TRAINING MATERIAL CASE STUDY 1 BRAZIL: RIO GRANDE DO NORTE RURAL DEVELOPMENT PROJECT Agriculture and Rural Development Department The World Bank 1818 H Street, N.W. Washington, D.C. 20433 USA September 1980 RIO GRANDE DO NORTE RURAL DEVELOPMENT PROJECT Postscript - 1981 Implementation of Phase II was formally initiated in April 1979. Four new components were added to the original five: inland fisheries/small reservoirs, marketing, seed production, and land titling. In fact, from the Brazilian point of view, the project included several more components -- CAP, cooperative development, input supply, rural electrification, education, community development, and roads. The World Bank did not participate in the financing of these components, but Project Unit staff dedicated a considerable amount of time to monitoring and supervising them. Some two years into Phase II, both Mr. Falloux and the new project officer, Ms. Pomerantz, felt that Phase II implementation results were disappointing. In particular: 1. Access to credit by small farmers was much improved in terms of additional local bank offices and simplified lending procedures, but credit provision was subject to delays and shortages at the local level. Land purchase credit continued to be unavailable. On some occasions an absolute shortage of funds at state level was responsible for project credit difficulties, but more often the problems arose because of faulty communication between local bank managers and the state superintendent, and because credit monitoring was inadequate for management purposes. -2- 2. Coordination between research and extension had not markedly improved. Researchers continued to view their work as largely an academic exercise, and were reluctant to release partial results. There were no regular channels or meetings to inform local extensionists about on-going research. While research did produce some interesting technological packages for irrigated or semi-irrigated lands, few recommendations for rain-fed agriculture were generated. Professionally, most researchers considered work on the latter to be more difficult and less rewarding. 3. The expansion of the extension network had not been effective. The new extensionists, many not from the area, were trained outside the state. Because of funding delays, most remained unemployed for about eight months after the conclusion of their training course. After the extensionists' arrival in the field, the severe drought in Northeast Brazil forced the Federal and State Governments to focus most of their attention and available human resources on the Emergency Drought Relief Program. Extensionists became, in effect, the local administrators of the program, leaving them little time for direct technical assistance to farmers. Planned seed production and marketing results were also far below appraisal estima:es, largely because of the drought, but also due to administrative difficulties. 4. Two of the Phase 1I components, inland fisheries and land titling, took over a year to begin implementation. While land titling, once started, was proceeding at a satisfactory pace, the project officer was concerned that the titles granted might not be permanent, since land registration records and cadastral surveys were either incomplete or non-existent for the project area. In addition, landless laborers and sharecroppers on large estates continued to have insecure land tenure and few possiblities for acquiring land. In the case -3- of inland fisheries, it was not clear that the target group was directly benefitting. The benefits seemed to be accruing to larger land-owners with previously constructed private reservoirs. 5. The increasingly difficult economic situation in Brazil, and consequent limitations on public spending, also hampered project performance. While, at the start, the project was the first "new-style", poverty-oriented rural development project in Northeast Brazil, by 1981 it was just one of forty-six projects in the regional POLONORDESTE program. That program, which had grown to an annual budget of some US$150-200 million, was subject to serious funding delays and cutbacks. For Bank-assisted projects, the Bank typically funded 35% of project cost, usually on a reimbursement basis, and Government the balance; delays in receipt of the Government's share (the 65%) have a devastating impact on the rate of implementation. Financial delays and shortfalls created a permanent climate of uncertainty, and in some cases, cynicism, among project staff and administrators. In addition, State spending (as opposed to Federal, the major source of funding) was also restricted. As a result, project-funded vehicles, equipment, and staff were frequently used for other programs or kept in the capital city. Despite Bank efforts, the project unit and field managers had difficulty in obtaining the logistical support necessary for regular field supervision. In such a situation, implementation problems were often left undetected and unresolved for longer than necessary, particularly in the more distant municipalities. So, after more than five years of implementation, the project was beset by a number of problems, the most important of which were funding delays and shortfalls, limited, albeit improved, execution and management capabilities of the involved agencies, and a prolonged and severe drought which brought -4- additional problems to the area. It was clear that the project would not be completed on schedule even though the completion date had been extended by 15 months when Phase II started. In five years, less than 25% of the Bank loan had been disbursed. Given this situation, some federal officials were in favor of cancellation of the loan, or in any event, were not willing to consider a further extension. They felt that project performance had demonstrated the limited agricultural potential of the area and that resources could be more profitably invested elsewhere, perhaps even in another portion of the state. Some of them also felt that the project had been poorly designed, particularly since it had ignored water resource develoment in one of the most drought-vulnerable areas of Northeast Brazil. Other federal and state officials held an opposing viewpoint. They believed that the project should be continued and even extended. They pointed out that the area, while difficult, was typical of a large part of the semi-arid Northeast; that solutions to the problems associated with the area could be found, and that, in fact, the project was beginning to provide some results. Many felt that poor project performanc,e was not essentially the fault of poor design, the area's limited agricultural potential, or poor management and execution. They placed much of the blame on the Federal Government's funding delays and shortages, which, according to them, had tied the project staff's hands. They argued that it was unfair to compare the project with other rural development projects in the Northeast, since other projects had benefitted from innovative aspects of the RGN project without having to undergo the same trial-and-error learning process. Officials pointed out that many of the Phase II difficulties could be overcome in the short run, given adequate funding, and that water resource development to increase small farmer resistance to drought -5- could be added at this stage of the project. During this debate, which lasted months, the Agricultural Planning Commission, to which the project unit belonged, began preparation of a new rural development project in the Litoral Agreste, an area of 53 municipalities located in a somewhat higher rainfall area between the current project area and the coast. Project preparation, partially financed out of the proceeds of the Bank loan, sparked a good deal of enthusiasm among local planners. In fact, some project unit staff became more involved in preparation than in on-going supervision of the current project. The general perception was that the Litoral Agreste had greater agricultural potential and better infrastructure than the existing project area and that, consequently, project implementation would be subject to fewer delays and difficulties. Water resource development was to be a major feature of the new project, particularly important in this drought-prone state. For a variety of reasons, including political considerations, the state was eager to begin project implementation with Bank financial assistance as quickly as possible. However, the detailed studies required, particularly for water resource development, would take at least two to three years to complete. The State argued that implementation, on even a limited scale, should begin more quickly. At this stage the project officer and government planners reflected on the past and considered options for the future. Questions they needed answers to included the following: (i) Had the original objectives of the project and the revised objectives for Phase II been correct in light of the lack of institutional capacity in the region? Had sufficient been known about this institutional capacity when plans were made, or about -6- other demands on limited resources? (ii) Had the Government, with Bank assistance, effectively attempted a pilot project on too large a scale. (iii) Had design of both Phase I and II been faulty for not including a water resource component? (iv) Had the addition of new components in Phase II distracted attention from more fundamental activities for development in the area? Should the components exclusively financed by the Brazilian government have been included within the definition of "the project"? (v) Had the jump in municipalities from 11 in Phase I to 55 in Phase II been too drastic? (vi) Could design have accommodated both concepts referred to in the description of the project (page 5)? (vii) Should the project be cancelled? In either case, what should be the Bank's position in the preparation and financing of the Litoral Agreste project? Should linkage of the two projects be encouraged? (viii) Why had design of Phase II not taken sufficient account of the changing climate and associated implementation difficulties of the growing POLONORDESTE program and worsening economic situation? (ix) The drought accentuated many difficulties and int7oduced others. How could design accommodate such crises? (x) What lessons from the project could be incorporated into the design and preparation of the Litoral Agreste project and other state-level projects which the Bank might assist in the Northeast? DRAFT 9/5/80 Connie Mejia:jh BRAZIL - RIO GRANDE DO NORTE : RURAL DEVELOPMENT PROJECT Upon returning to the World Bavk from the Mid-Te=t Evaluation (MTE) follow-up mission in October 1978, Francois Falloux, the project officer,sat down to reflect on the information he and the Brazilian government team had collected regarding Phase I of the Rio Grande do Norte Rural Development project. Though performance under Phase I had not been uniformly up to appraisal expectations, he felt strongly that important achievements had been made. He wanted to explore the nature of both these achievements and the obstacles to progress that still remained. He also wanted to determine how both factors related to the project's design in order to recommend any needed changes to the design of Phase II. In January 1976, the Bank made a US$12 million loar to the Government of Brazil to finance the Rio Grande do Norte (RGN) Rural Develop- ment project. The five-year, US$30 million project Vas to begin in April 1976; it was to be the first step in a larger effort to deal with rural poverty in Northeast Brazil. For both the Bank and Brazil, the project was experimental. For the Bank, the project was an early attempt at "new style" rural development projects, as delineated by Mr. McNamara in his Annual Meeting speeches of 1972 and 1973, and elaborated in the 1975 Rural Development Sector Policy Paper. The RGN project was to use a minimum-package approach to raising the incomes of the small farmers in the project area, while developing the institutional capacity to carry out similar projects on a larger scale. For Brazil, it was the first regionally focused, integrated rural development project aimed specifically at the small farmers of the Northeast. Also, it was the first project in a regional rural development program -- POLONORDESTE 1/ -- that by the time of MTE had over thirty similar projects and an annual budget of approximately US$150 million. Because of its experimental character, it had been planned at the time of appraisal to be implementedin two phases: Phase I was to comprise the first two project years and concentrate on eleven municipios selected to represent the 53 municipios of the total project area; Phase II activit- ies were to take place over the following three project years in both the initial eleven municipios and the incremental 42. The decision to proceed with Phase II was to be subject to a comprehensive review of Phase I, performed by the Brazilian Government in consultation with the Bank. Due to various delays in early project implementation, Phase I was extended from two to three years, to end March 31, 1979. The "comprehensive review of Phase I" was conducted by a joint POLONORDESTE/Bank MTE mission in July-August 1978 and by a follow-up MTE mission in October. Northeast Brazil Northeast Brazil contains the largest concentration of poverty in Latin America. It comprises approximately 20% of the country's territory 1/ POLONORDESTE is a special development fund for the Northeast. - 2 - and is inhabited by 30% of its population. The area produced only 11.6% of Brazil's Net Domestic Product in 1970, however, and per capita income averaged less than US$200 -- less than a third of the national average. Early Government attempts to improve incomes. in the Northeast were focused on urban areas. They consisted of fiscal and monetary incentives to encourage industrialization, They also included supplementary social programs. Sixty percent of the population w?s rural, however, and largely beyond the reach of program benefits. In its approach to rural development, the Government generally drew a distinction between agricultural policies and policies aimed at improving the welfare of the rural poor. The former were generally designed to stimulate output to meet increasing domestic demand for food and fibers and to encourage production for export to contribute to foreign exchange earnings. The rural poor were not considered viable targets for production- oriented programs; rather, the Government felt that their problems were best alleviated through socially oriented programs focusing on education, health, nutrition, sanitation, and related areas. At the time of appraisal, however, the emphasis of governmental programs was beginning to shift. POLONORDESTE in particular was designed to raise both productivity and incomes of small farmers through integrated rural development projects to be carried out in priority areas; these were to be selected on the basis of both large concentration of poor rural population and development potential. Rio Grande do Norte Rio Grande do Norte (RGN) is a relatively small state in the Northeast (see Map, Exhibit I), but its problems are representative of those faced in that nine-state area. The capital, Natal, is located on the coast, with a population of 350,000. Most of the industries were centered around Natal, but were not sufficiently developed to absorb the state's expanding labor force. With the exception of a narrow, humid coastal belt, the state has a semi-arid climate and is subject to periodic devastating drought. Records available at appraisal showdd that severe droughts occurred every six to ten years, and that the Government responded with Emergency Relief Plans involving massive expenditures to relieve the ensuing destruction and hardship. Due to the lack of long-term planning implicit in such plans, much of the money spent through them was not effective as investment. Those physical improvements which were made with Emergency Relief Funds were usually done with insufficient technical advice and were destroyed by flooding when the drought was finally over. The health problems of RGN are worse than those of Brazil as a whole. The state has a higher rate of infant mortality and a higher incidence of both communicable and nutrition-deficiency disease than the country generally. In an effort to alleviate these exceptional health problems, the state initiated a pioneer program of low cost village-level "mini-posts" -- small primary-care clinics -- and supported a family planning program. -3- The project area covered 21,000 km2 or approximately 40% of the state and was located entirely in the drought-prone interior. Average rainfall varied between 400 mm and 700 mm annually, but was e.ctremely erratic both in seasonal occurrence 7nd from year to year. A history of drought made agriculture very risky and adversely affected farmers' willingness to innovate and invest. Most of the area was completely rainfed. The area was divided roughly into four sub-areas which comprised 53 municipios (counties). Soils were generally poor throughout the area except in valley bottoms, but there was great variation among the sub-areas in soil quality and type as well as in climate, geology, crops and infrastructure. Only 18% of the area was under crops at appraisal; the rest was either natural pasture or unusable. The project area coincided roughly with the state's perennial cotton-growing area. The perennial cotton plants had a five-year cycle and good drought resistance. Agricultural cash income was generated mainly by this high-quality, but low-yield, long-staple cotton. Cash income was supplemented in some instances by cattle (mostly beef and rarely dairy). The chief food crops grown were maize and beans. These were usually interplanted with first-year cotton and were cultivated particularly by small-scale farmers. Food crops served a dual purpose for these farmers. They were sold at harvest time in May (at low prices) to provide cash needed for employment of extra workers for the harvest of cotton in June. When the cotton crop was sold, the food crops were bought back for family consumption (at considerably higher prices). There were between 20,000 to 25,000 farms in the area, 75% of which occupied less than 50 ha each and together accounted for 25% of the land area. Intensity of land use varied with farm size; farms of less than 50 ha averaged 40% to 45% cultivated area. Those greater than 200 ha were cultivated over 10% to 15% of their area. This difference was not due entirely to differences in soil quality. Table I shows selected performance indices by farm size. According to the 1970 census, the population of the area was 415,000 -- about 26% of state population. The rural population was about 270,000 and was estimated to grow at an annual rate of about 1.3%, as compared with the urban population growth rate of 4.6%. The difference in rural vs. urban growth rates was attributed partly to the generally worse health of the rural population, and partly to rural emigration to cities in the Northeast and irdustrialized Central South areas. Of the 52,000 rural families in the project area, more than half owned no land: about 35% were temporary workers and 20% were share-croppers. Thirty-five percent were owner-operators of small farms l/ and 10% were owner-operators of large farms. Those owners with less than 10 ha usually share-cropped additional fields for the larger owners, and all but the largest landowners were relatively poor by Brazilian standards. There was a wide range of share-cropping systems used in the area. A common example allowed for a 50:50 split of the cash crop between owner 1/ Small farms are defined as chose less than 200 ha. -4 and share-cropper. The owner provided original plantings and pesticides, and the share-cropper had the right to interplant food crops. In addition, the owner often provided the share-cropper with credit and sold him food crops in the off-season. The cash crop was typically marketed jointly. Cash crops were generally sold to a private cotton gin or to a buyer/transporter. The latter is defined here as a buyer who pays the producer a faragate price and takes responsibility for transporting the crop to the cotton gin. These middlemen often bought the crops prior to harvest. There were only four cooperative cotton gins operating as buyers in the area, and together they accopnted for only about 18% of total sales by area farmers at time of appraisal. All of the cotton gins operated with considerable excess capacity. Institutional sources of credit were used by relatively few farmers; it was estimated that over 80% of all farms in the project area did not borrow from institutional sources. Among the informal sources of credit commonly used by small-scale owner/operators were the crop buyers mentioned above. The condition of such informal loans was tied crop sales at prices disadvantageous to the farmer, making the implicit interest rate high -- often 50 to 80 percent. Interest rates implicit in the transaction described above between share-cropper and landlord were in the same range. I The producer services offered by extensionists in the area were geared toward the larger scale farmers, (those with farms over 200 ha) and therefore reached a small percentage of the rural population. Most of the extensionists were from large farm-families, and both their attitude and their knowledge were biased toward large-farm production. It was estimated that fewer than 10% of the area farms were affected by any extension service. The extension service's limited success was also due to the fact that most of the agent's time was taken up with administrative matters, in particular with completing the paperwork required for credit applications -- usually from large farmers. Research on improved cultivation techniques was carried on by two separate agencies and one university. Most research focused on cotton and did not treat questions important to small farmers, such as food crop cultivation. Some of the research identified technical innovations, but at the time the project began, the results had not been coordinated into technical packages economically attractive to small farmers. Roads and private trucking services for transporting crops were considered adequate considering the traffic flow. Input supply, on the other hand, was not considered adequate, particularly for improved seeds, but a new agency was in the process of reorganizing to improve service. Finally, a semi- private radio service (Radio Rural) was operating in the area, transmitting farm and rural news to the large number of radio receivers in the project area. -5- The Project From the beginning, there had been a difference in concept between the objective of the project as prepared by the Government and that which the Bank wished to encourage. The Brazilian proposal chosen from several alternatives was focused primarily on cotton production. This proposal had no target-group focus and was quite costly per beneficiary. Because of its main objective, this proposal was referred to in Brazil as the "Long-Staple Cotton Project," and the same name was later used for the rural development project actually approved. The appraisal mission was atare of the b1is toward large farmers in agri- cultural projects but decided to gamble that a reorientation to a target-group focus might be effected through careful supervision of this pilut project. The appraisal report stated the following objective: "to develop the institutional capability to conceive, plan, and implement in a coordinated manner actions to improve the incomes and welfare of a target group of small-scale farm families at cost levels low enough to allow replicability." This was the objective agreed upon with federal- level officials through the negotiating process. The first specific objective was to reach 15,000 families through the directly productive components. These compc&ents were expected to double the cotton production of participants and generate a marketable surplus of food crops. These two effects were expected to result in a 50% increase in the incomes of fully parti- cipating families within five years. The second objective was to provide complete community health coverage to about 80% of the rural population. The design was to include components directed toward development of appropriate institutional as well as technological changes. These experimental aspects of the project were extremely important from the Bank's point of view, since this was to be one of the first poverty-group oriented, "new-style" projects, and there was a recognized need to develop strategies for raising the income of poverty groups through increasing -heir productivity. It was hoped that innovations would be developed to help address some of the problems basic to rural development projects -- e.g., how to reach landless workers and bring them into the productive mainstream. In the meantime, however, for the RGN project, it was decided to aim at the most easily reached sub-group within the general poverty group. For this reason the main target group was owner-operators of small farms. It was thought that share-croppers and landless workers would only be reachable through the health component. From the Bank's point of view, it was considered of critical importance to establish the viability of the small farmer as a producer. The major focus would be the identification of suitable technical packages to small farmers. It was expected that improvement of delivery mechanisms would involve considerable institu- tional reorganization and increased participation of small-scale farmers. In order to take advantage of the experimental features of the project by allowing the results of research and testing (at both the agronomic and insti- tutional levels) to be incorporated, the project was designed in two phases. Phase I was to be completed in two years and was to focus on 11 municipios covering about 20% of the cultivated land in the project area. The objectives during Phase I were to refine and disseminate technical packages financially attractive to small farmers; to develop methods of delivering extension, credit, and health services to the target groups and to strengthen the existing rural development institutions -6- in the project area. The decision to proceed with Phase II would be based on comprehensive review of Phase I by the Government in consultation with Bank staff. The objective in Phase II would be to expand actions developed and tested during Phase I to more (or.all) of the project area. The decision to support only a small number of components was based on several considerations, including the fact that there had been no previous experience with this type of project in Brazil, and that there was limited managerial capacity in state institutions. In view of these limitations, it was thought wisest to keep the design as simple as possible. It was recognized that the share-cropper population might be gradually included as owner-operators, but the question of land tenure was too politically sensitive at the time to permit discussion. Annex I gives additional background on this issue. An irrigation/water control component was considered, but was not included for two main reasons. First,irrigation projects were traditionally biased to benefit large farms. Second, the heavy emphasis on infrastructure that such a component would involve would distract from the main objective of developing ways of increasing small farmer productivity. The major components finally included were : (i) credit for small-scale farmers; this would involve a major reorientation of Bank management toward small farmer borrowers; (ii) extension service expansion to reach the target group; (iii) cro experimentation and farm trials; to focus on technologies suitable for adoption on small farms; (iv) expansion of existing health program to address the real needs of the rural population; it was also hoped that the type of community organiza- tion begun by the health service would facilitate farmer organization for productive purposes; (v) organizational support for project administration and monitoring and evaluation. In addition, 1% of baseline costs was to go to development of training materials and to the actual training of farmers, and another 1% to the preparation of rural development projects for other areas of the state. Table II shows the planned phasing of investment by component. The individual components will be described in more detail below. Credit In order to focus attention on provision of institutional credit to small-scale farmers, this component was to include experimental aspects as well as direct borrowing by farmers: (i) alternative credit delivery systems were to be simlified and tested; (ii) procedures were to be developed -7- to imprJve small farmer access; and, (iii) crop-credit insurance was to be made available to small farmers. Small farmer loans were to be made for seasonal production purposes,for on-farm investment,and for area expansion. A ceiling was placed on individual loan amounts, and new borrowers were to represent 80% of the total number receiving credit. Both of these requirements were an attempt to reach the smaller scale farmer and discourage the larger owners. The success of the credit component was considered essential to the project, since it would allow the farmers to adopt the technical recommendations developed through farm trials. Thl-, credit component largely failed in both its new borrower and small farmer borrower goals. New borrowers represented 20% to 35% of all borrowers (with variations among lending institutions), compared with the target of 81%, and the percentage declined by about 30% in the second credit cycle. There were several reasons perceived for this poor performance: (i) Misunderstanding on the part of the Project Unit (PU), extension agency,and participating banks regarding project objectives, Emphasis at all levels was on increasing cotton productivity in general. New borrowers received low priority, as borrowers with institutional credit were easier targets. (ii) For extension agents, new borrowers required considerable one-time paperwork. Old borrowers with papers on file were considered more desirable. Old borrowers were also sent to extension agents for small loans by bank managers, since project loans required less evaluation work. (iii) Experienced borrowers put pressure on bank officials for cheaper project credit; this was especially serious because other sources of credit dried up at about the time of initiation of the project. (iv) Banks often required notarized documents attesting land ownership and/or third-party guarantors of new borrowers who were small land owners. In regard to size of farm owned by borrowers, Phase I failed again in accomplishing the objective of reaching its target group, if the criterion is total volume of funds lent. Tha MTE mission estimated that 20% to 30% of the total value of loans went to farms larger than 200 ha. Some of these loans went for the purchase of livestock and heavy machinery, for which project funds were not intended. This was ascribed to the considerable local influence of large land owners. If the number of loans going to different sizes of farms is considered, however, Phase I performance is closer to targeted levels. Table III gives more details on the distribu- tion of loans for farmers. Since small-scale farmers were expected for the most part to be new borrowers, many of the factors mentioned above explain their low participation rates.* In particular, bank officials gave low priority to project objectives and activities; -8- (a) They believed their performance was evaluated on the basis of total loan volume, and therefore were inclined to make larger loans. (b) Siasonal credit was limited to perennial cotton production, except for food crops interplanted with cotton, thereby excluding an important small farmer activity. In addition to the above problems, the credit component fell short of its objectives in three other areas: (i) land-purchase credit was not financed (see Annex I); (ii) credit-crop insurance was not promoted. The PU staff believed that it was inappropriate for a perennial crop like cotton, despite the fact that the national credit-crop insurance scheme had been important and highly successful in coffee-growing areas; (iii) project-financed purchase of traction animals was about half that intended. In spite of these problems, there were significant achievements made under the credit component: (i) It was shown that it is possible to provide credit to small farmers without undue risk (since their repayment rate was close to 100%). (ii) The possibility of simplifying credit application and processing procedures was proven, and some simplification did take place. (iii) Project-financed cotton expansion greatly exceeded appraisal estimates, despite a major drop in cotton prices in Year II. Due to start-up delays, a short- fall in new cotton planting was expected, but plantings of new cotton exceeded expectations for the first three years. This was partly because old borrowers increased planted area by a much greater percentage than anticipated. (iv) A major innovation in the delivery of credit to small and landless producers -- though it was neither executed by the banks nor financed by the project -- was the advance-crop-purchase scheme (CAP). Since it was executed through the extension service and had its major impact in that area, it will be discussed below. -9- Extension The main objective of the extension component was to reorient and expand extension activity to reach small producers while continuing to service medium and large farmers. This was to happen through i) enlarging the service in the project area; ii) training technical staff to focus on small producer systems and problems; iii) restricting the credit functions of extensionists to technical recommendations for proposed credit plans; and iv) increasing farmer coverage and lowering per farmer costs by organizing and working through groups of small farmers. In addition, it was expected that time would be given in Phase I to: coordinating extension and research activities; organizing input delivery and marketing systems; obtaining limited and relevant monitoring and evaluation information; and improving the technical capacity of the service. In the first two years the service was enlarged, but its orientation was not significantly changed from its previous large-prodtcer focus. In Year III, the service became involved in testing CAP. Compra Antecipada de Producao 1/ (CAP) As a result of the Bank's request to seek alternative methods of reaching small farmers, a Ministry of Agriculture work group was formed. This group developed a proposal, and the project area was selected for testing. The principle upon which CAP worked was to provide seasonal credit to small producers by allowing them to "sell" (pledge to sell) their crop in advance of harvest. This innovative system was to have three effects: (i) break dependency on landlords and middlemen for credit markets; (ii) regulate farmgate prices by guaranteeing an outlet for cotton and food crops at state-capital prices less handling and transport; and, (iii) prepare farmers for access to institutional credit. With respect to the first two objectives, the program was an overwhelming success. By 1978, 5,000 farmers were organized into 350 groups and were receiving extension services. This represented 134% of the number of farmers expected to be reached in Phase I through 33 extension workers, During 1978, a total of 1,912 contracts were made under the CAP program, with a total value of 9,718,300 cruzeiros. With respect to the third objective, the program did prepare farmers for institutional credit in the sense that it brought them into contact with an organized money market, but prepared them poorly for bank credit, since the interest rate under CAP was 0%. In the eyes of the project officer, however, CAP had much more important results than its immediate successes. By dramatically increasing contact between extensionists and small farmers, it not only demonstrated to the agents the possibility of servicing small farmers and share-croppers, but it also made them much more willing and enthusiastic advocates of the small producer. This led to significant changes discussed below. 1/ Advance Purchase of Production - 10 - Until Year III and the experience with CAP, agricultural extensionists (many of whom were also large farmers) preferred to see themselves as agronomists rather than as small-farmer advocates, but their technical recommendations were often inappropriate to the farmers' needs. Since they understood their goal to be one of increasing cotton productivity, they often recommended reducing interplantings of food crops, which are of crucial importance to small farmers. In addition, serious funding problems led to poor morale and low incentive to innovate In contrast, social extension- ists often displayed more understanding of small producers' problems and were more motivated to overcome obstacles to get services to them. Other problems of the extension component in Phase I were organizational: there were considerable staff-instability and coordination problems, with the state directors of the agency often ignoring project objectives. One major innovation of the project was to carry out the pre-service and in-service training of agents in the project area. The pre-service training was successful in promoting cooperation between the agricultural and social extensionists and establishing contact between the agents and individuals in the community. The in-service training, however, was either not done or was timed inappropriately for farmers to be able to use recommendations made by the agricultural experimentors. In addition, there was a serious lack of continuity in the extension service resulting from rapid rotation of staff. Trained staff were quickly promoted out of the project area to state- and federal- level posts. Their replacements did not receive the same high quality of initial training and in some cases were unclear regarding project objectives and even regarding the function of the extension service itself. Airlied Research and Farm Trials The objective of this component was to develop technologies adaptable to the small-farmer target group and specific to the project area through applied research at experimental stations and farm-level demonstrations. The results were to be translated into techrical packages financially attractive to the target group. After a slow start, tests were done on alternative intercropping systems, fertilizers, planting time, planting systems, spraying, weeding techniques and soil preparation. The tests were generally successful, and several specific recommendations were made in spite of delays in data analysis. There were still some problems noted at the end of Phase I: a planned survey of households was not done; tests on farm implements, animal traction and post-harvest on-farm storage of food crops were not done; and extensionists did not begin to use research results until the third year. Coordination between research and extension was hampered by institutional jealousies and conflicts between the two agencies involved. Health The objective of this component was to expand and improve the existing innovative extension-supported community health mini-post program. After some long delays, new centers and mini-posts were established. Though staff of local posts was inadequately trained at first, training was later improved. There were still problems in getting the mini-posts adequately supervised, and the program was affected by the same funding delays that plagued the other components. Greater community participation was achieved, but there was no carry-over as yet to farmer organization for productive purposes. - 11 - Nevertheless, the component was judged to have satisfactorily achieved its Phase I objectives. Administration and Funding The project was to be administered by a Project Unit set up with a project coordinator and technical advisers who were to work closely with implementing agencies. Each adviser was to chair a committee of representatives from the executing agencies to coordinate work on the particular components. A separate monitoring and evaluation group was also to be set up. The Project Unit was intended to have responsibility for coordinating the efforts of the various agencies, approving their plans, and approving requests for Bank funds. Project funds were not to be channeled through the unit. This was an important limiting factor in the Project Unit's ability to pursue project implementation in an aggressive manner. The bureaucratic procedure required for actual implementation of the program was quite complex. Annual work plans had to be submitted by the executing agency and the appropriate Project Unit adviser, to be reviewed by the regional and national POLONORDESTE bodies and the state-level rural development policy coordinating body (Conselho Director). These bodies required a complicated process of fund request, allocation approval, and release of funds. The process was cumbersome, and, along with the lack of authority and drive of the national POLONORDESTE reviewing body, was largely responsible for the funding delays that affected all components. The Project Unit was composed of high-quality staff, but morale was low due to the frustration of not being able to coordinate or direct effectively the implementing activities of the executing agencies. This in turn was largely due to thc- lack of formal authority of the Unit. Whatever informal authority the individuals may have had was undermined by funding delays. The Unit was also hampered by a lack of autonomy in recruiting of staff and setting salaries. The Monitoring and Evaluation group of the Project Unit was practically non-existent in Phase I. The support/administrative staff was never hired, and what staff there was had to spend most of its time filling out the quarterly reports required by POLONORDESTE. Reviewing the documents which recorded progress during Phase I, the project officer felt that some of the more important achievements were reflected only by implication. This was partly because of the nature of the achievements and partly because of the context in which he viewed the project itself: the project affected a tiny microcosm of a small state in one area of Brazil, but it served as a laboratory for institutional changes which would be essential to progress in rural development throughout the country. The records did show the development of credit delivery alternatives (CAP) which promised to be transferable to other projects and states. Equally important was the fact that a good trained team of local technicians had been established, in increasingly close contact with the target group of small farmers, and was ready to proceed with Phase II. Perhaps of overriding importance, however, was the change in attitude toward the small farmer as a viable target, which he had observed at all levels -- government planners, extension agents, bank managers were all more enthusiastic about the potential of small farmers as producers. This change had begun to result in institutional and organizational changes which might justify a more optimistic expectation that measurable targets - 12 - would be met or even exceeded in Phase II. An example of such a change was that the Bank. of Brazil, after apparent reluctance to participate in the credit component of the project in Phase I, had undergone a fundamental reorientation. Regulations had been devised to require branch managers to focus on small farmers in providing credit, and supervision was provided at the state level to insure compliance. As a result of the new regulations, managers perceived less risk in dealing with small farmers, and were less vulnerable to pressure from large landholders. This and other similar changes promised a Phase II performance much more satisfactory than that of Phase I. There were still several questions that he had to consider for possible modification of Phase II design, however: (i) Had the scope of the project been comprehensive enough? (ii) How could the Bank facilitate improvements in organiza- tion to reduce funding delays? Would the Bank be over- stepping its role in doing so? (iii) How could what was learned with respect to the organiza- tion of small farmer groups be used to make other components more effective? Could these ideas be applied to areas which had previously been considered too expensive (e.g., improvement of water control and irrigation)? (iv) Was the design flexible enough to make use of what had been learned? (v) Had the timing been appropriate? (vi) How could the difficult and sensitive problem of land titling be dealt with to bring more of the share-croppers and landless workers within range of the directly productive components of this and other projects? ANNEX 1 Page 1 LAND TENURE AND REFORM The Brazilian Government recognized the problems stemming from the regionts extremely uneven pattern of land tenure and had made some attempts to alter the situation prior to 1975. PROTERRA (Program for the Redistribution of Land and Stimulation of Agro-business in the North and Northeast) had been instituted to finance land redistribution, but had had minimal overall impact due to administrative shortcomings and inadequate support from INCRA, the land reform agency. A land-purchase credit program under PROTERRA and administered by Banco do Brasil had had some success, but the credit terms had limited its impact: a 20% down payment and 12-year repayment period put the program completely out of reach for landless workers and sharecroppers and for most small farmers as well. At appraisal the Government had begun an intensive review of its land reform policy and programs. An additional obstacle to land purchase by small farmers was related to the supply of land on the market. Small farmers were interested in buying small additional parcels of land, but the land offered for sale was generally in enormous tracts, and the land owners were unwilling to divide the tracts for many small buyers. At appraisal, the Government agreed to support project objectives through a revision of the terms for credit for individuals to purchase small parcels of land and become owner-operator farmers. Credit under the revised terms was to be available under POLONORDESTE in the project area. The new terms were to be a 7% interest rate and 100% financing of the purchase price. Loans were to have a 20-year maturity, including a five-year grace period. These terms were later modified to 12% interest, 20 years amortization and six years grace. It was recognized that this still did not alter the problem of the structure of the land market. At MTE it was suggested that the extension service could play an important role in facilitating land purchase credit by using its farm-group approach to pool resources and buy large pieces of land for subsequent division among group members. NORTHiAST bRAZIL RIO GRANDE DO NORTE RURAL DEVELOPIENT PROJECT Income Levels - Selected Performance Indices by Farm Size (1973) Farm Size hanges (ha) 0-9.9 10-l49.9 50-99.9 100-200 Percent of total receipts Crops 79 65 75 66 Live stock 19 3h 24 314 Othr7rs 2 1 1 - Percent output from crops 83 65 81 70 Percent output from livestock 17 35 19 30 Percent output marketed 13 47 54 58 Purchased inputs (% total costs) 30 16 16 12 Output/man (Us$) 376 589 850 1,224 Output/hectare (U:S4) 65 41 29 19 i:t. income (US$) 1119 612 1,229 1,1159 3otuce: SIUDEj-E-IBIRD/DRC Survey (1973). Information refers to zone C of the Survey, the Semi-Arid Sertao, representatire of Llbe project area. NIORTif!IlS BRAZTL RIO GRANDS DO NORT RURAL DEVEIDPMENT PROJECT Phasing of Investment (Cr$ mill ion) - - - - - - - Year- - - - - - - - - - - - - - - - Credit (Table 1) 1 2 3 1 5 Total Seasonal Credit . .7 2.9 2.6 4.2 6.5 16.9 Cotton establishment .8 2.0 3.14 5.3 8. 19.9 On farm investment 3.7 5.14 6.0 9.6 146 '39.3 Agricultural Extension (Table 2) 3.6 2.5 3.3 5.3 9.2 23.9 t!?perimentation a id Farm Trials (Table 3) 1.8 3.2 3.7 5.o 6.o 19.7 Farmers Develop mnt (Table )1) .2 .3 .3 .4 .6 1.8 Health (Table 5) 1.3 2.2 3.1 3.1 3.2 12.9 Social Extension, Self-Ielp Demonstration .2 .2 .3 - - .7 (Tbe6) Project Administration (Table 7) 2.2 1.9 1.9 1.9 1.9 9.7 ProjecL Special Fund .5 .5 .5 .5 .5 2.5 Project Peparation Studies .5 .8 .7 - - 2.0 Monitoring-Ovaluation (T-ble 7) 1.1 1.1 1.3 1.1 1.1 5.7 L Estimate 16.5 23.0 27. 1 36.11 52.0 155.0 Pi-Jce iontingenc ies 2.3 5.5 8.9 16.0 29.1 61.8 Total Expected Cost of Project 18.8 28.5 36.0 12.4 81.1 216.8 . ....,:2oa_eJ..n, u i-u too, p,<-t)j.s - LInClWUd in c. 11dUU1r o 1' t lue V a Uov 4.130poUJIts - .- .t k I.n c t aininr costs, ext-acted from the above, is shown in Tabh] 8. Octobr 2, 1975 NORTHEAST BRAZIL RIO GRANDE DO NORTE RURAL DEVELOPMENT PROJECT DISTRIBUTION OF SUBLOANS BY FARI SIZE 1976 - 1978 Number and value of loans Percentage of total loans Year .(50 ha ;50 .ha (50 ha >50 ha No. Value No. Value No. Value No. Value 1976 Investmenta- Sub-Total 1976 360 4,980 176 - 8,787 67.2 36.2 32.8 63.8 1977 Seasonal Sub-Total 1977 177 932 64 1,625 73.4 36.5 26.6 63.5 1977 Investment Sub-Total 1977 426 8,425 257 11,617 62.4 42.0 37.6 58.0 1978 Seasonal' Sub-Total 1978 330 3,863 158 6,234 67.6 38.3 32.4 61.7 TOTAL 1,293 18,200 655 28,263 70.0 38.3 ! 30.0 61.8'/ tdi n/Starting in July. Since the investment cycle starts after the cotton harvest in July and August, the invest- ment credit of 1976 probably would not have been much greater if the project had started operations at the beginning of the year. b/Through April. c/Given that the cruzeiro values are current rather than constant, these value percentages are calculated as simple averages of the four sub-total. percentages. Source: Project Unit, CEPA/RN 4 . 1дКП 11ид4 ._..�_,,.._._..�_..�..�____�_._7_.__.--..,�_., _.(-�'-'.` .1 аииии3�пб еи� т ю' С( г L BRAZIL � ^� . лгыит � '`:,, R10 GRANDE DO NOR7E ,�Х ..и�.«.гг�«о.П«� � '�+ь� Сwlпи3агно ,о 1о w ли su i�URA�. N�1�1L П�ПНL �4VLL�Г1ё1E1Vt ! RVJE�tl еsяггс ,п r1 � C,U11UП li�П3 ,l_�..... L. ь._ 1. ... _1.. .. L�J f -. � ь ь. .. FhmuXAroш 3.11йMf1tA1 �.• е иwиш. Н --- sиь-агао3 Pro�ect Агва, RP�iпfoll & lnhostrucPurв �� --- а,иьиlгRииа3 л �. далйQа� sоитн аGлеаlсм --- .- $исиПJигу HuaJa l)1+I4 / � ?. e�vw3 �. / � '4 5'_. lS Мипкiниl диипдагiи3 N аfг�н Т � -�•- 5tutы 9оьпдипь. I Q� � � мсСди - •-✓-. `^. BЛN1.1N1�1JEPWURK /� � ( � 7..иw.и.,Jи,,...о...яь.ш МОцUАс� нг+s .�.ь.�мw м нц+хг N йе 8�опсhпs \ . . -9 .иьиа.,дшнг.г.. � Bmrco Go Brutil еС' 7'�дг+(ЕгR.7 '!+J riU�'РЕ , i Br3nco Оо Nлгда3га Ро дти! +� _...._ иwд}�,.,., ; � КА1[Jh�l1 �- � , / 1 ~ � ieo}iyeri (mП,tlhmafылj % . �i1 � ` � / � 1 / �,� --------.� ,---,, � f ��' • � �` �: С, � /� ль, 7 � ` `-7 tоьолиеtииз ( � / рР � � � г / 7о.+оглмада I �- � i ''-I / �� 1 � Ei 1 ,�" �Г„ �� ''е ! � Х r.�w ; адва мiвiм I� р� �� ина�..� г ■ pucim5 '1 ( �и,1 3 ('3 . v � �г,/ � �,, �/,��_ 4_ О �1 3и, А.«ы.м ,� �j иатаL � / l О l А чи, а � ю. диг.д 4 ,� � � с., о ь. - p� ' жг �м ➢ � ( � i 5...�о ) �� ' ; ' . 1 iа.Л9' г... J � � � , � .f . 'f �фё«ŭw � . r /� 1 1 _�� � ги«.и.,���,. z 5йо 3оп' �мnRria3'' _ lос.дш. � 3.«�ш. _ �.-� !1 �, а. / � ( t � �ч<д ri ��� �� �'' t . / //` 1 �/ ,� ' г , � 1 � l^� ' iy�p Г i � � ��л 4«�� 1 ( 3П�' 1]GJO� \ . `v3b.:J. � '.V/� ` �L.к{г� � Рл1и { % �iJ/I _ ♦.Jl� � � � M�Wkи ■+ Аи 1k75 rtинUS � .�А�м ` /£3аиrдиiа /Т I4���` / � � �О � ! ндЕцS�юеиtЕ } . Р о � ,/' l 4 ' ) 1 r ti 1 �\ ю 1сг� ! J 1- .- 1 � ..i�(., ` � � � ' ганGаАа 1 ` i /1 амлп� гй«ги. _ ! `' � � , � ! % А � 5ай3л сн�ц вло 7os¢ О 1 , !1 1" у� г�I п 1 �' f \� � �^"� lиААацг.r,и',: � [�мPCSti� ` 4 Д.. . I, У � � \ �'�I � 1 'r..�'С -"1 1 � 7ии . и�.р ди r j Г �АСл1иtЕ �" Сгы ' � 2 t �.• l , ``..� `` А � . /1 /' ио�3,сиа ь+a�a.n.wn, ��� � ai.,a ,и f : � � �1 е f лс i �V �, � � o.+�ww�w�, . � геипгое ci мида а� '-`,-. 1 гаiео '��s„er•�o� �S ьа, : г ча еиш f' �-� � в..�, s�.,,. . �' 1 �l ' '-.� i � ..!-..��.г� �`�:..-�.,� оаии.ь«, ! J. '✓"1.. , I>�' f-. �,� ° ' '�..._.. -�•. ..-` / : i �` � � ' 3 '` � '' '' / Гл. � J�^-�` �згаадгдьАа И---, � '�-�1аи��мгюsевiПо� �, Ч]РГ ( � ` Ir7 PUlrenг �S wйu `� � \ ! ��йЕlМа5 н _ :�s.ь.,, ��, � �� � � � ,~ r /о r r'-�� . �_� А � 7В' � 37' � � ]а' �5•. .� ._ � .�_.._ _�_ �