations that the PMU would contract consultants on terms and conditions satisfactory to the Bank to assist in survey prepara- tion and to establish data processing and analysis procedures, the recommended survey design and analysis procedures to be communicated to the Bank, for review and comment prior to implementation. The consultant's terms of reference would specifically stress that data collection should be tailored to data analysis capacity and that reports should be prepared in a short enough time to be of full operational use to the PMU. 6.07 The long-term planning and special studies sub-group would be responsible for: (i) identifying, organizing and, usually, contracting out studies of particular problems affecting project implementation that are too far-ranging or time-consuming for operations staff to investigate; and (ii) examining, on an annual work-plan schedule, long-term plan- ning issues affecting project implementation. Examples of the type of issues that would be examined are: manpower status and projections; agro- industrial feasibility studies, including processing and marketing possi- bilities for small livestock products, such as sheep and goat skins, meat and cheese; ecologically, technologically and economically viable crop  - 41 - combinations; marketing channels and their cost efficiency; production, market and price projections for project area crops; and intra-basin variation in output and input marketing, including variation in prices, margins, channels, seasonality and volume. 6.08 Field Offices. Because of the very large size of the project area and the need to assure good field supervision and coordination of the project, small field offices have been established at Seabra, Itaberaba and Feira de Santana. The chief task of the regional coordinator in each will be to provide administrative continuity and a communication link between the PMU, executing agencies and beneficiaries to ensure speed and efficiency in project execution. Local and regional project committees, with representatives of these groups meeting at least monthly, are already being formed to help achieve this objective. As well as resolving implementation difficulties, the role of the regional coordinator as a two-way communication channel should provide beneficiaries with an impression of where they fit into the broader regional development plans, help ensure their cooperation, and provide grass- roots feedback to project administration and executing agencies on implemen- tation progress and problems. 6.09 The present administrative organization of the POLONORDESTE pro- gram in Bahia is likely to remain in effect over at least the medium term. Assurances were received from the state during negotiations that it would maintain, for the entire duration of the project, the PMU with appropriate staff and resources, and would fully support other state agencies participating in the project. Operating Agreements and Procedures 6.10 Since the overall project covers a wide number of related activi- ties, and the timing and scope of some will undoubtedly require adjustment as experience is gained, a certain amount of flexibility in project execution would need to be maintained. For this reason, and to help assure continued timely budgetary support, each participating agency would prepare jointly with the PMU an annual plan detailing the activities and expected expenditures under the project for the next year. The yearly planning period would corres- pond to POLONORDESTE's April 1 - March 31 fiscal year. The integrated package of annual plans, as reviewed and approved by the state POLONORDESTE Management Council would be submitted to the regional and federal POLONORDESTE coordin- ating groups as a basis for yearly budget allocations. Agreement was reached during negotiations that the annual plans -- which provide the critical operational detail needed to implement and, as necessary, adjust the initially defined longer term project actions -- would be submitted to the Bank by November 30 each year for review and comment. Agreement was also reached that the annual plans would receive the prompt review of the regional POLONORDESTE coordinating group (so as to avoid delays in budget approvals) and that the Bank would be advised of the approved plan by January 31 each year. 6.11 Consistent with a commonly used system in Brazil, the state would enter into written agreements (convenios) with certain of the executing agencies to specify their respective responsibilities under the project and  - 42 - help assure timely project completion. Such agreements are normally utilized in POLONORDESTE projects when: (i) the executing entity concerned is not directly part of or responsible to the Ministry of Agriculture or the State Government, the principal channels for POLONORDESTE funding other than credit; (ii) the executing agency is largely autonomous; or (iii) special responsi- bilities (e.g., counterpart funding, adoption of new procedures) need to be spelled out. In the proposed project, convenios would be particularly useful in assuring adequate participation of, and coordination among, the numerous entities participating in the health, education, storage-marketing, mechani- zation service and input supply components (all of which require special counter-part actions by bodies such as local municipalities or other state of federal institutions). Draft convenios required for the execution of these components have been reviewed by Bank staff and found satisfactory. During negotiations, assurances were obtained from the state that it would enter into and maintain during the project execution period all such agreements as are necessary to assure active cooperation of the various agencies participating in the above-mentioned components. Receipt by the Bank of the respective signed agreements would be a condition of disbursement for each of these components. Implementation Schedule 6.12 A chronogram of the main project activities is given in Annex 8. Evaluation 6.13 Although directly or indirectly, all of the staff working on different aspects of the project would need to be involved in evaluating project impact and effectiveness, overall responsibility for project evaluation would lie with an evaluation group separate from the rest of the PMU, but funded by the project and located in the Secretariat of Agriculture (possibly within CEPA-BA). It would be directly responsible to the Management Council and would carry out on-going and post facto evaluation of the project. Staff of the evaluation group would comprise two professionals (one of whom would be a social scientist) and administrative support. The group would have an orientation toward the sociological, as well as economic, aspects of rural development to help focus evaluation on the intended key project beneficiaries. 6.14 On-going evaluation would complement the monitoring activities of the PMU and would require close collaboration between the evaluation group and the rest of the PHU in the planning of work programs and sharing of information and ideas. It would involve a continuous advising of project management of the extent and quality of project progress and a rapid isolating of present or potential implementation difficulties. On-going evaluation "products" would comprise reports of topic-specific evaluation studies (see below), evaluations of PMU monitoring and farm survey results, and a major mid-project evaluation report (during project year 3). On-going evaluation studies would be conducted primarily by consultants on a study-by-study basis but consideration would be given to the possibility of the long-term involvement of a Brazilian university graduate program in evaluation studies. Studies would be qualitative as much as quantitative, covering topics suggested by the evaluation group, project  - 43 - management, executing agencies, the Management Council, POLONORDESTE officials or the Bank. The studies would provide an independent examination of possible difficulties in project implementation and would be a source of suggestions for remedial action, adaptation or expansion. They would focus on the quality of project interventions, and the extent to which project objectives are achieved and to which the various segments of the target population are benefitted. Specific evaluation studies could include: (a) evaluations of the effectiveness of particular executing agencies or development strategies (e.g., credit delivery, land titling, off-farm storage and marketing activities, use of informal produ- cer groups as extension "contact groups", participation of small- scale farmers in cooperatives) and the productivity and income changes resulting from the project; (b) participant evaluation of project activities -- comparisons of the expectations of participants (beneficiaries and executing agencies) to actual project intentions and attainments; and (c) intensive sociological studies of a small number of participant and non-participant communities, to serve as a counterpart to the farm survey in establishing a data base for project evaluation. The evaluation group would contract a consultant, on terms and conditions satisfactory to the Bank, to assist in establishing the project evaluation strategy and work program. Agreement was reached during negotiations that the detailed evaluation plans would be provided to the Bank by December 31, 1978 for comment. 6.15 The evaluation group would also prepare a post facto project evalua- tion report to serve partly as a project completion report. It is expected that both the on-going and post facto evaluation would provide the state with an important input in the formulation of future rural development projects. Agreement was reached during negotiations that the project completion and post facto evaluation report would be provided to the Bank within six months of project completion and that the report would include information on the execution and initial operation of the project, including costs and estimates of the effects and impact of the project, as well as information on the performance of all involved parties. VII. PROJECT PRODUCTION, DEMAND AND MARKETING Production 7.01 By full development (about year 8), there would be substantial increases in crop production among project participants deriving from: a moderate expansion in the area cropped (a 58% increase, to around 60,000 ha, including some 32,000 ha interplanted); the introduction of improved manage- ment, technology and inputs; and off-farm investments, particularly in roads,  - 44 - storage facilities, agricultural extension and research and seed production services. At estimated 1977 farm-gate prices (taken as 1974-76 average farm-gate prices in constant terms), incremental annual production is expected at full development to amount to about Cr$491.1 million, or US$32.7 million equivalent. Value of production per hectare is estimated to increase about 75% to Cr$7,290 (US$486 equivalent) at full development. 7.02 Approximately 45% of incremental production (by value) would come from beans, 11% from manioc, 8% from citrus, 9% from miscellaneous other fruit, 11% from maize and 10% from string and leaf tobacco. The relative importance of the different crops to project participants would change with project production. In addition to the small-scale introduction of peanuts and soy beans, the production of citrus and other sub-tropical fruits would be considerably augmented. Manioc would remain a key crop, but would account for under one-half its pre-project share of participant production. Maize, garlic, tobacco leaf and castor beans would decline, while beans, fruit and tomatoes would increase in importance. Because of the "demonstration effect" of project extension activities and infrastructural development, production increases are also probable from many other farms in the project area. The degree of such increased production is not easily predictable. 7.03 Production of the estimated 675 livestock beneficiaries would increase through improved management and production technology. Incremental meat production is expected to amount to around 800 tons annually at a value of about Cr$9.6 million (US$0.64 million equivalent). Estimated incremental milk production (about 2.88 million liters) would have a value of around Cr$5.8 million (US$0.39 million) annually. Demand, Markets and Prices 7.04 Potential project participants account for a small share of current (1977, estimated) state production, although project activities could raise their contribution significantly. The Paraguacu Basin could become an important source in Bahia of beans, garlic and onions, and possibly also of maize and tomatoes. Without project and incremental production of potential participants as a percentage of estimated 1977 state production is as follows:  - 45 - Potential Participant Prod. as % Estimated 1977 Bahia Prod. Crop Bahia production 1/ Without Project Incremental ('000 tons) Manioc (root) 5,547.0 1.8 2.4 Beans 140.0 5.4 20.6 Maize 265.0 3.1 12.9 Rice 38.5 0.3 2.9 Tomatoes 42.4 1.3 8.2 Onions 10.0 2.8 14.2 Garlic 1.4 11.9 36.5 Castor beans 154.0 1.3 1.8 Tobacco (leaf) 34.0 4.7 6.5 Participant production of milk and meat at full project development would amount to little more than 1% of current Bahian production of each. 7.05 While project activities could increase significantly the produc- tion of target crops, population and urbanization trends, available data on income elasticities of demand in potential market areas and on production trends in other potentially competitive production areas, suggest that project production could face demand constraints or price reductions to the farmer (see also para 7.06), were it not for planned project investments in access roads and improvements in market services. Chief markets for participant food production will be the population of the project area and the adjacent greater Salvador area. Project area population is projected to be 1.4 million and 43% urban by 1990, compared to slightly more than 1.0 million and 35% urban presently. Through 1990, Salvador is expected to expand 4.5% annually from a current population of 1.4 million, so continuing to have a population greater than the entire project area. For demand calculations it was assumed that the commercial trade of food products would first be towards cities in the project area with over 5,000 inhabitants and that any balance would go to the Salvador market. Upon these assumptions, maize and rice production is unlikely to leave the project area. Manioc (marketed only as flour) 2/, fruits, onions, garlic and tomatoes will find ready markets in the project area and the balance 1/ Data are not available on current state production of string tobacco, fruit, soy beans and peanuts. Estimated state production data, particularly of minor crops (rice, tomatoes, onions, garlic, tobacco), are very approximate. 2/ For some time, there have been discussions in Bahia (as elsewhere in Brazil) about the possibility of developing an industrial capacity to process alcohol from manioc with the intention of using the alcohol as a gasoline additive. Proposals made so far appear to have doubtful or marginal economic viability. If viable plans were developed in the future, the market for project area manioc may be considerably better than now assumed.  - 46 - in Salvador. No marketing difficulties are foreseen for participants' indus- trial crops. String tobacco will be marketed to the existing specialized markets in Bahia, Minas Gerais, Goias and Mato Grosso; leaf tobacco will continue to be sold to national manufacturers and to overseas markets. The slight incremental producttion of castor beans will be marketed through existing processing channels. Meat production will find a strong demand, chiefly in Salvador, and milk production will find a ready market at the chief dairies of the region (in Salvador and Feira de Santana). 7.06 It is difficult to estimate changes in the prices of agricultural and livestock products that might result from project activities. As incre- mental production of most crops is slight compared to state production, pro- ject production is unlikely to have a negative influence on prices. Other factors would affect prices, however. The greater utilization of storage by producers might help decrease marginally local seasonal price variations, although it would also allow producers to take advantage of whatever variation remained. Regional differences in prices within the project area should decrease with project road activities and consequent developments in transport, just as greater regional connectivity in Brazil would contribute to bringing Bahia prices more in line with national prices (which are, in most cases, somewhat lower). In sum, improvements in local and regional connectivity will result in two opposing forces that could affect project area prices: greater connectivity within the project area could increase prices received by producers, while improved national connectivity could result in prices at the state level (and eventually within the project area) declining relative to national prices. It has been assumed that in the medium term the prices for project area agricultural and livestock products will decrease a little in relation to prices received elsewhere in Brazil, but that the producer's share of the final price will increase slightly as a result of more efficient and competi- tive marketing and transportation systems and better farmer organization. Marketing 7.07 No major changes in the present marketing system are foreseen for the marketing of project production as it is believed that the marketing system should be compatible with the volume and location of production, farming systems and socioeconomic environment of project beneficiaries (see para 7.09). However, during the project period, marketing options open to producers are expected to expand with the state's efforts to promote minimum price and advance-purchase marketing schemes (see para 7.10), the provision of off-farm storage facilities, the planned strengthening of cooperative societies and other developments in physical infrastructure (particularly road construction and upgrading) under the project. 7.08 Marketing of agricultural production in the project area is now characterized by minimal involvement of government or other large institu- tions; producers must find their own markets. Part of food production is consumed on-farm, some is kept for seed and planting stock, and the balance is sold to consumers or intermediaries at local public wholesale/retail  41 markets. in.uostr4`. ki;ct: d meat are generally sold on-farm or in market towns to agents of processors, and milk is collected by truckers for transpo-<. o-c T_sere is practically no on-farm processing of produce, as fre from converting manioc root to flour. There is limited post- harvest stora c-t of_oduce .y small producers; however, hardly any use the few public off-.iarm s e facili ties in the region. Producers generally receive 50-60% of the s- ales cpnne on less perishable items (rice, beans and maize) and 30O407 cnc r4ore perishable products. Milk producers receive 70-75% of the retail t rie -ca. 9etcii margins are about double wholesale margins. 7.09 Majo.-:c anges 1a the marketing system are not advocated at this time for five cihief reasons: (i) many target participating communities are isolated and distart fErom significant markets; (ii) the existing marketing system is not necessari-;y Inefficient given the isolation of some partici- pating cotecitie 3 te atomized nature and small-scale of production, and the extra-market-ing functions (credit and input supply, for instance) pre- sently perforocd by intermediaries; (iii) demand co-astraints for project productiorn are nolt foreseen (iv) there does not exist an adequate adminis- trative structure r which a more comprehensive marketing system could be readily devel-cas a-dA tv it is not yet certain at what rate participant producers. to s.eeU- cut new off-farm storage/marketing facilities as a result of te `.naging szales and modes of production. 7.10 Whi e ex st4n marketing arrangements should be able to handle incremental prouc.ution, two project-related innovations could broaden market- ing opticns. ',ne is che planned testing of an advance purchase scheme in part of the project- a-ea by the Brazilian Food Company (COBAL). The scheme would provide a guar2nteed p.a-harvest market for maize, beans, and, possibly, rice at least at the officia-11Y quoted (but, to date, rarely available) minimum price. 1/ Tlhe second innovation would be the implementatLon of the Production Financing Commission's (CP) minimum price loan/purchase option. This would require the implementat'on of existing regulations by CFP and its agent, BB -- and producers being Better organized and made aware of, and encouraged to use, the option - ts th,an the creation of new regulat-ions. The project itself will not finance the operation of these schemes, but to ensure the necessary storage infrastructure is avaLlable, the project would inelude the construction of six 700-> ca y waret