Brazil - Agricultural Extension Project

Docum-ent of The World Bank FOR OFFICIAL USE ONLY Report No. 1879-BR STAFF APPRAISAL REPORT BRAZIL AGRICULTURAL EXTENSION I PROJECT May 3, 1978 Projects Department Latin America and the 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 EQUIVALENTS Currency Unit = Brazilian Cruzeiro Cr$ Cr$ 15.435 = US$1.00 a/ Cr$ 1,543.50 = US$100.00 Cr$ 15,435.00 = US$1,000.00 Cr$ 1 million = US$64,785.82 WEIGHTS AND MEASURES (Metric System) 1 Kilogram (kg) = 2.20 pounds 1,000 kg = 1 metric ton = 0.98 long ton 1 meter (m) = 1.09 yards 3.28 feet 1 kilometer (km) 2 = 0.62 mile 1 hectare (ha) 2 = 10,000 m = 2.47 acres 1 square kilometSr (km ) = 100 ha = 0.39 square mile 1 cubic meter (m ) = 25.39 cubic feet GLOSSARY OF ABBREVIATIONS (see pp i - ii) EMPRESA BRASILEIRA DE ASSISTENCIA TECNICA E EXTENSAO RURAL (EMBRATER) January 1 - December 31 a/ Selling rate November 30, 1977  FUR OFFICIAL USE ONLY GLOSSARY OF ABBREVIATIONS ABCAR - Associa,ao Brasileira de Credito e Assistencia Rural (Brazilian Association for Credit and Rural AssistancB) ACAR - Associa,ao de Credito e Assistencia Rural (Credit and Rural Extension Agency) ASTER Associa,co do Asistencia Tecnica e Extensao Rural (Territorio) (Agency for Technical Assistance and Rural Extension (Territory) BB - Banco do Brasil (Federal Bank) BID - Banco Interamericano de Desenvolvimento (IDB - Inter-American Development Bank) BNB - Banco do Nordeste do Brasil,S.A. (Bank of Northeastern Brazil) BNCC - Banco Nacional de Credito Cooperativo (National Cooperative Credit Bank) CATI - Coordenadora de Assistencia Tecnica Integra&d (Office of Coordination of Integral Technical Assistance) CEPLAC - Comissao Executiva de Plano da Lavoura Cacaueira (Commission for Recuperation of Cacao Culture in Rural Areas) CFP - Comissao de Financiamento da Producao (Production Financing Commission) * CIBRAZEM - Campanhia Brasileira de Armazenamento (Brazilian Storage Company) CNAE - Campanhia Nacional de Alimentacao para Escola (National School Food Campaign COBAL - Companhia Brasileira de Alimentos (Brazilian Food Company) COMPATER - Comissao Nacional de Pesquisa Agropecuaria, Assistencia Tecnica e Extensao Rural (National Commission for Agricultural Research, Technical Assistance and Rural Extension) CONAB - Conselho Nacional de Abastecimento (National Food Supply Council) CONCRED - Comissao Nacional de Credito Rural (National Rural Credit Commission) CONDEPE - Conselho Nacional de Desenvolvimento da Pecuaria (Council for the Development of Livestock) DIPOA - Departamento de Inspecao de Produtos de Origen Animal (Department of Inspection of Products from Animal Origin) DNOCS - Departmento Nacional de Obras Contra as Secas (National Department for Anti-drought Operations) DNSC - Departamento Nacional de Servicos de Comercializacao (National Department of Marketing Services) EMATER - Empresa de Assistencia Tecnica e Extensao Rural (State Technical Assistance and Rural Extension Corporation) EMBRAPA - Empresa Brasileira de Pesquisa Agropecuaria (Brazilian Agricultural Research Corporation) ENBRATER - Empresa Brasileira de Assist^ncia Tecnica e Extensao Rural (Brazilian Technical Assistance and Rural Extension Corporation) FUNRURAL - Fundo de Assistencia ao Trabalhador Rural (Fund for Welfare Assistance to Rural Workers) IBC - Instituto Brasileiro do Cafe (Brazilian Coffee Institute) ICM - Imposto sobre Circulasao de Mercadorias (Circulation of Merchandise Tax) 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 autt orilation.  - ii. - INAN - Instituto Nacional de Alimentagao e Nutri,ao (National Institute for Food and Nutrition) INCRA - instituto Nacional de Colonizacao e Reforma Agraria (National Institute for Colonization and Agrarian Reform) MA - Ministerio da Agricultura (Ministry of Agriculture) MEC - Ministerio da Educa,ao e Cultura (Ministry of Education and Culture) PIASS - Programa de Interioriza,ao das Acoes de Saude e Saneamento (Rural Health Program) PIN - Programa para Integraqao Nacional (Program for National Integration) -POLAMAZONIA - Programa de Polos Agropecuarios e Agrominerais da Amazonia (Program for Agricultural and Agro-mineral Nuclei in Amazonia) POLOCENTRO - Programa de Polos Agropecuarios da Area Central (Program for the Development of the Central-West) POLONORDESTE - Programa de Desenvolvimento das Xreas Integradas do Nordeste (Development Program for Integrated Areas of the Northeast) PROTERRA - Programa de Redistribuigao de Terras e de Estimulos a Agro-industria do Norte e Nordeste (Program for Land Redistribution and Agricultural Incentives in the North and Northeast) SEPLAN - Secretaria de Planejamento (Federal Planning Office) SIBRATER - Sistema Brasileiro de ATER (Brazilian System for Rural Extension - Federal level) SNIR - Sistemn Nacional de Informaeao Rural (National System for Rural Information Services) SUDAM - Superintendencia para o Desenvolvimento da Amazonia) (Superintendency for the Development of Amazonia) SUDECO - Superintendencia para o Desenvolvimento do Centro-Oeste (Superintendency for the Development of the Central-West) SUDECOOP - Superintendencia de Cooperativismo (Superintendency of Cooperatives) SUDENE - Superintendencia para Desenvolvimento do Nordeste (Superintendency for the Development of the Northeast)  BRAZIL AGRICULTURAL EXTENSION I PROJECT STAFF APPRAISAL REPORT TABLE OF CONTENTS Page No. I. THE AGRICULTURAL SECTOR ............................ I A. The Sector in the Economy .... .............. I B. Land Distribution and Use .... .................. 2 C. Agricultural Development Policy ................ 4 D. Agricultural Support Services .................. 5 E. Bank's Participation in the Sector ............. 8 II. AGRICULTURAL EXTENSION ............................ 10 A. Historical Perspective ......................... 10 B. Present Status of Agricultural Extension ....... 11 C. Interactions with Other Agricultural Services .. 15 III. THE PROJECT ........................................ 18 A. Introduction ................. .. ................ 18 B. Description, Objectives and Scope .............. 18 C. Detailed Features .............. .. .............. 20 D. Project Costs ................ .. ................ 21 E. Financing .................... .................. 23 F. Procurement ................. .. ................. 24 G. Disbursement ................. .. ................ 24 H. Accounts and Audit , .. 25 IV. PROJECT IMPLEMENTATION ............................. 26 A. Organization and Management .... ................ 26 B. Field Implementation and Extension Methodology . 27 C. Staff Training and Technical Assistance ........ 33 D. Monitoring and Evaluation ...................... 34 V. PRODUCTION AND TECHNOLOGY .......................... 37 The project is based on the findings of an appraisal mission which visited Brazil during June-July 1977. The mission comprised Messrs. R.P. Armour (Mission Leader), G. Ashkenazi, K. Kwaku and C. Vidalon (Bank), and A. Apodaca, G. Hull, A. Ramirez and G. Kawata (Consultants).  TABLE OF CONTENTS (Continued) Page No. VI. DEMAND AND MARKET PROSPECTS ............................ 40 VII. BENEFITS AND JUSTIFICATION ............................ 41 A. Beneficiaries and Target Population Groups ........ 41 B. Cost Recovery ...... ............................... 41 C. Economic Analysis ..... ............................ 42 D. Social Benefits .................... 43 E. Project Risks ...... ............................... 43 VIII. SUMMARY OF AGREEMENTS REACHED AND RECOMMENDATION ...... 44  TABLE OF CONTENTS (Continued) ANNEXES Page No. 1. Monitoring and Evaluation .................................... 46 2. Economic and Social Benefits ................................. 49 List of Tables 1. Target Population Basic Data ............................. 54 2. Incremental Production Attributable to Project .... ....... 55 3. Project Base Costs ....................................... 56 4. Project Costs and Incremental Value of Production ........ 57 3. Supporting Tables and Charts T-1 Area, Population and Demographic Density .... ........... 58 T-2 Land Distribution and Ownership ........................ 59 T-3 Gross Domestic Product and Agricultural Product ........ 60 T-4 Area Yield and Total Annual Production for Basic Food and Export Crops ...................... 61 T-5 Performance of Ten Major Export Crops .... .............. 62 T-6 Distribution of Agricultural Properties by Size ........ 63 T-7 Allocation of Land to Crops and Livestock .... .......... 64 T-8 Regional Distribution of Extension Units (1977) ........ 65 T-9 Rural Population and Number of Farmers Receiving Extension Service Assistance (1977) ............................ 66 T-10 Number, Level and Distribution of Technical Staff ...... 67 T-11 Incremental Staff Appointments by Program Emphasis ..... 68 T-12 Incremental Staff Appointments by Category .... ......... 69 T-13 Incremental Staff Appointments by Regional Emphasis .... 70 T-14 Incremental Staff Appointments by Project Year ......... 71 T-15 Cost Estimates per Program by Project Year .... ......... 72 T-16 Regional Distribution of Project Costs .... ............. 73 T-17 Proposed Financing Plan and Allocation of Loan Proceeds 74 T-18 Estimated Schedule of Disbursements .... ................ 75 T-19 Technical Assistance - Consultants - Manyears .... ...... 76 T-20 Physical Targets for Project Development Programs ...... 77 C-1 Organization Chart: EMBRATER/EMATER Structures and Function ......................................... 78 C-2 Organization Chart: EMATER - Structure Common to all Agencies ...................................... 79 C-3 Brazil's Agricultural Policy - Summary Chart .... ....... 80 C-4 Civil Works: Construction Schedule .... ................ 81 C-5 Implementation Schedule by Project Years .... ........... 82 C-6 Summary of Program Objectives and Implementation ....... 83 4. Selected Documents and Data Available in Project File ........ 84 MAPS BRAZIL: Major Regions and Location of Training Centers (IBRD No. 13120) BRAZIL: Physical Features (IBRD No. 13416)   BRAZIL AGRICULTURAL EXTENSION I PROJECT L. THE AGRICULTURAL SECTOR A. The Sector in the Economy Geography 1.01 Brazil, which occupies almost 2half of the South American continent has a total land area of 8.5 million km (Annex 3, Table 1). Although geographic, topographic and climatic features are favorable to plant and animal production in over 85% of the country, the majority of this area (60.5%) is in native forest and natural woodland (Annex 3, Table 2). Only 4.3% is in permanent cropping and a further 19.8% in improved or permanent pasture. The Brazilian topography is of pirticular significance from an agricultural stand- point. About 4.9 million km or 58% of the total land area, is at elevations of less than 300 meters and much of it is relatively flat or rolling terrain conducive to mecha2ized agriculture. In the range of 300 to 599 meters, there are 2.3 million km or 27.4%. Finally, the area of highest elevations (600- 2,890 meters), represents only 14.6% of total area, but here also, at levels of approximately 800 to 1,200 meters, there are large areas of flat table lands suitable for agricultural production. It is evident, therefore, that in fundamental terms the country has great agricultural potential and the development that has already occurred, though substantial, is merely the fore- runner of much greater development in the future. The factors most limiting to agricultural expansion are poorly developed support services, poor infra- structure where Brazil's sheer size plays a dominant role, and the general low natural fertility of many of the soils, particularly in the underdeveloped regions of the Central-West and North (see Map Nos. IBRD 13120 and 13416). Gross National Product 1.02 Brazil's economy has traditionally maintained a strong agricultural base. The relative importance of agriculture (Annex 3, Table 3) has shown a gradual decline in overall contribution to GNP during the past decade (from 13.3% in 1966 to 10.5% in 1975) but agriculture still provides over 60% by value of merchandise exports. Recent significant increases in the value of agricultural export crops, particularly for coffee, cacao and soybeans, have now moved Brazil into third place among the world's exporting countries of agricultural produce. The relative growth of Brazil's agriculture during the period 1968-76 is shown overleeaf:  - 2 - ANNUAL GDP SECTORIAL GROWTH, 1968-76 (percent) Average 1968-73 1974 1975 1976 Agriculture 5.3 8.5 3.4 4.2 Industry 13.0 8.2 4.2 10.9 Manufacturing (13.6) (7.6) (3.8) (10.5) Civil construction (10.5) (11.2) (3.8) (12.8) Public utilities (11.9) (12.0) (10.2) (10.1) Services 10.3 10.8 4.4 8.8 GDP 10.2 9.6 4.2 8.8 Source: Getulio Vargas Foundation (FGV), October 15, 1977. 1.03 The growth in crop value for the period 1973-75, has come mainly from coffee, soybeans, cacao, maize, tobacco and citrus. The incremental value for soybeans alone was a significant 81%. Brazilian agriculture is particularly sensitive to price incentives for annual crops, such as cotton, soybeans, tobacco and wheat, and annual changes in the area under cultivation reflect this price sensitivity at the farm level (Annex 3, Table 4). Agricultural Exports and Imports 1.04 Ten crops provide the bulk of agricultural exports: namely, coffee, sugar, soybeans, cacao, maize, castor bean (oil), tobacco, cotton, oranges and sisal. All are sensitive to world price fluctuations, particularly the annual crops (Annex 3, Table 5). 1.05 Agricultural imports are not significant except for wheat and dried beans. Wheat imports fluctuated significantly, for 1976 over 1975, by an increase of 57%, based on dollar value. Brazil currently fluctuates between being a net exporter or net importer of both rice and edible beans. Two other products which currently have a significant role in the Government's economic strategy, to reduce costly imports, are natural rubber and petroleum products. It is hoped to stimulate domestic production of the former, while a program to substitute with alcohol, up to 10% of the gasoline used in automobiles, is now seriously being considered under a plan to produce methanol from sugarcane, cassava and sweet sorghum. Importance of agriculture in the Labor Force 1.06 Despite its gradually decreasing importance relative to GNP, there were still 12.8 million people engaged in agriculture during 1976. This represented 39.5% of the total active labor force and may be compared with 19.8% for industry, and 40.7% for services.  - 3 - B. Land Distribution and Use Farm Size and Ownership Pattern 1.07 Figures for the 1970 cadastral survey show a total of 4,923,790 holdings in Brazil covering 294,144 thousand ha, or 34.8% of the total land area. The distribution by size of holdings was shown to be as follows: Number of Number of Area in Farm Size Holdings Total Area Holdings Each Group (ha) (Units) (ha .000) (% (%) 0 to 10 2,519,630 9,083 51.40 3.06 10 to 100 1,934,392 60,069 39.42 20.41 100 to 1,000 414,746 108,743 8.44 37.07 1,000 to 10,000 35,425 80,059 0.71 27.21 >10,000 1,597 36,190 0.03 12.25 Unclassified 18,000 - - Totals 4,923,790 294,144 100.00 100.00 From the table above it will be seen that, using composite figures, 91% of all holdings are in the 1-100 ha range but control only 23.5% of the land area, while holdings of 100 ha and over represent only 9% of the total number of units but occupy 76.5% of the land area. 1.08 Of additional interest is the distribution of small land holdings by regions. In the 1-10 ha range the majority, 39.48%, are in the South, closely followed by 34.99% in the Northeast. The same pattern (41.87% and 26.98% respectively) is true for the 10 - 100 ha range. In the 10 - 100 ha range, the percentage of holdings within the area of project emphasis, regions north, northeast and central-west is 36% (Annex 3, Table 6, Section I). There is also wide variation in the relative percentages, based on property size within regions, as follows: in the 0 to 100 ha range; north 63%, north- east 80% and central-west 49%. Only in the north and central-west is there a significant number of very large holdings, i.e. in excess of 10,000 ha (see also Annex 3, Table 6, Section II for detail). 1.09 Under owner/occupier classification farm units are divided as follows: direct ownership 64% and renter/sharecropper 36%. These statistics are of considerable importance when considering government policy for providing a meaningful stimulus to agricultural development through availability and/or direct provision of necessary inputs such as credit, fertilizer and support services such as research and extension. The successful transfer of technology, at the farm level by extension methods, is particularly influenced by the size and relative efficiency of the production units being serviced.  -4- Land Use 1.10 The total area in crops is currently thought to be 37.1 million ha, or 4.4% of - -al land area. Of this area, 9.4 million (25%) is in perennial or permanen :-e crops, 27.3 million (74%) in annual field crops, and 361,000 (1%) in horLicuitural crops. The area used for livestock production (155.5 millioti h; uo 1{8.4% of land area) is divided between 111.1 million ha (72%) in permanent pasture and 44.3 million ha (28%) in temporary pasture. In addition, some 27.2 million ha of forest are utilized for extractive activities, i.e. gathering of natural r^- 3ber, nuts and essential vegetable oils (see also Annex 3, Table 7). C. Agricultural Development Policy 1.11 In the 1970's there have been two National Development Plans (PNDs), the first, PND I, from 1972-74 and PND II from 1975-79. Under both plans agri- cultural strategies have been very similar and an annual (GNP) growth of 7% was predicted for the sector. This rate of growth was achieved in 1974 (para. 1.02) but for the remainder of the period growth rates have been between 3.4 and 5.3%. Preliminary estimates for 1977, however, would indicate a dramatic improvement to 9.8% due largely to increased values for coffee, cacao and soybean exports. In both development plans there has been a notable policy emphasis towards increasing the area under production as well as stimulating increases per unit area by a series of fiscal incentives to encourage the production of key domestic food and export crops. The major features and goals of such development policies affecting agriculture are as follows: (a) Land Settlement and Rural Development Programs: Examples are the defining of regional development "poles," chiefly in the North and Central-West and development plans for these areas (POLOAMAZONIA and POLOCENTRO). In the Northeast, special rural development schemes have come under the POLONORDESTE plan, linked to such programs as the Program for National Integration (PIN) and the Program for Land Settlement and Stimulus to Agro-Industry in the North and Northeast (PROTERRA); (b) Specific Crop and Livestock Production Programs: These have been based on specific objectives such as the formation of "export corridors," with the crops chosen being key elements in the drive for increased exports of agricultural products. Examples are programs for pasture and herd development (PROPEC and PRONAP) and, for crops, minimum pricing policies for selected crops under the Production Financing Commission (CFP). (c) Production Credit, Food Distribution, Storage and Nutrition Programs: Many of these, particularly credit and nutrition programs, are designed to assist the low-income farmer. The regula--ization of production incentives comes under the National Food Supply Council (CONAB) food storage is controlled by the Brazilian Storage Company (CIBRAZEM) and food distribution is the responsibility of the National Food Supply Superintendency (SUNAB).  - 5- (d) Institutional Strengthening. In order to provide the necessary new technology, and ensure that new productior. ,echniques thus developed also reach the farmer, the Govern- ment has recognized the need for integration and administrative strengthen- ing of the agricultural research and extension services at all levels. As a result, steps are being taken to improve these key agricultural support services as described in paras 1.13 and 1.14. D. Agricultural Support Services Introduction 1.12 Considering the size of the country, the importance of agriculture to the national economy and the enormous potential that remains for further development, agricultural support services are not particularly well developed apart from the banking system. Big increases in production have been achieved in recent years but these have been due largely to opening up new land, and the stimulus of rising commodity prices rather than through the impact of conventional support services such as extension and research. Government is, however, increasingly aware of the need to improve such services in order to achieve increased yields per unit area thus improving levels of efficiency in the agricultural sector. Agricultural Research 1.13 The Brazilian Corporation for Agricultural Research (EMIBRAPA) is the institution now responsible for the coordination and promotion of research throughout Brazil. It was founded in 1972, through the amalgamation of a number of different research organizations at both federal and state levels. Through fiscal control at the federal level it can either act directly to execute research at its own research centers, or act indirectly through numerous state level research institutions. It also acts as the planning, coordinating, executing and evaluating agency for federally integrated devel- opment programs where agricultural research has a significant role to play. EMBRAPA's long range plans, initiated in 1973, now call for expenditure of over US$100.0 million per year for the 1977-80 period. The new production technology being developed must, however, be successfully applied at the farm level before significant returns on this investment can accrue. It is here that the role of the extension service can have a significant impact in the Government's development plans for agriculture in the years ahead. 1/ The present proposal for an Agricultural Extension I Project, to cover the operational period 1978-81 is designed to ensure that the results of agricul- tural research are made available, through transfer of technology, to all producers but with particular emphasis on the small farmer where the need is greatest. 1/ See also Appraisal of the Agricultural Research I Project, Report No. 1031a-BR, April 15, 1976.  -6- Agricultural Extension 1.14 The Brazilian Corporation for Technical Assistance and Rural Extension (EMBRATER) is the agency currently responsible for coordinating agricultural extension work in Brazil. Like EMBRAPA it is a semi-autonomous agency under the Ministry of Agriculture. However, it was formed three years after EMBRAPA and is still only in the initial stages of taking full control over its mandate at the federal level. EMBRATER's primary function is to formulate, coordinate, monitor and evaluate the implementation of extension policies but, unlike EMBRAPA it does not itself directly execute extension programs. This is done by the state extension services (EMATERs) which are affiliated to EMBRATER. Since the establishment of EMBRATER there have been significant innovations in the entire national extension system (known as SIBRATER), particularly the introduction of plans to increase federal funding for specific programs, and to improve the support services being provided to low-income farmers. Further data on EMBRATER and the present range of the extension services are provided in subsequent sections of this report. Agricultural Credit 1.15 Agricultural credit services in Brazil were first formally organized under the Agricultural and Industrial Credit Department (CREAI) of the Bank of Brazil (BB) in 1937. Despite the generally unfavorable economic conditions of the period, the growth of agricultural credit was rapid and has continued to grow steadily since that time. The major defect of the system in the early years was that BB did not have branches in all rural municipalities and did not always possess adequate resources to meet the demand. However, by 1964 BB was servicing almost half a million (461,633) crop loans and 152,652 live- stock loans. To provide even broader credit service to farmers the Federal Government, in 1965, created the National Rural Credit System (SNCR) which allowed for the participation of the entire national banking system, and provided the Government with greater rural development and production policy controls through the channeling of credit resources. The main functions of the SNCR are to: (a) establish operational norms; (b) create new credit bills to respond to policy initiatives; and (c) to rationalize and facilitate the financing of rural credit at all levels by defining those who are entitled to credit facilities. Repayment terms are also supervised but may not, in any case, exceed limits set by the National Monetary Council (CMN). Interest rates, of 7-21%, are also highly negative in an economy which has registered annual inflation rates of as high as 40-45% during recent years. Despite the above, however, it cannot be said that credit is equally, and easily, available to all producers. The small farmer, with no land collateral and limited education, is often excluded from participating fully in institu- tionalized credit programs. As noted (para. 1.11) the Government has, more recently taken several important policy decisions, within the rural develop- ment framework, to address this problem directly. 1.16 In order to make more credit available the SNCR introduced a resolu- tion (No. 69) which established that private banks should apply 10% of their deposits to rural credit programs. SNCR also created a special fund (FNRR) attached to the Central Bank which would channel these funds from the private  7- - 7 banks and other sources (including foreign loans) thus providing discounting facilities on agricultural loans made by the private banks, and increasing substantially the amount of rural credit available. Under these and other direct Government policy incentives such as credit lines within regional and crop specific development programs (PROTERRA, PRODOESTE, and FUNAGRI) the total number of credit contracts grew from 1,145,209 in 1969 to 1,252,841 in 1971 with corresponding increases of total loan values from Cr$9.3 billion to Cr$12.9 billion respectively. The amount of agricultural credit issued in 1976 was estimated to be Cr$79.0 billion (current Cr$) or 52% of the total agricultural product for that year. Ministry of Agriculture forecasts indicate that a substantially increased amount would be available for credit operations in 1977. Since many of the official (federal and state) and private banks have either very limited or non-existent technical units, with staff agro- nomists for credit supervision in the field, the task of credit supervision, at the farm implementation level has increasingly become the responsibility of local and regional extension services in recent years (see also para 1.15). Agricultural Education 1.17 In 1976 there were 40 institutions of higher learning offering 57 programs in the agricultural sciences. Most of the schools have been started during the past 10 to 15 years; few are older than 25 years. Most of the animal sciences, forestry and fisheries study programs are of recent origin, since, traditionally, only agronomy and veterinary sciences were taught at the older, well established institutions. Total enrollment in the agricultural sciences in 1973 was 15,090 students, or only 16 per 100,000 inhabitants and an average of 3.06 students per 1,000 farm enterprises. The majority of the courses are offered at the undergraduate level and currently there are only 500 places per year for graduate level studies in 29 programs. Opportunities for graduate education are also still concentrated in the South and Southeast regions. Almost 80% of all graduates are employed by federal and state agencies. The number of agronomists, at the undergraduate level, is believed to be fully adequate to meet present demands but candidates with advanced degrees are still deficient in numbers to meet current demand and the number of post-graduate level courses offered is extremely limited. 1.18 The supply and demand position for skilled and mid-level (high school graduate) technicians is one which affects recruitment at all the major federal and state agencies dealing with agricultural development; but particularly so for the rural extension system which relies heavily on this level of personnel for the "in field" operations at which the most frequent extension agent/farmer contacts are made. Current assumptions 1/ are that total demand at this level will be 230,000 for the 1977-85 period, 185,000 (or 80% of this demand) would be for development projects in the North, Northeast and Central-West regions. However, estimated demand for both state and federal extension agents for the next four years (1978-81) period is not expected to exceed 4,000 or 2% of the projected total demand. 1/ Source: Appraisal of a Vocational Training Project in Brazil, Report No. 1549a-BR, May 1977, Annex 3.  - 8 - E. Bank's Participation in the Sector Lending Strategy 1.19 In its lending to Brazil, the Bank has sought to help the Government achieve a number of important rural development objectives, which are inter- dependent and complementary. One of these objectives is to support institu- tional development and policy reform, designed inter alia to help maximize public savings and ensure that they are used economically. This is also an objective of the proposed agricultural extension project, which would assist EMBRATER to improve the coordination and emphasize the production orientation of Brazil's agricultural research efforts in the Northeast, North and Central- West of the country. The proposed project would include extensive training and technical assistance components to strengthen EMBRATER. The institution- building objective has also been particularly important in previous Bank assistance to the Agricultural Research Corporation (EMBRAPA) under a loan (1249-BR) signed in June last year. EMBRAPA is now developing specific research programs, in the macro-regions mentioned, which are designed to provide the necessary agro-technology for expanded agricultural production and development of the rural sector. 1.20 A key element in the execution of these programs is the development of new technology through the expanded and strengthened national research programs under EM4BRAPA, followed now by a similar policy of strengthening the technical assistance and rural extension system (SIBRATER) under EMBRATER, the recently formed coordinating agency designated to finance and coordinate extension services on a national scale. The Agricultural Extension I Project, as proposed by EMBRATER to the Bank for partial financing, would appear to be well designed for the future execution of EMBRATER's mandate. The project goals are also compatible with the Bank's role in lending to the agricultural and rural development sector and the proposed loan would be a logical comple- ment to the Agricultural Research I Project, thus ensuring that new technology developed under that program would reach producers and become effective both in generating increases in agricultural production and in raising the living standards of the rural poor. 1.21 Another important lending objective for the Bank has been to help Brazil ease the foreign exchange constraint on its development by supporting projects designed to increase Brazil's export capacity and, where economical, to substitute domestic production for imports. The proposed extension project supplements previous projects to help Brazil reduce its dependence on imports of some key agricultural commodities (e.g., wheat, dry beans and natural rubber); maintain or develop self-sufficiency in the production of several basic food crops (e.g., beans, cassava and rice); and to gradually increase exports of others (e.g., beef, maize and cacao).  - 9 - Proposed Loan for Agricultural Extension 1.22 The proposed loan for an agricultural extension project would be the thirteenth for the development of Brazil's agricultural sector. The Bank has made two loans, totalling US$66 million, for livestock development; three amounting to US$155.2 million to support agro-industry and grain storage credit programs; one for US$6.7 million for the Alto Turi Land Settlement Project in the state of Maranhao; one for US$23 million for the Lower Sao Francisco Polders Project, one for US$40 million for Agricultural Research; one for US$12 million for the Rio Grande do Norte Rural Development Project; a loan for the Minas Gerais Rural Development Project of US$42.0 million; one for Rural Development in Ceara of US$17 million and a similar Rural Development loan of US$24 million, in the state of Paraiba has recently been negotiated. In addition, there have been three loans, totalling US$196.0 million for fertilizer production and a recent Bank loan (1302-BR) of US$19.0 million to strengthen the National Nutrition Institute (INAN). These last four loans have also a direct impact on agricultural production and development of the rural sector. 1.23 The proposal to assist Government in strengthening the national agricultural extension service, through EMBRATER, is entirely consistent and compatible with Bank lending strategy for the sector. The aims of the project would further strengthen government initiatives to obtain increased agricultural production, and, at the same time, it would have significant impact on rural development in general, but with a major emphasis upon programs of assistance to the rural poor, in those regions which have been designated as having priority in the Government's plans for agricultural development as a whole.  - 10 - II. AGRICULTURAL EXTENSION A. Historical Perspective 2.01 Several commodity organizations have provided limited advisory s-v4ces to their members since the early 1940s, examples being the Cocoa i<_.opment Institute (CEPLAC), the Brazilian Coffee Institute (IBC)9 the Council for the Development of Livestock (CONDEPE); and the National Plan for the Improvement of Dairy Livestock. Various agro-industrial firms have also provided such services. The Ministry of Agriculture, too, has provided limited assistance through what were referred to as "rural missions." The first state- wide service to be established was the Office of Coordination of Integral Technical Assistance (CATI) in the state of Sao Paulo. This was followed in 1949 by the formation of a Credit and Rural Assistance Association (ACAR) in the state of Minas Gerais. The latter was based on extension methodology that had been successful in the USA, and was in effect the forerunner of the exten- sion services now found in all the states in Brazil. Various international organizations and bilateral agencies supported those first extension efforts, under which the original US methodology was modified to fit the Brazilian situation. The pilot program in Minas Gerais was successful and the system spread throughout the other states. 2.02 In 1956, the Brazilian Rural Credit and Assistance Association (ABCAR) was established to integrate all existing state services into one national system. It was a non-profit civil corporation with semi-autonomous affiliates (ACARs) in 20 states 1/ and two territories, and was guided by a National Board of Directors (Conselho Deliberativo) which approved all plans and budgets. The ABCAR system ostensibly provided general extension services, but in practice concentrated largely on advising farmers on the use of credit and the preparation of credit applications. In many states, in collaboration with other agencies, the ACARs also initiated programs to improve local health and nutrition standards. There was great demand for such programs and since the state extension services were for the most part the only agencies with a network of local offices, they were by default and necessity involved in many functions that were not strictly speaking agricultural extension. 2.03 Between 1956 and 1965, the ABCAR system had established 254 local offices with 652 technicians. By 1970, it had 1,053 local offices and 2,556 technicians, and by 1974 it had grown to 1,440 local offices and 4,400 tech- nicians, assisting about 344,000 farmers. Each extension agent serviced about 40 farm credit applicants, while social assistance personnel reached about 12U families per year. Sixty percent of the ABCAR budget came from the Miaistry of Agriculture while the remainder came from State and Municipal Governments, and other development agencies, associations, and private enter- prises. In many states, individual ACARs entered into agreements with banks providing rural credit, further augmenting their income. For example, under 1/ The state of Sao Paulo has continued to operate its own independent agricultural extension organization (CATI).  - 11 - a contract with the Banco do Brasil (BB) the ACARs received 2% per annum of the outstanding balance of each loan serviced. Many such arrangements still exist. 2.04 The ABCAR system was quite successful in its broad approach to agri- cultural extension, but it suffered, perhaps inevitably, from a series of constraints, particularly a chronic shortage of funds, lack of continuity in funding, and a lack of continuity of policy and personnel. As the system developed, a heavy financial burden was placed on the Ministry of Agriculture because many of the programs ABCAR sponsored were, in reality, outside the field of agricultural production and productivity. Furthermore, attempts by the Ministry of Agriculture to modify programs met with resistance from the rural population, who felt that if the ACARs did not continue to provide social services, no one else would. Another major problem was that, due to the lack of effective research programs, extension agents had little new technology to pass on to farmers. Instead, they devoted most of their time to supervising credit, and the various other programs carried out on health, nutrition, child care, handicrafts, and general home economics. As a conse- quence true extension work suffered. Some ACARs, however, did develop their own field trials to test cropping practices, and were able to recommend improved agronomic practices to farmers. Lastly, there was considerable variation between states in the quality and coverage of the extension services provided, due largely to political circumstances and the fact that some states were much more developed than others. B. Present Status of Agricultural Extension The Brazilian Enterprise for Technical Assistance and Rural Extension (EMBRATER) 2.05 In the early 1970s, based on policies enunciated in the National Development Plan (PND I), the Government began strengthening agricultural services, assigning full responsibility to the Ministry of Agriculture. In 1972 the agricultural research services were reorganized and a new institution, EMBRAPA, was established. The new institution in effect took over a range of research programs that were previously administered by a series of separate organizations. The Bank has played a role in this development, and is currently financing a number of EMBRAPA's activities (Brazil Agricultural Research Project--Loan 1249-BR). Two years after the formation of EMBRAPA, in February 1974 the Brazilian Enterprise for Technical Assistance and Rural Extension (EMBRATER) was established 1/ to replace ABCAR, and the old state ACARs were renamed EMATERs. The purpose of this reorganization was to try and improve the quality and effectiveness of both research and extension services and to ensure that the activities of both were more closely coordinated. To formalize such coordination a National Commission for Agricultural Research and Technical Assistance (COMPATER) was also set up, on which are represented both EMBRAPA and EMBRATER (para 2.12). 1/ Law No. 6.126 of 6/11/74 and Decree No. 75.373 of 14/2/75.  - 12 - 2.06 The formation of EMBRAPA has led to a much more centralized admin- istrative structure for research than existed previously, but this is not the case with ex- nsion, and the new EMATERs, being affiliates of EMBRATER, have retained a 1,-rge measure of the independence of their predecessors, the old state ACARs. irie new institutions do, however, have closer links with each other, and. 1r emiphasize that such links exist between the federal agencies, the entire structure is referred to as the National Technical Assistance and Rural Extension System (SIBRATER). Thus, the latter in effect encompasses both EMBRATER, the central coordinating body at federal level, and 23 EMATERs at state or territorial level. 2.07 The new SIBRATER system is still in the process of becoming estab- lished, but it is envisaged that, although the individual EMATERs will imple- ment extension programs according to local circumstances, EM4BRATER will be influential in three respects: (a) Administration. EMBRATER holds a capital share in each EMATER and participates in the selection and nomination of the management. Legal norms in the different state extension agencies are equally applied, and financing is on the basis of signed contracts between EMBRATER and the EMATERs. (b) Planning. Legal instruments allow EMBRATER to play a role in planning projects. General policies and goals are outlined at the federal level to serve as a framework for these plans. (c) Operations. EMBRATER headquarters staff are assigned to the state affiliates for purposes of consultation, orientation, and training. 2.08 Being a new organization with new staff, EMBRATER has yet to become fully established, and like many new organizations, has experienced teething problems. There have been delays in the provision of funds and the organiza- tion has been reluctant on occasion to assume the role of leader and catalyst as aggressively as some would wish. On the other hand, since the individual state EMATERs have had a long tradition of independence, this is understand- able and EMBRATER is probably wise to tread with care in the early years. Fortunately, there is general agreement in Brazilian extension circles on the need for a central coordinating agency such as EMBRATER, and the institution has for the most part substantial support from the state EMATERs. State expectations are in fact high and EMBRATER has a major task ahead of it to satisfy such expectations. In order to achieve the desired goals in the area of planning, coordination and execution at both federal and state levels, a much stronger organizational structure will have to be built which can be more responsive to the special needs of both state and federal programs. Indications are that, assisted by the proposed project, this can be achieved. State Extension Services 2.09 Extension Coverage. According to the 1970 agricultural census there were approximately 4.9 million farm properties in Brazil, and in 1977 there were 9,140 agricultural technicians working in the SIBRATER system.  - 13 - Of these 7,077 can be classified as working at the field level, in which case there is at present an extension agent/family ratio of 1:700. A great many farm families are not, however, being reached by the SIBRATER system. Exact data are difficult to obtain, but it is thought that about 565,000 farm families are being contacted through group meetings and individual visits. The figure does not include those being reached through the mass media. It is believed, therefore, that each technician on average contacts about 80 farmers, but there are wide differences between and within individual states and geo- graphic regions, namely: North 70:1, Northeast 55:1, Central-West 101:1, South 183:1, Southeast 112:1. It should be borne in mind that a great many of these contacts, probably the majority of them, are still primarily to supervise credit on behalf of the banking system, and that in most states few of the contacts involve small farmers. It is evident therefore that there is a substantial need to increase the number of extension agents, improve their effectiveness, and increase the number of contacts per agent (see also Annex 3, Tables 8 and 9). 2.10 Administration. Circumstances differ widely from state to state, particularly the degree of extension coverage, but with most EMATERs a state is divided for administrative purposes into regions. Within each region there are a number of field offices, each of which service farmers in a number of municipalities. Details of the actual number of staff in each state (1977) are provided in Annex 3, Table 10. 2.11 In all states actual extension work is very much program oriented, and consists of four main categories. These are (a) crop programs, with 38 crops involved; (b) producer specific programs, i.e., low-income farmers and "artisanal" fishermen; (c) production support programs, based on coopera- tives, use of agricultural inputs (such as improved seed, fertilizer, credit) and training for rural workers, and (d) social welfare programs. Many of the programs are closely interwoven, particularly the social welfare programs for low-income farmers who may well be the beneficiaries under a variety of programs. At the state (EMATER) level each program has a manager who is responsible for the operation of the specified program in the state.- He supervises the programs at the local (municipal) level. The total number of municipalities receiving extension services in each state or territory varies enormously, from two in Rondonia to 686 in Minas Gerais. However, program work in any one state usually reaches an average of 50-60% of the municipalities for any one program. In the case of specific crop programs the state managers can, to a limited extent, call on subject matter specialists through EMBRATER headquarters, or the local research service (EMBRAPA) as may be necessary. Each project manager is responsible for the supervision of about 6 regional supervisors (senior level technicians) and each of these supervisors has 10 or 12 field staff (middle level technicians) under his supervision for a total of 60-72 field personnel per managerial unit. Given the great diversity in the degree of agricultural production and rural development in each state these averages can, however, vary enormously. Administrative linkages between the EMATERs and EMBRATER are shown at Annex 3, Charts I and 2.  - 14 - 2.12 Extension Methodology. Thorough extension methods vary between states, for the most part the classical personal contact system is most widely practiced, i.e. with the extension worker meeting the farmer on a person to person basis. Group methods have not been widely practiced, partly because the necessary technical facilities are lacking and partly because of the emphasis on credit supervision which cannot be other than on a personal basis. In the two years since assuming control of the exten- sion system network (SIBRATER) at the national level, EMBRATER has not imposed a uniform methodology on delivery (information dissemination process) at the field level. Nor should it attempt to do so until many of the problems unique to the rapidly evolving agricultural industry, strongly influenced by government policy and economic incentives on a macroregional basis, can be adequately studied to provide the most effective methods to be applied in each case. EMBRATER is, however, increasingly conscious of the need to find ways and means of reaching more farmers, and recognizes that where small farmers are concerned there should perhaps be more emphasis on group methods and the greater use of well designed audio visual aids. Officials believe there is urgent need for studies and experimentation (methodology studies) to arrive at more effective methods. There is also recognition that a variety of approaches are probably necessary depending on such factors as farm size, population density, cropping patterns etc., all of which vary greatly between and within states. 2.13 Current Financing Arrangements. Based on budget estimates for 1977, allocation for the National Agricultural Technical assistance and Rural Extension System was US$147.2 million. Of this, US$134.7 million was allocated to the 23 state and territorial associate companies (EMATERs and ASTERs) and the remainder (US$12.5 million) was allocated to EMBRATER headquarters. Allocation of the state/territory funding by region was as follows: North 12%, Northeast 32%; Central-West 15%; South 19% and Southeast 22%. The source of funds varies considerably in the case of the EMATERs and ASTERs. However, approximately 60% of EMATER funding is derived from federal sources and is channeled through EMBRATER, Brasilia. In the case of the territorial ASTERs funding may be 90% federal and also be controlled through EMBRATER. Much of the federal funding, particularly in the territories where land settlement is occurring, is tied to special development programs such as POLAMAZONIA and POLOCENTRO. In the case of the northeast there are special programs for rural development, under POLONORDESTE, and programs specifically designed to assist the low-income farmer. 2.14 Field Offices. Due to the diversity of the country, there are many variations in local extension units and none can be considered as being typical of the whole system. At one extreme there are two-man units, staffed by either a veterinarian or an agriculturalist and a clerk, while in some areas they may be as many as 15 or 16 on the staff, e.g., five higher-level technicians, five middle-level technicians or field assistants, two social assistants (home economists), and four lower-level field workers. Usually these local extension units are housed in offices supplied by the municipality -- usually a rented house or office in the municipal building,  - 15 and some are combined with regional extension units. In some areas however, the location of the municipal offices may bear no relation to the agricultural production areas, and, therefore, the local offices are not readily accessible to farmers. Most of the local offices have a vehicle for every two technicians plus a few items of demonstration equipment and, occasionally, a slide projector. Frequently, however, office buildings and equipment are very deficient. Staff training facilities are also for the most part inadequate, particularly for pre- and in-service training, and in very many cases there are few back-up or subject matter specialists on which field staff can rely for specialist advice and assistance. 2.15 Extension Constraints. The contraints experienced by the old ABCAR system have already been noted (para 2.04), particularly lack of continuity of financing and policy making, a chronic shortage of funds, wide variations in quality and coverage between states, and dilution of effort caused by the need to both supervise credit and service a variety of social programs. Many of these constraints still exist, in fact by their very nature it is apparent that they cannot be rectified in the short term. Perhaps the most basic difficulty relates to lack of continuity of funding, because, in the absence of a secure source of funds over the long term, it is hardly surprising that staff and facilities are often inadequate and that individual EMATERs feel compelled to undertake such functions as supervising credit (for which they receive a fee from the banks) and servicing social programs (for which funds are obtained from many sources). Government has become increasingly aware of the nature of these difficulties, and believes that the creation of EMBRATER is one of the first steps that needed to be taken in order to address basic funding and policy issues. It also feels that, as in the case of research, the Bank can play a useful role by helping to provide continuity of funding and the wherewithall to both expand services and improve their quality. Officials realize, however, that due to the size and variability of the country and the complexity of many of the problems involved, a number of the constraints mentioned may well take years to remedy. C. Interactions with Other Agricultural Services National Role of EMBRATER 2.16 A primary function of EMBRATER is to collaborate with other govern- ment agencies involved in stimulating agricultural production and carrying out rural development policies, and to promote, encourage, and coordinate tech- nical assistance initiatives. With this in mind, in April 1975, EMBRATER published "Guidelines for the Operational Structure of EMBRATER," prepared by 15 working groups consisting of interdisciplinary teams of 155 technicians, representing 45 public and private institutions, plus 15 officials of EKBRATRER. The purpose of the guidelines was to establish the mechanisms required to coordinate the work of EMBRATER with the other national agencies handling agricultural research, rural credit, marketing, rural cooperatives, and rural development programs involving education, nutrition, health and vocational training. These mechanisms are still being worked on in detail.  - 16 - Relationship to EMBRAPA 2.17 A National Commission of Agricultural Research and Technical Assistance (COMPATER) has been set up as the coordinating link between EMBRATER and EMBRAPA, and is responsible at the federal level for synchronizing, review- ing, and coordinating the programs of both agencies. The primary role of COMPATER is to ensure that both the research and extension agencies are responsive to national development priorities and goals. Members of COMPATER are: one representative of the Ministry of Agriculture, as Chairman; the presidents of EMBRAPA and EMBRATER; one representative each from the Central Bank, the National Farmers Confederation and the National Agricultural Workers Confederation; and two representatives from agro-industrial interests in the private sector. COMPATER is jointly financed by EMBRAPA and EMBRATER. EMBRAPA's Department of Diffusion of Technology (DDT) and EMBRATER's Department of Technical Support (DTS) act as liaison units between the two. At the state level, the representative departments responsible for diffusion of technology in the two institutions work together in planning, executing, and evaluating research projects and in training staff. Extension bulletins (Pacotes Technologicos) are now being prepared as a result of discussions between researchers, extensionists, and producers, and research staff from EMBRAPA are helping with the training of extension personnel. Similarly, extension staff assist in planning field days on EMBRAPA experimental farms. Relationship with Rural Credit Institutions 2.18 EMBRATER wishes increasingly to free extension staff from adminis- trative functions relating to credit so that staff can concentrate on exten- sion activities and the dissemination of new agronomic technology. Discussions are routinely held at the highest levels with the National Commission for Rural Credit (COMCRED), the Guarantee Program for Agricultural Production (PROAGRO) and the Banco do Nordeste (BNB), and Banco do Amazonas (BA). It has been decided that EMBRATER will continue to assist the recipients of rural credit but will increasingly emphasize loans to small and medium landholders through cooperatives or farmers' associations. This is a desirable shift in policy, but it is appreciated that it is not a policy that can be implemented overnight since the supervision of credit to larger farmers has been a tradi- tional function of the state extension services since their inception. Collaboration with Marketing Organizations 2.19 EMBRATER believes that much can be done to improve agricultural marketing facilities and is currently emphasizing assistance to cooperatives and informal farmer associations. In this respect, EMBRATER marketing special- ists are establishing coordinating mechanisms at both national and state levels with the Brazilian Food Company (COBAL), the Commission for Financing Production (CFP), the Brazilian Storage Company (CIBRAZEM), the National Department of Marketing Services (DNSC), the National Superintendency of Supplies (SUNAB), and the Department of Inspection of Products from Animal Origin (DIPOA). Preference is being given to producers who are organized in groups or cooperatives, and in 1976, 226 cooperatives, comprising 212,773 members, were given technical assistance. Training and support were provided on planning, management, maintenance of equipment, storage processing and marketing.  - 17 - 2.20 EMBRATER is also collaborating with private industries that produce agricultural inputs such as fertilizers, insecticides, fungicides and veterinary products to provide technical assistance programs. Joint training programs, demonstration plots, campaigns and production of informative materials are resulting in more uniform recommendations to the farmers. EMBRATER has established an accreditation system for private firms which want to undertake technical assistance work under contract with the Banco do Brasil and other financial institutions. Such accreditation includes guidelines on procedures and sanctions for possible irregularities. Collaboration with Social Institutions 2.21 Special agreements are being signed with the Ministry of Health (MINISAUDE), the National Nutrition Institute (INAN), the Ministry of Education and Culture (MEC), and the Ministry of Labor (M4TSS) to introduce programs that indirectly support agricultural production in the fields of health, nutrition, literacy, formal rural education and manpower training. These agreements at the national level will include coordination of such programs at the state and regional levels. Some of the major agencies in this area are: the National Employment System (SINE); the Program of Assistance to Rural Labor (PRORURAL); the National Foundation for Occupational Safety, Hygiene and Medicine (FUNDACENTRO); the Rural Health Program (PIASS); the Center for Education and Feeding of Preschool Children (CEAPES); the National School Food Campaign (CNAE); the National Food and Nutrition Institute (INAN); the Intitute of Food Technology (ITAI); the National Food and Nutrition Program (PRONAN); the Public Health Services Foundation (SES); and the National Service for Rural Training (SENAR) generated by the Ministry of Labor. The interaction of the activities of these various agencies is extremely complex. Insofar as EMBRATER is concerned, it believes that it has an important catalytic role to play to encourage other agencies to improve social and support services to the rural population, and to support and service such agencies if their programs require an input from trained extension staff. Traditionally the state extension services have been heavily involved, within the limits of their available resources, in providing many such services directly. This has largely been because the state extension services, being relatively decentralized and with a network of local offices, have been most closely in touch with rural people. There is increasing recognition that the provision of social services is not the real function of an extension service, but as with credit supervision, the state services feel they cannot withdraw until the other agencies, whose job it is to provide social services, are better developed. 2.22 A broad overview of the interlocking agency responsibilities under the various policy-making councils, and EMBRATER's role in many of the rural development programs can best be visualized from the table provided at Annex 3 (Chart 3).  - 18 - III. THE PROJECT A. Introduction 3.Oi As already noted Brazil's economy has traditionally maintained a strong agricultural base and the country has great potential for further agricultural development (paras 1.01 and 1.02). Government believes that in order to realise this potential agricultural support services require substantial expansion and improvement, particularly research and extension services. It is also increasingly accepted that such services need to cater to a much greater extent than in the past to the needs of smaller farmers, of whom there are very many (para 1.07). Steps have already been taken to strengthen and improve the country's research services (para 1D13) and plans are now in hand to similarly strengthen and improve the extension services (para 2.05). Both activities are complementary one to the other, and since the Bank is already heavily involved in the former, it is a logical sequence of events for it also to be involved in the latter. 3.02 In a letter received from Brazil's Minister of Agriculture, dated January 21, 1976, the Government requested Bank assistance to strengthen Brazil's National Technical Assistance and Rural Extension System (SIBRATER). A Bank identification mission in June 1976 was followed by a preparation mission during November of the same year. Subsequently, a pre-appraisal mission assisted the proposed executing agent for the project, the Brazilian Enterprise for Technical Assistance and Rural Extension (EMBRATER) to prepare the project proposal document which was received by the Bank on May 20, 1977. 3.03 The Bank appraisal mission, under terms of reference dated June 7, was in the field during the period of June 16, to July 20, 1977. Subsequent changes to the original project document, necessary to comply with Government's current development plans, were made by EMBRATER's project preparation team. These changes were reviewed by a post-appraisal Bank mission during late November 1977. B. Objectives, Description and Scope 3.04 The main focus of the proposed project is essentially institution building and, as an overall objective, the project would aim to strengthen EMBRATER's capability to provide national leadership and overall coordination in order to provide more efficient and effective extension services on a national scale. More specifically the project would aim to: (a) improve the continuity of funding to both EMBRATER and its affiliate EMATERs and ASTERs, thus providing them with the financial strength and security necessary to improve, modify and expand extension services to farmers;  - 19 - (b) make extension services increasingly available to about 1.3 million farmers, an increase of about 130% over current levels; (c) improve the quality of the services already being provided and expand their coverage, thus enabling farmers to have better access to up-to-date data on cultural practices, market prospects, credit arrangements and production incentive programs; (d) improve the methods of communication now under use to include group methods by introduction of audio-visual aids, stimulation of cooperative efforts and greater availability of simple educational bulletins and pamphlets covering all aspects of agricultural production; (e) generally raise the income levels of farmers, particularly an increasing number of small farmers; (f) strengthen the technical planning, monitoring, evaluation and administrative capabilities of both EMBRATER and EMATER staffs. 3.05 In order to achieve these objectives the proposed project would comprise: (a) Programs to improve and strengthen existing state agricultural extension service facilities by: providing for additional staff, library documentation and current data services, audio-visual aids, training centers, state offices and facilities for crop production and small scale (artisanal) fisheries cooperatives; (b) Further development of technical and administrative staff capability by: providing pre-service and in-service training, specialized short courses and graduate training for selected senior staff; (c) Provision of technical assistance by means of: consultant services and extension methodology studies designed to improve existing extension programs; (d) Specific operational programs with special emphasis on assistance to low-income farmers by means of: production systems, field demonstration plots, pilot farm cooperatives and "artisanal" fisheries colonies. 3.06 Investments would be made under three main categories: (a) capital investments for civil works, vehicles, equipment, library and documentation systems and office furnishing; (b) technical services to include long- and short-term fellowship programs, specialized short courses, pre- and in-service training, use of consultant services and methodology studies, and (c) incre- mental establishment costs to cover salaries of additional staff, materials and supplies, and the necessary operational costs for such incremental staff during the four-year project period.  - 20 - 3.07 The project would be national in scope but two-thirds of the emphasis would be placed in the priority areas of the North, Northeast and Central-West. The remaining Southeast and Southern regions would be considered non-priority areas. The project is thus complementary to the Bank's Agricultural Research I Project (Loan 1249-BR). It also supplements the PROCENSUL project, partially funded by an IDB loan, which provides assis- tance to research and extension programs in the non-priority areas described above. C. Detailed Features Capital Investments 3.08 These would include: (a) the construction of eight training centers, 10 offices for state EMATERs, 20 base-operation complexes for "artisanal" fishery colonies (including storage facilities) and 10 offices for model fisheries cooperatives. Land purchase, although not financed under the loan, would be a capital cost at 40 of the 48 building sites; (b) 4,263 vehicles of all types including field vehicles, trucks and small buses; (c) equipment and furnishings for 10 offices and 24 libraries; (d) printing, audio visual and communications equipment for 600 documentation centers; and (e) farm and fisheries equipment for 8 training centers, 10 production systems units, and 20 pilot model fisheries colonies. Training and Technical Assistance 3.09 Pre- and in-service training courses for a total of 4,725 1/ incoming staff and annual training for an average of 3,400 permanent technical staff would be provided at the eight training centers. In addition, short courses in specific subjects in the areas of library science, communications, data processing and fisheries would be provided for 381 field and administrative staff. A fellowship program to provide post graduate training for selected subject specialists would include 174 at the MSc level and 36 at the PhD level. Seventy-five manyears of consultancy services, either local or expatriate according to need and availability, would be provided in eight specific study areas as follows: (a) Administration; (b) Monitoring and Evaluation; (c) Training and Human Resource Development; (d) Architecture and Engineer- ing (civil works supervision); (e) Extension Methodology Research; (f) Communications; (g) Fiscal Management; and (h) Data Processing. Under the same technical assistance program funds would be made available to carry out extension methodology studies in three selected states representative of the challenges to be encountered at the various levels of the rural development process. Methods whereby group communication can supercede individual contact between extensionist and farmer would be tested and evaluated. 1/ Includes a 30% increase over designated recruitment level to cover normal attrition during project period, but particularly during first year of service.  - 21 - Incremental Establishment Costs 3.10 In crder to achieve the rapid expansion in extension coverage from 565,000 farmers in 1977 (the current level) to 1.3 million farmers in 1981 (as proposed under the project) an additional 3,734 extension staff would be recruited. This is at first glance an appreciable figure but it should be viewed in the context of a very large country (8.5 million sq. km) and the fact that new staff would be dispersed between 23 separate state extension services and EMBRATER headquarters. It should also be borne in mind that even with the additional staff the combined extension services are only pre- dicated to be able to reach about a quarter of the country's farmers. The bulk of the new staff would be technicians, either actual field and extension program staff (2,692) or technical support staff (368). The remainder (674) would be clerical and administrative staff. Thus the emphasis of this compo- nent of the project is on strengthening technical capability and coverage at the field level. Details of the work (programs) on which they would be deployed are provided later in this report and are tabulated in Annex 3, Tables 11 and 12 As a result of the project the total numbers of all staff involved in the SIBRATER system (EMBRATER, EMATERs and ASTERs) would increase from 14,760 in 1977 to 18,494 in 1981 an increase of 25% over a four-year period (Annex 3, Table 13 and 14). The project would also finance, in addition to the salaries of incremental staff, the incremental operating costs incurred by such staff. D. Project Costs 3.11 The total cost of the project over the four-year period (1978-81) is estimated at US$284.9 million, of which about US$40.4 million, or 14%, repre- sents the foreign exchange component. Costs are summarized in the table overleaf and are detailed in Annex 3, Tables 15 and 16. Costs were estimated at the end of November 1977 with a current cruzeiros/dollar exchange rate of Cr$15.435 to US$1.00. Physical contingencies 1/ were estimated to be 4% of baseline cost. Price contingencies of 19% 2/ were calculated assuming that commitments would not begin before May 1978. Baseline costs of US$231.5 million were estimated from prices at appraisal (June/July 1977) and adjusted to reflect November 1977 prices. 1/ Physical contingencies were calculated as follows: 20% on civil works, this high contingency is necessary due to the large number and great diversity of the construction component with several small buildings at remote sites not clearly identified; 5% on vehicles and farm equip- ment; 7% on all other capital items; and 10% on materials, supplies and other operating costs and 0% on all other items. 2/ Price contingencies were calculated in US$ terms as follows: for civil works, 9% per year for 1978-79 and 8% per year for 1980-81; for equipment, 7.5% per year in 1978-79 and 7% per year in 1980-81; for material, supplies and salaries, 7% per year for the period 1978-81.  - 22 - BRAZIL AGRICULTURAL EXTENSION I PROJECT Total Project Costs (US$ millions) % of Total % of Baseline Foreign Capital Investments Local Foreign Total Costs Exchange Civil Works 13.98 3.50 17.48 8 20 Land Purchase 0.13 - 0.13 * - Vehicles 13.06 5.59 18.65 8 30 Office Equipment & Furnishings 3.78 2.03 5.81 2 35 Printing & Audiovisual Equipment 3.42 5.14 8.56 4 60 Communications Equipment 0.43 1.72 2.15 1 80 Farm Equipment 1.31 0.56 1.87 1 30 Fishery Equipment 1.92 0.21 2.13 1 10 Library Equipment & Furnishings 0.71 0.31 1.02 * 30 Miscellaneous Capital Items 2.84 0.71 3.55 1 20 Sub-total 41.58 19.77 61.35 26 30 Training and Technical Assistance Pre- & In-service Training 5.24 - 5.24 2 - Short Courses 0.17 0.67 0.84 * 80 Short- & Long-Term Fellowships 1.15 3.45 4.60 2 75 Consultants 2.25 3.38 5.63 3 60 Methodology Studies 2.71 - 2.71 1 - Sub-total 11.52 7.50 19.02 8 40 Incremental Establishment Costs Staff Salaries 124.72 - 124.72 54 - Materials & Supplies 9.28 3.98 13.26 6 30 Other Operating Costs 13.20 - 13.20 6 - Sub-total 147.20 3.98 151.18 66 3 TOTAL BASELINE COSTS 200.30 31.25 231.55 100 14 Physical Contingencies 6.59 2.20 8.79 4 25 Price Contingencies 37.63 6.95 44.58 19 16 TOTAL PROJECT COSTS 244.52 40.40 284.92 123 14 * Less than 0.5%  - 23 - E. Financing 3.12 The proposed Bank loan of US$100.0 million to the Federative Republic of Brazil would finance 35% of total project costs. The Government, through specific budget allocations to EMBRATER, would finance US$111.0 million (39%) and the rema4ning US$73.9 (26%) would be financed by the state/territory exten- sion agencies, under individual contract agreements with EMBRATER. The pro- posed loan would be for 15 years including three years of grace. In a country like Brazil, which is making a vigorous effort to mobilize domestic resources, it appears appropriate for the Bank to give some assistance in the financing of local currency expenditures on projects whicb, like the present one, have very high priority and yet have a very low foreign exchange content (in the present case, i4%). If the Bank is to be effective in supporting projects of this kind5 having a significant impact on the agricultural production of basic food crops and the welfare of the rural poor, it seems reasonable for it to contribute at least 35% of the costs even though this entails financing some local currency expenditures. During negotiations assurances were obtained that (a) the Government provide, in a timely manner, the funds necessary for EMBRATER to execute the project and that (b) contractual arrange- ments satisfactory to the Bank would be entered into between EMBRATER and its associates under which each EMATER/ASTER would agree, inter alia, to submit to EMBRATER for approval, by October 31 of each year, its annual budget for the following year, which budgets would include adequate allocations from the state to cover expenditures required for the project and not covered from other sources. The conclusion of such arrangements would be a condition of disbursement in respect of expenditures of the EMATER/ASTER concerned. The proposed phasing of investments is given in Annex 3 (Tables 15 and 16). Proposed financing would be as shown below: Government Bank Total US$M % US$M % Investment Capital Investments 36.2 59 25.1 41 61.3 Training and Technical Assistance 8.4 44 10.6 56 19.0 Incremental Establishment Costs 107.5 71 43.7 29 151.2 Base Line Cost 152.1 66 79.4 34 231.5 Physical and Price Contingencies 32.8 61 20,6 39 53.4 Total Project Cost 184.9 65 100.0 35 284.9 F. Procurement 3.13 The major item for procurement is civil works (US$17.48 million). This covers the construction of a total of 48 buildings. Brazil has a compe- titive local construction industry capable of carrying out project works being financed. Major contracts would be just under US$1.0 million each for the eight training centers and 10 state EMATER offices. At the remaining 30 construction sites contracts would be for approximately US$150,000 each. Due to the very dispersed nature of these contracts it is doubtful whether foreign  - 24 - contractors, or even major national contractors, would be interested in such small individual construction contracts. Contracts would be awarded on the basis of competitive bidding advertised nationally in the case of the training centers and state EMATER offices and locally advertised in the case of other project facilities and in accordance with local procedures, which are satisfactory to the Bank. Foreign bidders would not be excluded. In remote areas, where it is difficult to obtain contractors through competitive bidding, work for force account would be permitted. About 4,260 vehicles (US$18.65 million) would be procured under the project, over a period of four years, for use in widely dispersed locations. Since vehicles would be purchased in small lots, and adequate maintenance and availability of spares is of critical importance, locally manufactured vehicles of types already used by government departments are preferred. Therefore, contracts for vehicles would be let on the basis of competitive bidding advertised locally and in accordance with local procedures which are satisfactory to the Bank. Farm and fisheries equipment (US$4.0 million) and office equipment and furniture (US$5.8 million) would be required at widely dispersed locations and contracts for those items would also be awarded on the basis of competitive bidding advertised locally and in accordance with local procedures. 3.14 Printing, audio-visual and communications equipment (US$10.7 million) is often highly specialized and after-sales service is of critical importance for such items. Since servicing must be available locally, local competitive bidding under procedures acceptable to the Bank would be followed for items estimated to cost US$10,000 or more, and prudent shopping would be followed for items of lesser value. 3.15 Library equipment and furnishing (US$1.02 million) which are of local manufacture, with a very competitive field of suppliers, would be procured according to Goverment procedures which are satisfactory to Bank. Training and technical assistance services, (US$19.02 m) comprising pre-service and in- service training, methodology studies, consultancy services, fellowships for advanced training and specialized short courses would be either obtained abroad (32%) or in Brazil (68%). In view of the difficulty of locating competent specialists abroad, and of arranging and monitoring the overseas training, EMBRATER would utilize the services of one or more agencies with qualifications and experience and on terms and conditions satisfactory to the Bank, to assist it in obtaining technical assistance services from abroad and in administering the overseas training included in the project. Assurances to this effect were obtained at negotiations. G. Disbursement 3.16 The Bank loan would be disbursed against (a) 100% of all direct foreign exchange expenditures (US$14.8 million), and (b) 35% of all other project expenditures (US$85.2 million), but excluding those for materials, supplies, other operating costs and land purchase. Details of the expected disbursements of the Bank Loan by project component are given in Annex 3, Table 17.  - 25 - 3.17 Disbursement requests would be submitted by EMBRATER to the Bank accompanied by appropriate documentation showing proof of bidding procedures, assignment of contracts for services and receipts of payments made against such documentation. In the case of incremental staff salaries, the appro- priate documentation would be a summary of those staff salaries applicable to the project and based on positions incremental to the staff list at December 31, 1977. The estimated schedule of disbursement is shown in Annex 3 (Table 18). H. Accounts and Audit 3.18 Separate accounts would be kept by EMBRATER of expenditures made under the project. These would be audited anniually for each fiscal year, in accordance with sound auditing principles, consistently applied. Accountts would be submitted for audit within three months of each year, and submitted to the Bank promptly. Certificates of expenditures, including satisfactory evidence of incremental staff salaries, would be audited aL least once in six months and the report furnished to the Bank. EMBRATER's accounts have been audited in recent years by the firm of Laudon, Blomquist and Co. Suicable assurances covering these accounting and auditing requirements by an independent auditor acceptable to the Bank were obtained at negotiations.  - 26 - IV. PROJECT IMPLEMENTATION A. Organization and Management 4.01 Headquarters Administration. The federal agency (EMBRATER) would have full responsibility for the formulation of project policy and directives to ensure that project implementation achieves the desired objectives. To this end it would also coordinate the essential phases of project monitoring and evaluation of results at the total project (federal) level. Execution and actual implementation of project programs, at the field (i.e., farm) level would be the responsibility of the associate agencies (EMATERs and ASTERs), working through the state or territorial offices as the case may be. Project staffing has been evaluated and incremental staff included as necessary to carry out the tasks assigned. 4.02 Liaison between Federal and State Levels. As both federal and state organizations have basically the same administrative structure (see Annex 3, Charts 1 and 2), the easy and direct flow of communication between corre- sponding levels of administration affecting operations at both state and federal levels would continue as at present. The only change would be the establishment of a project coordinator in the Technical Support Unit, in the Operations Department of EMBRATER. The office of the coordinator would, as in other areas of administration, have direct access to counterpart staff in the Technical Support Units of the State EMATERs and would be directly respon- sible to the chief of EMBRATER's Operations Department. 4.03 Technical Support Unit. The necessary technical and administrative services required by the project coordinator to monitor the progress of project implementation would be supplied by those departments responsible, i.e., Planning, Program Development, Administration and Finance, and Operations. For the supervision of the civil works component, the consultant services of a civil engineering/architectural firm would be provided under the Technical Assistance component (Annex 3, Table 19). The project coordinator would handle project-related correspondence with the Bank through the office of the Department of Operations and the Presidency of EMBRATER. The project coordi- nator would also be responsible for the submission of project progress reports, at six-monthly intervals, as required by the Bank. Assurances to this effect were obtained at negotiations. 4.04 Policy Review and Program Planning. All programs which would be implemented under the project strongly reflect current policy planning for agricultural production and rural development. 4.05 Budgeting of Programs at State Level. All programs require close collaboration at the federal (EMBRATER) and state (EMATER) levels. The fiscal year corresponds to the calendar year for the (SIBRATER) system and the budget- ing process commences during May/June of the previous year in the department of operations in each EMATER. These plans then pass to the programming and budgetary unit for review and there undergo cost/ finance reconciliation in the department of administration and finance. When approved by the president's  - 27 - council, the state (EMATER) budgets then go to EMBRATER headquarters in Brasilia about July/August. A similar operational-planning-financial and advisory board review takes place at the EMBRATER level. However, federal policies on a regional rather than state basis are considered aDa state plans may be revised accordingly. Budgets, as approved by EMBRATER are then returned to the states in September and final adjustments made no later than October. B. Field Implementation and Extension Methodology 4.06 As already mentioned, for administrative purposes much of the work of the extension services is classified under various specific programs (para 2.11). Under the proposed project certain of these programs would be substan- tially strengthened and expanded, the details being as follows: Production Systems 4.07 Considerable research effort is being devoted to devising improved cropping patterns and agronomic practices for adoption by farmers, the results of which are being translated into technical bulletins advising farmers on the merits of various farming systems. Some 247 such bulletins have already been produced, 195 dealing with crops and 52 with livestock. The methods used are simple and straightforward and involve cooperation between extension and research personnel. The program is currently administered by the Departments of Operations at both federal and state levels. The expanded program proposed under the project would be administered in the same way. There is one manager for the federal program, stationed at EMBRATER headquarters, and he supervises the five regional managers in the major regions of the country. Regional managers make contact with program supervisors in each state through the EMATER or ASTER. State and regional managers report to the national program manager and plan the Production Systems Program and budget for each year based on state, regional and national priorities derived through consultation with pro- ducers, research staff and field extension staff. Crop production specialists from EMBRATER are also used as advisors during the planning and evaluation stages. Based on the experience of the past few years (1974-77) EMBRATER staff have noted that the farming systems programs for annual crops require revision every two years while those for biennials or perennials may only require revision every three to five years. In the case of livestock, certain individual aspects, i.e., pasture management, nutrition or breeding may be the subject of specialized bulletins issued every three to five years. 4.08 The incremental staffing requirement for the expanded program would be 24 additional technical staff, mainly crop specialists (21) and three regional managers. Staff distribution by region is shown at Annex 3, Table 13. About 80% of the staff will, in effect, work in the priority areas. The new program goal for the project period is for 285 new bulletins with distri- bution as follows: annual crops 122; perennial and tree crops 83; and live- stock 80. An average of 5,000 copies of each bulletin would be printed for use by EMBRAPA and EMBRATER staff. This program is also useful in that it provides a non-formal, in-service training facility for technical field  - 28 - personnel of the extension service. Appropriate candidates are selected from the region to attend the meetings at which the farming systems are elaborated. The selected candidates also make periodic visits to the experimental fields and farm trial plots involved in the specific program during the growing season. Such attendance is usually limited to five extension agents (trainees) per subproject. Based on the figure of 285 such subprojects for the four-year period, this training exercise should reach a total of 1,425 technicians. Field Demonstration Plots 4.09 Extensive use is already made of demonstration plots, often coupled with field days for farmers groups, to disseminate ideas on new agronomic practices. They are for the most part elaborately planned some 9-12 months ahead of time by extension and research staff working in close cooperation. The entire demonstration plot program is operated as a national program with EMBRATER providing overall policy and planning to provide the desired regional emphasis. There are a project manager and two assistants in EMBRATER while each EMATER has a manager for field operations for the program within the state. Project managers, at the federal and state levels report to the Chief of Operations in each case. As a direct result of increasing the number of demonstration plots, an assistant project manager would be assigned in 11 states and an additional 60 field technicians would be required under the program. These would be employed at the state level for program implementa- tion, with approximately five additional staff members per state in 12 selected states in the priority regions. Ten technical support staff per state would also be added and 10 at EMBRAPA headquarters to operate publications and audio visual aids programs based on the field demonstration plot technique. 4.10 Program goals call for 676 plots in the first project year; 598 the second year, 715 the third year and 586 by the fourth year. As only those for annual crops are changed on an annual basis and those for permanent crops, and pasture for livestock, may continue for several years, the figures represent the number of plots being maintained in any one year. As with the farming systems program, training opportunities would be extended to a fairly wide range of extension agents in the state where demonstration plot work is being carried out. Such training sessions are usually held in conjunction with field-days and although they do not constitute a strongly structured and carefully super- vised meeting, they are invaluable as a non-formal training exercise. It is estimated that at least 800 field technicians would benefit from such training over the four-year project period. Cooperative Program 4.11 The difficulties which are often encountered when developing the cooperative movement amongst farmers, particularly small farmers, are well known. Brazilian experience has been that one of the keys to success is good management, but it is not easy to identify the management required from the membership of the cooperative, nor is it easy to retain it once found. Various approaches have been tried extensively in the past by various agencies, includ- ing the extension services. One method that has worked is for management to be provided on secondment to a particular cooperative by the extension service. Currently 404 cooperatives are being assisted in a total of 21 states, 307 of  - 29 - which (or 76%) are in the project priority area and the remaining 96 (24%) in the non-priority area. The proposed project would increase this number to 624. The nlumber of farmers in each cooperative varies greatly but using an average of 250 per unit, assistance would be provided to an additional 47,500 families in the priority area and 7,500 in the non-priority area. EMBRATER will supply the administrative assistance of one manager and one accountant for each of the 220 proposed new cooperatives. Incremental staff for the program would number 460. The administration and supervision of this staff will fall directly under the Operations Department of the EMATER or ASTER concerned, while planning at the EMBRATER level would be the responsibil- ity of the project manager of the Operations Department for national programs. Fishery Program 4.12 For a number of years, dating back to the period of ABCAR administra- tion, the state extension services have been assisting fishermen, particularly those in fishing colonies strung out along the coast. Such assistance is very necessary since surveys have shown that particularly acute poverty exists in the artisanal fishing industry. At present staffing is minimal with only one extension agent for each local office where there is a well established colony of artisanal fishermen. In some cases a home economist is also available to work with the families of the fishermen. Local administration and supervision of these two-member teams comes directly under the coordinator of the local extension office. Under the expanded program, financed under the project, administration and supervision would remain as above, but staff would be in- creased to six as follows: extension agent, home economist, fisheries special- ist, plus a management team comprising a manager, accountant and administrative assistant. A further function of this latter three-man team would be to encourage the colony to form a cooperative. Total incremental staff for the program would number 150. It is proposed that 20 colonies would benefit under this program with regional distribution as follows: North five; Northeast ten; Southeast two; South three. Assistance for the formation of fishing coopera- tives would be provided to two colonies in the north and eight in the northeast. It is estimated that some 10,770 fishermen could benefit in the north and northeast priority regions plus 7,130 in the non-priority regions of the south and southeast. Library and Documentation Services 4.13 It is currently policy to improve library facilities within the entire SIBRATER organization in order to provide a comprehensive efficiently operated system whereby information and documentation exchange can effectively take place at the national level. Policy and general guidance on methodology is provided by a headquarters unit, Department of Information and Documentation (DID) at EMBRATER. Although the latter is charged with providing general operational guidelines the state units operate their own Documentary Systems Departments. The proposed component for strengthening library and documenta- tion services would provide the necessary library facilities for the receipt, storage and dissemination of appropriate materials for use by all technicians within the extension system. Such services would be standardized at local, state and federal levels to make the information readily available to anyone  - 30 - requesting it within the system. Such services would also be made available to cooperating governmental institutions and to universities, research centers and selected international institutions where interchange agreements can be effective. Additional staff requirements would be 32, while the existing staff of 52 would be upgraded through technical assistance and study programs under the project. Data Systems 4.14 Much of the current information available concerning the agricultural sector is, in fact, generated by the extension agents. However, most of these data, such as crop reporting (i.e., area, yield and production data) are collected for other agencies, such as the government crop storage company (CIBRAZEM) and the Brazilian Food Company (COBAL). Much of this information then passes outside the extension system and is lost. 4.15 The first objective of the reorganized data collection process would be to standardize the methodology used so that it can be brought under national, as opposed to merely regional or state focus. Functional steps such as collec- tion methods, processing analysis, dissemination and storage (data bank) would be systematized within the SIBRATER system. Secondary objectives would be to improve the quality of the service being provided by: (a) increasing accuracy and reliability of the information provided, (b) implementing an interlocking system with other similar and relevant data collecting operations to provide interchange of data, (c) centralizing the storage of data in order to provide rapid assistance to collaborating agencies and for EMBRATER's own planning and administrative needs. To achieve these objectives under the program EMBRATER, and its data processing units, would work closely with the National System for Rural Information (SNIR). 4.16 Currently, at both federal and state levels both the Library- Documentation and Current Data Systems Units are operationally responsible to the Departments of Planning. As there is a certain complementarity of functions between the two units, a formal administrative merging into one documentation unit is contemplated. For the purposes of project planning and financing, however, current data systems have been considered as a separate program. Currently, there are 23 operational units under the SIBRATER system; 22 of these being in the states, with the exception of Sao Paulo State which has its own autonomous system under CATI. The remaining unit is at EMBRATER headquarters, Brazilia. The existing units are classified by size, i.e. eight small, 11 medium and four large units. Staffing at June 1977 comprised 73 staff with an average of two per small unit, three per medium unit and five per large unit. Incremental staff requirements under the project would number 34 for a total staff of 107 at project completion; the corresponding distribution, according to unit size, would be three, five and seven. 4.17 A total of four manyears of technical assistance would be required during the first two years of work program preparation for operations. Such assistance would provide specialized planning for (a) a diagnosis of the requirements of each state (EMATER or associate company), (b) criteria to evaluate both supply and demand function of information in each state, and (c) development of an operational plan based on (a) and (b) above.  - 31 - Audio Visual Aids 4.18 As already noted, the current farmer-contact method used by the extension agent in the field is almost entirely on an intensive "one to one" basis between extension agent and farmer. One of the reasons for this close relationship has been the extensive use of extension agents for the super- vision of credit. As a result, however, farmer contacts per agent are fre- quently very low. Obviously, if contacts are to be increased to the 130-140 range, a different technique for dissemination of agricultural technology must be employed. Key features of the methodology to be used under the project include: (a) A gradual shift from the individual farmer approach to one of group contact by the extension worker; (b) The use of audio visual aids and printed matter, where applicable, to be used to the fullest extent possible; (c) Staff to be trained in the handling and maintenance of the new equipment; (d) A monitoring and evaluation system provided to achieve maximum adaptive response to the system at the farm level. 4.19 Production units for printed matter, films, photographic slides and radio programs would be set up in each state, usually at the central offices of the EMATER or ASTER in the state capitol or territorial headquarters. The administration of such units will be the responsibility of the state (territory) headquarters offices. Additional specialized staff will be required to operate the equipment in the fields of printing, photography, radio programming and audio-visual cassette preparation. Staffing require- ments for the entire project would be 371. These include 234 in the area of publications; 20 for the production of audio visuals; 90 for the prepara- tion and presentation of radio programs and 27 support personnel. Overall administration would be the responsibility of the Program Unit, under the Department of Operations at both the state (EMATER) and federal (EMBRATER) levels. As most of the technical functions required under the program are relatively new to existing extension staff, it will be necessary to train 36 staff members in the fields of printing, photography, radio programming and the production of audio visual cassettes. Twelve of these training courses would be given abroad and would last four months per course. In addition, 24 technicians would receive two months training courses in Brazil. Extension Methodology Studies 4.20 In the two years since assuming control of the extension system network (SIBRATER) at the national level, EMBRATER has not imposed a uniform extension methodology at the field level. Nor should it attempt to do so until many of the problems unique to the rapidly evolving agricultural industry can be adequately studied to provide the most effective methods to be applied in each case. A major factor influencing the continued use of current tech- niques, largely inherited from the previous ABCAR system, has been the reliance  - 32 - on the credit system as a partial source of funding for state extension programs. Another major factor is the multiplicity of programs having differ- ent levels of federal and state priority, usually reflected in the degree of funding from each. The exceedingly wide variation in population density, from a high of 43.9 per square km in the southeast to a low of 1.0 per square km in the north gives some indication of the difficulties involved. The need for methodology studies, and some trial programs, becomes an obvious first step which must be undertaken before the introduction of new extension techniques on a large scale. 4.21 The basic objective of these studies would be to provide the most effective combination of methodology and techniques which would enable limited numbers of field extensionists to reach an increasing number of farmers and their families. As a first step, three states would be chosen where the EMATER is willing to enter into an agreement with EMBRATER for the study program to be executed. Since each state has particular constraints which cause problems for extension services, i.e., based on (a) farm size, (b) land ownership pattern, (c) predominant crops being grown, and (d) relative devel- opment of basic infrastructure, the first essential would be a study of the particular problem in the selected states. This would mainly be a "base-line study" to determine the situation at that time, even before starting the trials of new methodology, to obtain a basis for selecting the new techniques to be tried. Objectives of the study should be clearly stated. This would be followed by a planning period to decide upon the best test methodology to be implemented. The third stage would be the actual implementation of the methodology chosen, including the training of personnel involved. The entire study would normally not encompass a period greater than two years, although monitoring should continue for an additional year. 4.22 EMBRATER management has been advised of the success of the training and visit system in several Bank funded projects in India. Also recently prepared rural development projects in the states of Ceara and Bahia have included a modification of this technique to be applied to the agricultural extension component of these projects. Modification has comprised basically a separation of duties, other than pure extension, to permit the field exten- sion agent to devote his full time to agricultural extension. It is believed that this could be an important first step in separating the credit function from the extension service. It is too early to evaluate these trials in Brazil but EMBRATER has shown a willingness to collaborate with state EMATERs in these efforts. Also under the project, the availability of funding for specialized short courses would permit key extension staff to view this, and other systems, in countries such as India and Mexico where innovation in agri- cultural extension methods is taking place. 4.23 The administration of these studies, including planning, execution, monitoring and evaluation would be the responsibility of EMBRATER. However, they would work closely with the EMATER concerned, based upon a plan to be mutually approved, before the study subproject is initiated. Administrative and field staff (10) would be provided by the Planning Department of EMBRATER and key technical staff, mainly consultants engaged for the study period only, would also be supplied through EMBRATER. All field staff for execution of the  - 33 - project would be supplied by the local EMATER through the appropriate depart- ment of planning and would be extension staff as normally assigned to the particular study area. 4.24 A total of ten manyears of consultants would be made available for the planning, monitoring and evaluation functions of the methods selected. Such technical assistance would be administered through EMBRATER. Assignments would tend to be long term, for the first year, or planning phase, but could be adjusted to short term for the monitoring and evaluation phases. Distri- bution would be for four manyears at the federal (EMBRATER) level and six manyears at the state (EMATER) level. In addition, funds would be made avail- able to contract university teams for specific studies. These would take place during the first three years of project execution and would be based on a detailed contractual arrangement between EMBRATER and the university concerned. Specific study plans would be prepared by EMBRATER and discussed with the university staff concerned prior to the signing of the contract. 4.25 Planning and base line studies would be carried out by EMATER personnel under the direction of EMBRATER staff, responsible to the Planning Departments in each case. Monitoring would be carried out by EMBRATER staff, at previously designated intervals and, after the basic two-year study, for an additional year. The local staff would participate by providing the neces- sary field data. Evaluation would be carried out at the end of the test period (several agricultural seasons) and would be carried out by the groups involved in the study, i.e., representatives from farmers/producers, research personnel, extension personnel (state and federal level) and the consultants engaged for the specific study. C. Staff Training and Technical Assistance Staff Training 4.26 Long- and Short-Term Fellowships. A total of 210 employees of the SIBRATER system (EMBRATER/EMATERs) would be trained at post graduate university level during the project period. Of these 174 would study for the MSc degree, mainly in areas designed to provide the increased number of subject matter specialists under the project, i.e., crop and animal husbandry, soil conser- vation, food storage, cooperatives and inland fisheries. A few would also study administration in the areas of library, documentation and current data processing. Of the 174 candidates 99 would study abroad and 75 in Brazil. In addition 36 candidates would receive post-graduate degree fellowships for study at the PhD level. Of these (PhD) fellowship programs, 10 would be abroad and 26 in Brazil. These staff members, on return from university, would fill positions as program specialists, and in advisory, planning and liaison roles between EMBRATER and the EMATERs at the state level for the evaluation of programs and would work closely with the research organization (EMBRATER) in the interpretation of government policy concerning specific crops. Doctoral candidates (three) would also provide inputs to methodology studies, monitoring and evaluation at the federal level. As under currently established fellow- ship programs, candidates would be required to sign a contract with EMBRATER  - 34 - stating their willingness to work for the institution on completion of studies for a period equal to that of the fellowship provided. 4.27 Specialized Courses. A total of 381 staff members, 256 abroad and 125 in Brazil, would attend short courses of four months and two months respec- tively. Courses would be selected principally in the areas of support services a' t-o train personnel in the use of new equipment, i.e., printing and litho- graDn, technical photography, audio-visual programming, data processing, docu- mentation, cooperative management and simple food storage techniques for both crops and fish. 4.28 Pre-service Training. This training would be carried out under the administration of the Human Resources Department of EMBRATER. Pre-service courses of 30 work days would be provided for all incoming (incremental) staff and would thus provide training for 4,725 students, allowing for the tradi- tional 30% "drop out" factor, over the four-year project period. In this way total incremental staff would be 3,634 as shown. Courses are designed for 40 students and capacity at the eight training centers to be provided under the project would absorb 50% (six months) training program capacity at each of these centers, i.e., courses would be provided at approximately two-month intervals. 4.29 In-service Training. This would comprise essentially refresher courses of 15 days' duration each year. Following a general updating of exten- sion methodology currently in use the curriculum would be more flexible as compared to the fixed course content of the pre-service training exercise. Study subjects would be designed to cover problem areas in current programs, i.e., crops, livestock, credit, cooperatives, fisheries, current data docu- mentation and communications, according to the requirements of (a) the region, (b) agents selected, and (c) policy and priority directives pertaining to the specific area of operations. The training officers of the Human Resources Unit, under the Department of Program Development at the federal and state levels, would be responsible for curriculum development of these courses. By year four of the project, it is estimated that 13,500 participants (8,000 technicians and 5,500 administrative and support staff) would participate in in-service training courses each year. Courses for the 8,000 technicians would occupy the eight training center facilities for six months of the year (10 courses of 80 students each) and courses for administrative personnel would cover subjects such as personnel management, accounting and finance, organiza- tion, marketing, data gathering and retrieval systems, etc., would be handled either at federal or state headquarters facilities or in cooperation with local schools or facilities providing business administration or communications courses, as the case may be. D. Monitoring and Evaluation 4.30 Monitoring. Overall monitoring and internal reporting of physical and financial inputs under the SIBRATER system is the responsibility of respective monitoring and evaluation units, and the financial accounting units, under the departments of Planning, Administration and Finance, respectively,  - 35 - at both state (EMATER) and federal (EMBRATER) levels. The unit responsible for complying with reporting requirements to the Bank and for handling all queries concerning project implementation would be the Technical Support Unit of the Operations Department, at the federal level (see para 4.03 and Annex 3, Charts 1 and 2). The project coordinator in the latter unit would work closely with the departments concerned to discriminate program components under the Agricultural Extension I Project from other programs within SIBRATER's general operations. 4.31 Evaluation. Effective evaluation requires continuous and accurate data gathering. Some project components, such as the documentation and current data systems, are expected to contribute to the effectiveness of project eval- uation. Also, under the project the Monitoring and Evaluation Units at state and federal levels are to be strengthened with additional personnel and requisite training courses. Uniformity of data reporting, at present faulty, would be strengthened by the programs under the documentation unit mentioned above. The evaluation process is one which is essential to the entire SIBRATER system and is currently being evaluated by EMBRATER administration. The focus- ing of Bank loan related programs within the SIBRATER system would be the basis for reporting procedures to the Bank and the responsibility of the project coordinator (see para 4.03). The project coordinator would also assume prime responsibility for the project evaluation report that would be required by the Bank upon project completion. An assurance was obtained during negotiations that a project completion and evaluation report be pro- vided to the Bank within six months of project completion. This exercise would form an integral part, and would not be alien to, the normal monitoring and evaluation process for the project. Detailed discussion of the Monitoring and Evaluation program to be implemented is provided in Annex 1. 4.32 Technical Assistance. Although the basic monitoring and evaluation units are in place, both at the state and federal levels, there is a need for greater integration of the two levels so that comparative evaluations can be made. As a first step uniformity in data collection procedure is essential and the key questions as to what, when and where data will be collected, and over what time period can only be answered after a uniform model for analysis and evaluation has been prepared. In order to assist in effecting the neces- sary changes, under the current administrative model, the project would finance consultants to deal specifically with monitoring and management pro- cedures for the entire SIBRATER system, of which project related activities form an integral part. Six consultant manyears would be allotted for this purpose, under the general technical assistance program (para. 3.09). The cost of such consultancy services, for both Brazilian and expatriate consul- tants is US$75,000 per manyear. 4.33 Civil Works Implementation. To ensure that the civil works program is adequately supervised and that implementation is commenced at an early stage in the project cycle, the Bank obtained assurances at negotiations that EMBRATER would: (a) engage the consultancy services of a competent three-man civil engineering/architect team to work with EMBRATER's Technical Support Unit (TSU) in the review of site development plans, detailed con- struction plans and review of contract bids for the civil works component; and  - 36 - (b) prior to August 1, 1978, present the site plans for construction projects to be carried out during the remainder of calendar 1978. Such site plans would include those for Training Centers, in pre-selected states; plus sites for state EMATER headquarters offices and buildings required under the "Artisan" Fishery Colonies program, in states yet to be designated. EMBRATER would also submit an annual (calendar year) civil works building program plus site plans no later than November 30 of the previous year for such activities in project years two, three and four. In the case of civil works involving the renovation of, or addition to, existing buildings EMBRATER would ensure that the EMATER or ASTER involved would seek, through EMBRATER, the approval of the Bank for such building plans before inviting bids in the proposed construction con- tracts. Assurances to this effect were obtained at negotiations. The pro- posed construction schedule under the civil works program is provided at Annex 3 Chart 4. Conditions of disbursement of funds against the civil works component, would be that: (a) EMBRATER had furnished evidence satisfactory to the Bank that it or the EMATER/ASTER concerned has legal title or other ade- quate right to the land on which it is proposed to construct project facilities (b) consultants had been retained and would be available for supervision of the implementation of the civil works component during the period of project execution. Assurances to this effect were also obtained at negotiations. 4.34 Project Goals and Implementation Schedules. A major objective of the project is to strengthen existing extension services through the program approach (4.07 to 4.12). There is also a need to adopt new methodology, as may be necessary and based on local conditions, to ensure that the individual field extension agent is increasingly relieved of duties pertaining to credit supervision and health and social welfare programs. In order to ensure an orderly and balanced development of the programs to be financed under the project, assurances were obtained at negotiations that EMBRATER would prepare and furnish to the Bank for review and comment: (a) by June 30 of each year appropriate guidelines regarding priority programs, regions and target groups for use by the EMATERs/ASTERs in connection with the preparation of their annual operational budget; (b) by November 30 of each year the proposed programs (PROATERs) and related budgets of the EMATERs/ASTERs for the follow- ing year. Assurances were also obtained that EMBRATER would: (a) by June 30, 1980, or such later date as may be mutually agreed by EMBRATER and the Bank, review with the Bank progress made under the project and the extent to which project objectives are being achieved; and (b) introduce such changes in project content or mode of implementation as the Bank shall reasonably request in the light of such review. A table showing physical goals to be achieved to complete project implementation within the four-year period is provided at Annex 3 Table 20. A project implementation schedule by project years is given at Annex 3, Chart 5. A summary of program objectives at full implementation is is provided at Annex 3, Chart 6.  - 37 - V. PRODUCTION AND TECHNOLOGY Impact on Production 5.01 Experiince in other countries has shown that providing extension services as visualized under the project should have, over a period of time, considerable impact on total national agricultural production. How- ever, since a wide variety of crops is grown in Brazil over a very wide range of physical circumstances by many different types of farmers, it is virtually impossible to quantify and detail the extent of the increased pro- duction. The efficiency of production should also be improved as a result of the adoption of improved technology by farmers but this too is difficult to quantify. To the extent possible the increases in production that should be achieved as a result of the project are mentioned in the section dealing with economic analysis (paras 7.04-7.06). Impact on the Extension Services 5.02 The project would have an appreciable impact on the country's extension services. It would encourage greater efficiency and output per extensionist, in part due to more emphasis being placed on group methods of extension and in part due to greatly improved back-up services. In addition, the proposed methodology studies would test and evaluate new technologies related to the main task of information dissemination and rural education. Although there are wide variations according to region, each extension worker currently contacts about 80 farm families, or a total of 565,000 families. Under the project it is proposed that, by use of additional facilities and new technology each agent should be able to contact an average of 140 families or a 75% increase in efficiency. Based on these assumptions, the proposed increment of 34% (2,419) technical field personnel to the total SIBRATER staff would show a corresponding increase of approximately 130% in the number of farmers contacted at year five of project, thereby providing assistance to a total of 1.3 million farm families. The project at full development would provide employment to an additional 2,692 middle-level (high school graduate) technicians, and in addition 1,042 support and administrative staff would be employed for a total labor increment of 3,734. The technical competency of the total 18,500 staff of the SIBRATER system would also be upgraded. The productive capacity of some 1.3 million families would also be enhanced by technology transfer and rural (farmer/worker) educational programs. The availability of local farm related skills is being improved under the SENAR program being partially funded by the Bank. 1/ This project is expected to benefit 309,000 rural workers by 1981 and would be complementary to the extension services. Selected local extension staff would be called upon to teach several of the courses related to agricultural skills. The upgrading of extension staff under the proposed project would, therefore, provide additional benefits to the Vocational Training Project. Impact at Farm Level 5.03 Based on number of farm properties in Brazil, 4.9 million, the rela- tive number of agricultural holdings receiving technical assistance under the 1/ Appraisal of a Vocational Training Project, Brazil, May 26, 1977.  - 38 - project would rise from a current 11.0% to 26.5% of the total. Quite apart from this extended coverage to be achieved by the strengthening of the national agricultural extension system (SIBRATER) on a national scale, there are several direct, and indirect benefits which will accrue as the programs become functional. These are changes directly affecting agricul- tural production, which would result from: (a) adaptation and adoption of new technology, produced by the research branch (EMBRAPA) and applied at the farmer (producer) level; (b) increases in yields per unit area which, apart from the direct economic effect, would permit more flexibility in the requirements for infrastructure and rural development objectives which are inherent in programs which merely expand the area being cultivated; (c) the extended use of farming systems demonstration plots would not only help to raise farm incomes but affect conservation measures so urgen-ily needed in some areas; (d) the extended use of important inputs, auxiliary -o more intensive production methods, such as improved seed, fertilizer, pesticides and insec- ticides, and the rational use of credit would further stimulate the national economy. 5.04 Several changes would take place due to the ability to both gather and disseminate information more quickly under programs such as documentation, audio visual and current data collection and processing systems. This would achieve the following: (a) greatly enhance the opportunities to inform the rural population of opportunities to improve their basic skills and standard of living through education, home economics, health care, nutrition and other socially oriented programs; (b) the current data provided, on crop prospects, markets, prices, weather and federal/state programs affecting specific crop production programs would also flow from these various aspects of data collection and dissemination. In addition, the overall contribution, through the interaction of increased staff capability, operating effectiveness, and well-equipped programs with adequate financial support at both federal and state levels would provide the federal and state institutions which are respon- sible for the performance of the agricultural sector, not only with the means to stimulate production but also with the current data flow necessary to enable them to make policy decisions affecting the agricultural sector. Such information on crop production, storage and marketing would provide the know- ledge necessary to successfully integrate rural development programs within the broad policy objectives of economic policy guidelines at the federal level. Such Government policy, that of providing specific economic and social welfare to the rural poor, would clearly benefit from the execution of the proposed extension programs under the project. Environmental Impact 5.05 During the current decade, and particularly during the last five years, Government has been criticized in international forums for neglect- ing the preservation of the natural habitat of the Amazon basin, and of the possible rapid and irreversible damage which can accrue to the delicate balance held by the biomass of the Amazonian forest complex. The Brazilian Government has responded, in a positive manner, to such criticism and many of the problems of rapid expansion along the controversial "Itransamazonica" highway have been addressed through control legislation by Government and implemented through the National Research Council (CNPq) and the National Institute for Colonization and Agrarian Reform (INCRA).  - 39 - 5.06 In addition the need for natural resource inventories, not only for the Amazon region (North) but for the other two regions currently selected as priority areas in the proposed agricultural extension project i.e., Northeast and Central-West, was given special attention in the preparation of the Agricultural Research I Project. 1/ Under the research program special area studies have been set up at specific natural resource research centers in the three areas mentioned. These are located at Belem, Petrolina and Planaltina and the research carried out at these stations and several sub- stations throughout the area has specific emphasis on the development of viable farming systems designed to conserve, rather than destroy, the ecosystems in which they are located. 5.07 Based on the premise that the technology developed at these stations will be successfully disseminated to the farmers in the three regions, it is clear that the environmental impact of the proposed project can have a positive effect. It cannot be assumed that the inevitable competition between produc- tion on the one hand and conservation on the other will always be decided on the merits of the latter. However, it has frequently been shown that the shifting cultivation cycle, so destructive of virgin forest, can be broken. In addition, the increase of production per unit area means that less land need be brought under cultivation. Improved farming techniques, covering a wide range from soil conservation, crop rotation and rational water use (irrigation and control of damaging water flow) all serve to preserve rather than destroy the ecological balance which is essential to increased agricul- tural production, and agronomic stability, of the cropping systems to be utilized on a long-term basis. 5.08 The relative success of the agricultural extension system in preserving the environment at the agricultural frontier area will undoubtedly depend on many factors, including official policy at the state and federal level. However, the concepts of good agricultural husbandry, embodied in the project goals, are entirely favorable from the viewpoint of conservation of the environment in this case. 1/ See Brazil: Agricultural Research I Project, Report No. 103la-BR, 1976.  - 40 - VI. DEMAND AND MARKET PROSPECTS Introduction 6.01 In a project of this nature where the goal is to increase the general productivity and economic efficiency of farming operations, consis- tent with Government planning and agricultural policy initiatives, the project is as likely to be influenced as much by national policy directives as by the more direct effects of the market place, be it local or inter- national in scope. Influence of Policy Guidelines 6.02 Project goals would be in direct response to Government agri- cultural development policy guidelines which are based on macro-economic considerations of (a) providing sufficient basic food items, at reason- able prices for a population of approximately 107.0 million, which is increasing at an annual rate of 2.9% and (b) maintaining, or even rein- forcing its predominant position as an exporter of agricultural produce on the world market. Currently this has been noted to contribute about 60% of the agricultural GNP and to be increasing at a significant rate each year. To maintain an adequate basic food supply for the nation, reduce costly imports, such as wheat, and to increase its share as a supplier of raw food and fish products in overseas markets, the Government must rely heavily on the national agricultural research and extension systems to assist farmers in reaching the annual goal of approximately a 7% increase in the GNP of the agricultural sector. Influence on Crop Production 6.03 Since the national extension system is designed to provide programs leading to farm-level production improvements over a very wide range of crops (currently in excess of 38 under specific programs) its influence on any one crop is not large although, in the aggregate, the effect on the overall rural economy can be shown to be very significant. The production goals set for both the agricultural research and extension services are well within the economic guidelines which Government has traced for the role which agriculture will play in the current (1975-79) and near future (1980-1984) planning periods. It seems unlikely, however, that any one crop will greatly exceed the production goals set since government policy is sensitive to both internal consumption and export demand trends and can readjust annual production targets accordingly. Indeed it has been pointed out (para. 1.11) that official policy has a strong regulatory influence on both agricultural production and marketing. Many of these regulations are implemented through directives to EMBRATER. Influence on Marketing 6.04 Many of the same principles, outlined in paragraph 6.03 above, apply here, in the area of regulatory prices, storage and marketing schemes for basic domestic food crops, e.g. beans, rice and export crops such as coffee, soy- beans, sugar. Here also, the extension system (SIBRATER) is one of the main agencies used to apply such regulations which frequently remove such commodi- ties both from free internal markets, at the state level, and in the world market for products such as coffee and sugar.  - 41 - 6.05 In general the project must be viewed as one providing a much needed stimulus to agricultural production, with specific small farmer emphasis but the influence of which, on demand and supply factors, is very much governed a priori, at the Government economic planning level, by policy measures issued through the Ministry of Agriculture. Assuming that such policies will continue, the improved technical capability of the national extension system should tend to mitigate in favor of market stability in the agricultural sector. VII. BENEFITS AND JUSTIFICATION A. Beneficiaries and Target Population Groups Beneficiaries 7.01 The -direct beneficiaries of the project would be the additional 735,000 farmers receiving extension services for the first time. All farmers, and farm labor, represented by the total of 1.3 million properties, which would receive improved agricultural extension service under the project, would also benefit to the extent that target goals are substantially achieved. In- direct benefits would also occur due to the improved services provided by the extension system, in the area of living standards and general welfare through the programs which are required for development of the rural areas and through which many direct social benefits accrue to the rural population in general. The extension system (SIBRATER) acts as an important catalyst in the areas of education, health nutrition and home economics in such areas of new development. Target Population Groups 7.02 Although agricultural extension services must assist all levels of production, there is opportunity for target population emphasis, particularly programs emphasizing the small farmer. This is possible because the larger farmer can often avail himself of new technology either direct from the ex- periment station or through visits to regional headquarters units of EMBRATER. For this reason, on a cost input basis, the proposed project would select programs designed to assist small farmers in an approximate ratio of two to one in favor of the latter, i.e., 66% of program and investment emphasis would be targeted towards the small farmer group. B. Cost Recovery 7.03 The most important sources of returns under the project to the state and federal governments are respectively the state value added tax (ICM), and the federal tax for assistance to rural workers (FUNRURAL). It is difficult to estimate the extent to which project costs would be recovered because exemptions from the ICM are given at different times as incentives to certain groups of pro- ducers (i.e., for certain crops or in certain regions), and state receipts from  - 42 - the ICM vary accordingly. Assuming recovery of 50% of the ICM tax and 100% of the FUNRURAL tax attributable to incremental production from the project, the approximate average annual tax yield would be about US$27.0 million at the state level and about US$9.0 million at the federal level. C. Economic Analysis 7.04 The design and scope of the project, the selection and relative size of project components and the organizational arrangements for project imple- mentation are based on current experience in the implementation of a similar project to strengthen the agricultural research service in Brazil. However, even while assuming a rational degree of successful project i4mplementation, the method of estimating a level of economic benefits to this -yp-e of (institutional strengthening) project still involves a large degree of assumption and imputation since many additional inputs are required before the full productive potential of new agricultural technology can be realized. 7.05 In calculating an estimated economic rate of return for the project two basic assumptions have been made. These are (a) that the acceptance rate of new technology by farmers would be slow, reaching no more than 15% in year three of the project and not increasing above 50% of the target population by year 12, or full project development; the figures, although admittedly conservative, have some basis of comparison from a closely monitored experiment in Mexico; 1/ and (b) it was assumed that increased production would follow the (existing) normal curve up to project year two after which a 3% increase in annual yields due to new technology would occur in the agricultural production curve. Border prices, calculated from November 1977 data, were used for all crops except for those commodities produced mainly for home consumption where constant farm gate prices were used. The cost stream used included 52.7% of all project incremental costs which is based upon the relative value of field crops in the project area and upon which this analysis is based. The life of the project was assumed to be 12 years with no residual value attri- buted after that period. 7.06 Based on these assumptions the internal economic rate of return was found to be in the range of 11 to 18% which reflects annual incremental crop yields, attributable to extension, in the range of from 2.5 to 3%. An internal economic rate of return of about 14% is considered most likely and this compares favorably with the estimated opportunity cost of capital of about 11%. See also Annex 2 for details of cost and benefit stream calculations. D. Social Benefits 7.07 In the past decade the emphasis in official government policy towards agriculture has been aimed primarily at achieving agricultural production through a rapid expansion of the area being cultivated. For the past few years, however, there has been a growing awareness, and official emphasis, on problems of a social nature in rural areas and particularly the need to raise 1/ The Puebla Project, Seven Years of Experience: 1967-73. CIMMYT, Mexico.  - 43 - che income of the small farmer and the rural poor who work in agriculture but have limited land resources. Agricultural price policies, and subsidized farm inputs, tend to benefit greater numbers of the medium and large farmers, and proportionately, the services of the extension agents, in implementing such policies, have favored the latter groups. Several programs, under the proposed project, are designed to alter this bias more in favor of the small farmer through demonstration field plots, simple farming systems demonstrations and formation of cooperatives which would focus on better farm management tech- niques, the use of low cost inputs and on rural health and welfare services designed to improve the standard of living for the rural poor. E. Project Risks 7.08 The quantifiable economic and social benefits are highly dependent upon a series of assumptions concerning other essential inputs related to production (para 7.04). The administration of EMBRATER, however, will have no control over many of these factors which are governed by national agri- cultural policies. Through the coordinating council (COMPATER), however, it can make known when the constraints to rational production at the field level are likely to occur. 7.09 Apart from the above the most obvious risks which project implemen- tation would face are as follows: (a) the administration of EMBRATER, still a comparatively new organization, faces a formidable task of strengthening its existing leadership, training new and existing staff and dealing with new programs within a relatively short period. It is expected that during the first two years of project life intensive supervision would be required both on the part of EMBRATER and EMATER administrations. (b) the degree to which the main function of the extension worker i.e., communication of new ideas, depends upon the educational level and receptivity of his audience to accept these ideas. The group method approach, which must be recommended to obtain greater coverage, can exacerbate such difficulties. (c) extension is dependent upon agricultural research to provide the bulk of new agro-technology to be passed on to the farmer. Active cooperation between both groups is essential. Currently, the risk in this particular area would appear to be minimal as good cooperation has been established in most states. (d) Price fluctuations on international markets and government pricing policies do not always favor the producer as wider social benefits are sought for other segments of the economy. The degree to which the Ministry of Agriculture can influence Government policies in this sphere will have a direct effect on the returns to agriculture as a whole.  - 44 - VIII. SUMMARY OF AGREEMENTS REACHED AND RECOMMENDATIONS 8.01 Assurance were obtained at negotiations that: (a) The Federal Government would provide, in a timely manner, the funds necessary for EMBRATER to execute the project and also ensure that the contractual arrangements between EMBRATER and its associates in- clude a clause whereby the associate (EMATER or ASTER) gives an annual assurance to EMBRATER to provide the necessary funding to execute its share of the project based on annual budgets to be approved by EMBRATER, no later than October, for execution during the following EMBRATER fiscal year (para 3.12); (b) EMBRATER would: (i) agree to engage the services of a competent agency or agencies to administer that portion of the technical assistance compo- nent which requires (a) the employment of suitably qualified expatriate consultants under terms of reference satisfactory to the Bank, and (b) the making of arrangements for fellowship study programs at universities outside Brazil (para 3.15); (ii) agree to maintain separate accounts to indicate all expendi- tures made under the project and undertake that such accounts be audited annually by an independent auditor acceptable to Bank, in accordance with sound auditing principles, consistently applied. Such audited accounts would be submitted promptly to the Bank but no later than four months after the closing of accounts for the previous year (para 3.18); (iii) provide the Bank every six months and in a timely manner a comprehensive account of progress in project execution, based on the reports of EMBRATER's existing monitoring unit for operation of the extension system (para 4.03); (iv) agree to prepare and submit to the Bank no later than six months after project completion a project completion report as a summation of the monitoring and evaluation activity for the project (para 4.31); (v) to ensure that the civil works program is adequately super- vised and that implementation is commenced at an early stage in the project cycle EMBRATER would: (a) engage the services of consultants to work with EMBRATER's Technical Support Unit (TSU) in the review of site development plans, detailed con- struction plans and review of contract bids for the civil works component; (b) prior to August 1, 1978, present the site plans for construction projects to be carried out during the remainder of calendar 1978 and (c) by November 30 of each year,  - 45 - the site development plans for the civil works construction program under the project to be carried out during the following year; such site plans would include those for Training Centers in preselected states plus sites for state EMATER headquarters offices and buildings required under the 'Artisan' Fishery Colonies program; and (d) in the case of civil works involving the renovation of, or addition to, existing buildings EMBRATER would ensure that the EMATER or ASTER involved would seek through EMBRATER, the approval of the Bank for such building plans be- fore inviting bids on the proposed construction contracts (4.33); (vi) furnish to the Bank for review and comment (a) by June 30 of each year, appropriate guidelines regarding incremental staffing requirements priority programs, regions and target groups for use by the EMATERs/ASTERs in connection with the preparation of their annual operational budgets; and (b) by November 30 of each year, the proposed programs (PROATERs) and related budgets of the EMATERs/ASTERs for the following year (4.34); (vii) by June 30, 1980, or such later date as may be mutually agreed to by EMBRATER and the Bank, EMBRATER would review with the Bank progress made under the project and the extent to which project objectives are being achieved and introduce such changes in project content or mode of implementation as the Bank shall reasonably request in the light of such review (para 4.34). 8.02 Conditions of disbursement would be that: (i) contractual arrangements satisfactory to the Bank had been entered into between EMBRATER and its associates under which each EMATER/ ASTER would agree, inter alia, to submit to EMBRATER for approval by October 31 of each year its annual budget for the following year, which budgets would include adequate allocations from the state to cover expenditures required for the project and not covered from other sources, conclusion of such arrangements being a condition of disbursement in respect of expenditures by the EMATER/ASTER concerned (para 3.12), and (ii) in the case of civil works contracts that (a) EMBRATER had furn- ished evidence satisfactory to the Bank that it, or the EMATER/ASTER concerned, has legal title or other adequate rights to the land on which it is proposed to construct project facilities; and (b) con- sultants had been retained and would be available for supervision of the implementation of the civil works component during the period of project execution (para 4.33). Recommendation 8.03 With the above assurances the proposed project would constitute a suitable basis for a loan to the Federative Government of Brazil for the equiva- lent of US$100 million. The loan would have a term of 15 years, including three years of grace, with interest at the rate prevailing at the time of distribution of the documents to the executive directors. May 3, 1978  ANNEX 1 BRAZIL AGRICULTURAL EXTENSION I PROJECT Monitoring and Evaluation 1. Introduction. In order to insure that project goals and objectives are being met, throughout project life, an effective system of project monitoring and evaluation to verify such measurable project objectives is essential. The responsible coordinating federal agency for the project (EMBRATER) and the state executing agencies for rural extension (EMATERS) are fortunate in that, administratively, both include monitoring and evaluation units which are responsible to the department of planning in each case. However, to insure that the monitoring and evaluation process is an effective management tool for both the federal and state institutions, there is still much to be done to unify the methodology, criteria and analysis procedures at the state and federal level. The need for greater integration in this sphere has already been discussed with EMBRATER officials and the details would be agreed upon and finalized during negotiations. 2. Data Collection. The first essential step would be to unify and standardize the data collection system. The simple questions of what, why, how and for whom current data are essential are still to be answered before data collection operations can be adequately coordinated at all levels. Since current data systems are a program being provided with specific inupts and emphasis within the proposed project, it seems fairly certain that such goals can be achieved at an early stage. 3. Internal Reporting. The data generating system within the EMATERS must provide the necessary reporting mechanism for both state and federal administrations. This process corresponds to the so-called internal reporting procedure. This must be supplemented by a set of key data essential for specific monitoring and project evaluation functions based on carefully selected and pre-determined indicators. It is in this area of external monitoring and evaluation where the services (6 manyears) of adminis- tration management consultants, specialized in this field, would be engaged under the project. The terms of reference for such consultants would be the basis for the methodology selected to provide the information desired. As a first criteria the current reporting system would be strengthened and unified to facilitate the accurate and rapid flow of desired management information. Accounting procedures are already established to record expenditures of the financial resources to be provided under the project. The supervision of field staff does not reflect the same unified appraoch towards the recording of physical goals to be achieved under the current system. Indeed with so many interlocking operational programs, and where field staff are responsible for aspects of several of these programs in their area, it is here that adequate management procedures must be perfected to supply adequate supervision without having an adverse affect on the actual efficiency and morale of regional and field level staff.  - 47 - ANNEX 1 Page 2 4. External Reporting. Although the accuracy of data available at the federal level depends heavily on inputs from the state units, the capability and corresponding facilities of the federal coordinating unit (EMBRATER) are much greater and are not expected to present logistical problems. All programs have an administrator (manager) at the federal level and they are currently well informed as to the progress of programs under their responsibility. The selection of meaningful key indicators and optimum sample size are matters still not well defined and here also the technical (consultancy) services to be provided under project financing can play a meaningful role. EMBRATER headquarters unit (Brasilia) would be responsible for supply data to other interested Federal Agencies, or institutions abroad, as the case may be. 5. Surveys and Ad Hoc Studies. Occasionally, such as in the case of rural development programs where social impact on the target population is of specific interest, it may be preferable to contract specific studies to an appropriate university or other institute of advanced studies in the area to provide the necessary manpower inputs. This also applies to direct studies of alternative extension methodology and here again consultancy services (10 man- years) would be provided for this purpose. Such studies might include: 1. Methods for effective group contacts of farmers by extension workers; 2. Effective use of audiovisual aids; 3. The relative quality of training curricula both for extensionists and farmers; 4. Village case studies to measure the impact of extension services on social and economic welfare of the family group; 5. Review of results achieved by specific programs oriented towards predetermined target groups. 6. Design and Planning for Efficient Monitoring and Evaluation. This would be the responsibility of the monitoring and evaluation unit, working under the leadership of the Planning Department. The latter would plan the necessary coordination of the state units. Initial inputs, and timely feedback from farmers, producers and their families would be primarily the function of field extensionists and social workers who would be contacted to provide such information through the established Current Data system. Staff members of the Monitoring and Evaluation Unit at the state level would be responsible for this task. 7. Reporting Procedures. Under the proposed project to be partially funded by a Bank loan, assurances would be received at negotiations that progress reports, in a format satisfactory to the Bank, would be submitted at six-month intervals. Such reports would deal with progress under the proposed "Agricultural Extension I Project" but would be compiled from data  - 48 - ANNEX 1 Page 3 monitoring processes being implemented to cover the entire SIBRATER system. Under accounting procedures it-would be necessary to establish a system where project costs can be discriminated from general costs of the national system. In the case of key indicators to monitor progress towards specific project goals it would be the responsibility of the consultants, being provided under the project, to elaborate suitable reporting procedures which would be of common interest to EMATER and EMBRATER managements, as well as being of specific interest to the Bank. It is essential that the monitoring system adopted provide overall control of the SIBRATER system and is not merely project specific. In addition, the monitoring and evaluation unit would prepare, within six months of project completion, the necessary "Project Completion Report" as requested by Bank under the conditions of the loan.  - 49 - ANNEX 2 BRAZIL AGRICULTURAL EXTENSION I PROJECT ECONOMIC AND SOCIAL BENEFITS A. General 1. The purpose of agricultural extension is to disseminate new tech- niques and information on improved agronomic practices to the farmers. As improved extension services generally tend, ceteris paribus, to increase output, ideally the expanded transfer of improved farming practices should both increase the net income of the farmer as well as provide a larger stock of lower cost foods to the consuming public. To the extent that the technology transfer is directed, especially toward lower income farmers, it should also have a positive effect in improving distribution of income. While these types of benefits are difficult to quantify, an attempt was made in this project to estimate only the increased production which might be attributed mainly to the "improved technology" impact of the extension project and, hence, the return to the planned investment. 2. It is realized that any attempt to attribute specific benefits to an extension project cannot be precise, since cumulative increases in produc- tivity would also require, apart from adequate ecological and climatic condi- tions and acceptance by farmers, a number of complementary investments in other fields such as research, credit, physical and social infrastructure and the like. As a basis for the calculations, it was decided to use the growth in field crops 1/ the productivity of which can be attributed only to extension as a proxy, or indirect measure of the benefits of extension services. These yearly increases in productivity, attributable to improved extension services, would be incremental to the present average yearly increment in GNP of the Brazilian agricultural sector which is 7% (4% in the area under cultivation and 3% from increase in productivity). The major variables and assumptions determining benefits which may accrue from an agricultural extension service are given in the following. B. Measurement of Project Benefits 3. Technology Acceptance by Farmers: Although expected benefits to improved agricultural extension activities rely heavily on the degree of acceptance of the new agrotechnologies by the farmer, it is in fact very difficult to estimate such acceptance rates with any degree of reliability. 1/ Including only field crops, the 1976 value of which accounted for 52.7% of all agricultural output.  - 50 - ANNEX 2 Page 2 It has been noted that in the case of India 1/ estimated acceptance rates were much higher than those for a well documented experiment in the Puebla region of Mexico. 2/ Since the conditions prevailing in Mexico appear more comparable to general experience in this hemisphere, for both agricultural research and extension activities, it is expected that the more conservative rates of accep- tance used in the present analysis could be easily reached, given the higher literacy level and degree of technical sophistication in much of the project area. In this case, it has been conservatively assumed that only 50% of the target population of 1,300,000 farmers will adopt new agrotechnologies by year 12 of the project, and hence new technologies would be implemented in approx- imately 50% of the area encompassed by the project. 4. The acceptance of the producers will be higher during the first years of the project, and due to diminishing returns lower thereafter, if we assume the present levels of research technology and extension services currently visualized. From project year 12 and onwards, no new acceptances are expected to take place. Estimated acceptance would accrue as follows: Percentages of Target Population Adopting New Agrotechnologies /1 Project Yearly Cumulative Year Percentage Percentage I 0 0 2 10 10 3 5 15 4 5 20 5 5 25 6 5 30 7 5 35 8 5 40 9 3 43 10 3 46 11 3 49 12 1 50 13 .and onwards 0 50 /1 These percentages were also used for computing areas implementing new agrotechnologies (see Tables 5.1 and 5.2). 5. Increased Production; It has been assumed that the extension proj- ect will be the cause of a shift in the production curve of those farmers accepting new technology. Due to the gradual impact of agricultural extension 1/ India: Appraisal of the Bihar Agricultural Extension and Research Project, Report No. 1753-IN, October 1977. 2/ The Puebla Project, Seven Years of Experience; 1976-73, CIMMYT, Mexico.  - 51 - ANNEX 2 Page 3 services, one year lag has been assumed between the beginning of the project and its significant effects, on agricultural production. Commencing in project year two, a shift in the aggregated agricultural production curve should occur; this shift will correspond to a shift in the production func- tions, i.e., in the productivity of all crops grown by project participants. The rate of incremental productivity adopted is relatively low and is of the magnitude of a yearly 3% increase in productivity for these farms where new technology has been adopted. 6. Analysis Assumptions; In order to measure the economic value of the assumed changes in productivity at border prices, the present value of production for project participants as well as the incremental value of produc- tion for future project participants accepting new technologies (see para 4) were computed. The following explicit assumptions were made in order to measure the value of incremental production attributable to the project: - Extension services benefit evenly all activities carried out by project participants accepting new agrotechnologies. - The existing farm types, area per farm and cropping pattern for low income farmers and other farmers will remain unchanged during project life. - Project target population of 1,300,000 farmers will be composed of 55% 1/ (715,000) low income farmers and 45% (585,000) other farmers. - The target group of farmers selected collectively provides adequate representation for all crops that are grown in Brazil. - The average present annual rate of inrease in agricultural production (7%) is distributed evenly between the project participants (1.3 million farmers) and the remainder of farm holdings, i.e., 3.6 million for the current total of 4.9 million farm holdings in Brazil. - The expected effects (3% increase in agricultural production) attributable to the improved extension services are distributed evenly among all the crops grown by the project participants. As field crops represent 52.7% 2/ of the value of agricultural production, a conservative assumption would be that 52.7% of project costs could be justified by yield increases for field crops grown by project participants and accepting new agro- technologies. 1/ Presently, the ratio is 20% (106,240) low income farmers and 80% (424,330) other farmers. 2/ Overall agricultural production value in 1976, Cr$284,706 million of which about Cr$151,500 million for field crops (52.7%).  - 52 - ANNEX 2 Page 4 Incremental project foreign exchange costs, although relatively low, were adjusted upward by 30% to reflect taxes, tariffs and subsidies. The life of the project was assumed to be of 12 years and no residual value is attributed after that period. No scrap value was placed on buildings and equipment. Border prices were used for most commodities. For some products such as manioc and beans, which are produced mainly for local consumption, constant farmgate prices were used. This was done although it is recognized that some of these crops are likely to have price changes in the future but such changes are presently unknown, and difficult to forecast. C. Project Risks 7. As it is expected that the benefits to be achieved will gradually extend nationwide, as there are so many uncertainties involved in such a project, and since the extension service must deal with a very wide range of client farmers with differing levels of expertise and education, the extension service will have to be highly dynamic and adapt itself effectively to changing situations in order to be able to attain the expected benefits. EMBRATER, which is a relatively new institution, will have to strengthen its administra- tive capacity to implement the project. Another type of risk lies in the fact that agriculture in Brazil is mainly rainfed; there is currently a relatively high degree of risk concerning achievement of the expected yield increases due to climatic variation. However, based on existing data concerning agricultural know-how in Brazil, it is clear that enough expertise exists to lessen the effects of these variables. D. Project Benefits 8. Social Benefits: The development and upgrading of the agricultural extension service which, in Brazil, also supports programs related to social welfare and health, adult education, nutrition and home economics, would also have a considerable social impact that is quite impossible to measure and hence was not taken into consideration in the present analysis. 9. Economic Benefits: The economic analysis for the proposed project was carried out based on 52.7% of project costs and benefits (see para 6). An internal rate of return was calculated for project components related to the incremental agricultural production of project participants and the rate of their acceptance of new agrotechnologies. The benefit stream was calcu- lated to include the incremental crop production by project participants and the phased increases of productivity achieved by participating farmers. The  - 53 - ANNEX 2 Page 5 cost stream included all project incremental costs required to achieve the production increases. Tables 5.3 and 5.4 provide details of the costs of the project and the expected benefits. 10. The internal rate of return for the project components (assuming a 12-year discount period, equivalent to the period within which acceptance of new agrotechnologies still occur) was estimated to be in the range of 18%. This compares favorably to the opportunity cost of capital (11%) invested in Brazil. 1/ The spread in the range given would appear to make ample provision in order to cover the possible risks and uncertainties which have been noted above. A sensitivity test shows that an increase of about 2.5% in productivity will bring the internal rate of return to nearly 11% or the equal of the oppor- tunity cost of capital invested in Brazil, and a 3.0% increase in productivity would provide a return of 18%, based on the assumptions as detailed in para 6. 1/ See Memorandum entitled Estimates on Opportunity Cost of Capital for LAC Countries by Regional Senior Economist, February 8, 1977.  BRAZIL AGRICULTURE EXTENSION I PROJECT Target Population Basic Data Area uncer,field crops Percentage of Acceptance Number of Farmers receiving Low Income Other Total Project of New Agro-technologies improved extension services Farmers Farmers Year Yearly Cunmulative Low Income Others Total -- -(ha)- - - - (ha)- - - -(ha)- - - 1/ - 1 1 0 0 106,240 424,330 530,570 743,680 5,227,745 5,9719425 2 10 10 309,000 478,000 787,000 2,163,000 5,888,960 8,051,960 3 5 15 512,000 532,000 1,044,000 3,584,000 6,554,240 10,138,240 2/ 1 4 5 20 715,000 585,000 1,300,000 5,005,000 7,207,200 12,212,200 5 5 25 715,000 585,000 1,300,000 5,005,000 7,207,200 12,212,200 6 5 30 715,000 585,000 1,300,000 5,005,009 7,207,200 12,212,200 7 5 35 715,000 585,000 1,300,000 5,005,000 7,207,200 12,212,200 8 5 40 715,000 585,000 1,300,000 5,005,000 7,207,200 12,212,200 9 3 43 7151000 585,000 1,300,000 5,005,000 7,207,200 12,212,000 10 3 46 715,000 585,000 1,300,000 5,005,000 7,207,200 12,212,000 11 3 49 715,000 585,000 1,300,000 5,005,000 7,2071200 12,212,000 12 1 50 715,000 585,000 1,i00,000 5,005,000 7,207,200 12Z212,000 a 5 / Present situation. 55% of target population by project year 4. ;V 45% of target population..  BRAZIL AGRICULTURE EXTENSION I PROJECT Incremental Production Attributable to Project Yearly Production Value Ineremental Value of Yearly Area Adopting New Cumulative Area Adopting New - - of Areas. *2/ Product4on Attributable Agrotechnologies 1/ Agrotechnologies AdoPtinR New AeroQtMlhnAn .ie4 - To Project 3L Low Income Other Total Low Income Other Total Low Income Other Total Farmers Farmers Farmers Farmers Farnmers Farmers