Colombia - Second Atlantico Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 4579 PROJECT COMPLETION REPORT COLOMBIA SECOND ATLANTICO DEVELOPMENT PROJECT (Loan 849-CO) June 23, 1983 This document has a restricited 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: Colombian Peso At appraisal: 1972 US$1.00 = Col$ 21.0 Appraisal year average: 1972 US$1.00 = Col$ 22.8 Intervening year average: 1972-1981 US$1.00 = Col$ 42.5 Completion year average: 1981 US$1.00 = Col$ 54.5 WEIGHTS AND MEASURES Metric System ABBREVIATIONS CAC - Cooperaci6n Agropecuaria del Caribe Caribbean Agricultural and Livestock Cooperative CECORA - Central de Cooperativas de Reforma Agraria Central Cooperative for Agrarian Reform HIMAT - Instituto Colombiano de Hidrologia, Meteorologia y Adecuaci6n de Tierras Colombian Institute for Hydrology, Meteorology and Land Improvement ICA - Instituto Colombiano Agropecuario Colombian Agriculture and Livestock Institute INCORA - Instituto Colombiano de la Reforma Agraria Colombian Institute for Agrarian Reform SENA - Servicio Nacional de Aprendizaje National Training Service FISCAL YEAR January 1 - December 31  FOR OFFICIAL USE ONLY PROJECT COMPLETION REPORT COLOMBIA SECOND ATLANTICO DEVELOPMENT PROJECT (Loan 849-CO) TABLE OF CONTENTS Page No. PROJECT COMPLETION REPOR1' Preface ............................ .......................... i Highlights .................................................... ii Basic Data Sheet .............................................. iv I. Background. 1 II. Description of the Project. 4 III. Project Implementat:ion. 5 IV. Organization and Management. 9 V. Financial and Economic Returns .10 VI. Performance of the Borrower and the Bank .13 VII. Special Issues and Lessons Learned .15 Annex 1 Comments from t:he Borrower ..................... ..... 19 TABLES: 1. Cost Table 2. Disbursements and Allocation of Loan Proceeds 3. Land Acquisition, 1962-81 4. Land Distribution and Tenure in 1981 5. Appraisal Estimates and Actual Cropping Patterns 6. Crop Production 1972-81 and Projections 1982-92 7. Estimated Net Returns from Commercial Crops 8. Cattle Component of INCORA Settlers: Cost-Benefit and Cash-Flow Analyses 9. Cattle Component of Private Farms: Cost-Benefit and Cash-flow Analyses 10. Net Agricultural Benefit Attributable to Project 11. Cattle Component: Total Project Area 12. Cattle Component for INCORA Settlers and Private Farmers: Comparison of Appraisal and Completion Report Estimates 13. Operation of Boquitas Pumping Station CHART: Rainfall and Evaporation MAP: IBRD 3245R3 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.   PROJECT COMPLETION REPORT COLOMBIA: SECOND ATLANTICO DEVELOPMENT PROJECT (Loan 849-CO) PREFACE This is a Project Completion Report (PCR) of the Second Atlantico Development Project in Colombia, for which Loan 849-CO, in the amount of US$5.0 million, was approved on June 29, 1972. The loan was reduced to US$2.2 million in February 1977 and closed on September 30, 1981, after a delay of two and a half years. The final disbursement was made in January 1982 and a further US$0.6 million was cancelled, so that the final loan amount was US$1.6 million. The PCR was prepared by Agricultural Division 2 of the Latin America and Caribbean. Projects Department following a mission to Colombia in March 1982 to collect final project data and evaluate project achieve- ments in the field. The draft report has been read in the Operations Evaluation Department (OED). However, an audit of the project by OED staff has not been undertaken. The draft report was sent to the Borrower--INCORA, HIMAT and the Ministries of Agriculture and Finance--for comment on April 14, 1983. Comments received from the Ministry of Finance and INCORA are attached as Annex 1 and have been taken into account in the final version. The valu- able assistance provided during preparation of this report by the staff of the Borrower, INCORA and HIMAT is gratefully acknowledged.   - ii - PROJECT PERFORMANCE AUDIT REPORT CODLOMBIA SECOND ATLANTICO DEVELOPMENT PROJECT (Loan 849-CO) Basic Data Sheet Appraisal Actual or Actual as X of Key Project Data Estimate Estimated Actual Appraisal Estimate Project costs (US$ million) 9.69 9.60 99 Loan amount (US$ million) 5.0 2.20 a/ 44 Date Board approval 06-20-72 06-29-72 Date effectiveness 11-24-72 11-24-72 Date physical components completed b/ 12-31-77 12-31-80 155 Proportion then completed (1) 100 50 50 Closing date b/ 03-31-78 09-30-81 161 Economic rate of return (X) 23 negative c/ - Financial rate of return (Z) - Institutional performance good poor Agronomic performance good poor Number of direct beneficiaries 1,800 (1977) 1,252 (1982) 70 Cumulative Disbursements FY73 FY74 FY75 FY76 FY77 FY78 FY79 FY80 FY81 Appraisal estimate (USS million) 0.4 1.9 3.5 4.5 5.0 2.2 a/ 2.2 2.2 2.2 Actual (US$ million) 0.1 0.2 0.3 0.4 0.6 0.8 1.3 1.5 1.6 Actual as X of estimate 25 10 8 9 12 36 59 68 73 Date of final disbursement 01-20-81 Principai repaid to 05-31-82 (US$ million) 0.25 No. of Mandays Specializations Performance Types of Mission Data Date Persons in Field Represented d/ Rating e/ Trend f/ Problems Appraisal 10/70' 3 33 abc Reappraisal 02/72 4 30 abbe Subtotal 63 Supervision 1 08/72 1 4 c 2 - 2 02/73 3 21 bbb 3 2 M 3 08/73 3 24 bbc 2 1 T 4 06/74 1 2 c 1 2 1 5 03/75 1 10 c 2 2 M 6 10/75 1 11 c 2 2 M 7 03/76 1 3 c 2 2 n 8 01/77 2 12 ad 2 1 M 9 08/77 1 6 a 3 3 MF 10 10/77 1 1 a - - 11 11/77 1 11 a 3 3 MFT 12 06/78 2 24 ad 3 3 MFTF 13 11/78 1 4 a 3 3 HFTP 14 01/79 1 1 a 3 3 KFTP 15 06/79 1 4 a 3 1 TF 16 08/79 1 10 c - - 17 11/79 2 14 ac 3 3 TF 18 03/80 2 3 ac - - TF 19 09/80 1 1 a 3 2 TF 20 06/81 1 1 c - - 21 06/81 1 1 b - Subtotal 168 Completion 03/82 3 27 abc Other Project Data Borrower: Republic of Colombia Executing Agency: INCORA Fiscal Year: January 1 - December 31 Name of Currency: ColS Currency Exchange Rate: Appraisal Year Average US$1.00 - Col$ 22.8 Intervening Years Average US$1.00 - Col$ 42.5 Completion Year Average US$1.00 - ColS 69.2 Previous Project: Name: Atlantico Irrigation Project Loan Number: Loan 502-CO Loan Amount (US$ million): 9.0 Date Board Approval: 06-27-67 a/ US$2.8 million cancelled on February 7, 1977. b/ Calculated from Board approval date. c/ Actual rate of return cannot be calculated since project costs in all years exceed incremental benefits. d/ a - agriculturalist; b - ecolomist; c - engineer; d - financial analyst. e/ 1 - problem-free; 2 - moderat:e problems; 3 - major problems. 1 - improving; 2 - stationary; 3 - deteriorating. F - financial; H - managerial; T - technical; P - political; 0 - other. June 25, 1982   - iii - PROJECT COMPLETION REPORT HIGHLIGHTS 1. The Second Atlantico Development Project (Loan 849-CO) was appraised in October 1970 and reappraised in February 1972. It was approved in June 1972. This project was a follow-up to the Atlantico Irrigation Project (Loan 502-CO), whose progress was unsatisfactory at the time of appraisal of the second-stage project. 2. The purpose of the new project was to develop the remainder of the project area avoiding the mistakes made under the first project. Thus, instead of an irrigation project, a rainfed farming development was foreseen on 17,000 ha (para 1.03). Through completion of a drainage system and construction of farm roads in an area protected against flooding under Loan 502-CO, extensive livestock farming was to be converted into intensive crop farming. The Bank loan was to partially finance: (a) construction of civil works and project buildings; (b) purchase of equipment for operation and maintenance and farm equipment; (c) foreign consultants and foreign training; and (d) interest on the loan accrued before April, 1977 (para 2.01). 3. The design for the main drainage system for the project area was technically correct, but it was not properly managed; its construction was very protracted; no funds were provided for maintenance during construction (para 4.03); farmers are not prepared to pay drainage charges to finance maintenance until the system is entirely satisfactory; the main interceptor drain was breached after an exceptional storm in 1979, possibly because of inadequate maintenance; the flooding in 1979 increased farmers' dissatisfaction with the drainage system; and the whole area remains predominantly as grassland with scattered 'shrubs. 4. The agricultural system proposed at appraisal and used to justify the project economically was incorrect, because it paid insufficient attention to the climatic and soil conditions and the experience and concerns of the farmers. Two crops a year were expected to be grown over 73% of the drained area and the balance was to be used for grazing, and growing sorghum for silage and forage crops for intensive milk and meat production (para 1.03). At the completion of disbursement 86% of the project area was still used for extensive grazing, with the balance under subsistence crops (para 3.05). 5. Farm incomes have increased during the project, but it is considered that this has been almost entirely the result of the improvements made to roads adjoining the project area and the recent improvements in the price paid to farmers for milk (para 5.09). The farm area allocated to each family was 8 ha, based on a budget in which most of the income was to come from the sale of crops. Since the farms are essentially used for cattle raising, it has been found that the farm size is not sufficient to support a family without off-farm income (para 6.05). Since 22% of the land allocated by INCORA was abandoned (para 5.02), a de facto increase in farm size has taken place through redistribution of the abandoned land among those farmers who remained in the project area.  - iv - 6. On-farm land improvements were not carried out by INCORA, as planned, because during the project the Government determined that all such works should be carried out by farmers with the assistance of credit (para 3.04). This proved to be an entirely unworkable system because the farmers were unwilling to undertake such risk. Furthermore, most farmers had bad debts from earlier years, accumulated in the cooperative formed under the first stage of the project (Loan 502-CO). This cooperative collapsed half way during the second stage project and the agricultural bank would not extend further credit without a guarantee from INCORA, but INCORA had fully committed its guarantee fund elsewhere in Colombia. 7. The economic rate of return estimated at appraisal was 23%. Now, it is almost certain that project costs in each year will exceed net incremental benefits so that no rate of return can be calculated (para 5.11). 8. The Government will have to improve and maintain the civil works already constructed and minor additional works are required to protect the limited area (800 ha) of good soils from flooding (para 7.10). Operational costs can be reduced by limiting INCORA staff to those required to provide technical assistance in the only viable field of milk and meat production, for which a large margin for improvement still exists (para 5.13).  I. BACKGROUND 1.01 This is a Project Completion Report (PCR) on the Second Atlantico Development Project in Colombia, for which Loan 849-CO for US$5.0 million, was approved in June 1972. The loan was closed nine and a quarter years later (September 1981) after US$3.4 million, or 68% of the original loan amount, had been cancelled. 1.02 This projecl: was intended to complete a program initiated by the Colombian authorities in the 1960s, to alleviate rural poverty in the project area (para 2.01 and Map IBRD 3245R3). The Bank has been associated with the program since June 1966 when a mission visited Colombia to evaluate the managerial capacity of the Colombian Institute for Agrarian Reform (INCORA) to undertake a 35,000-ha project, and the suitability for irrigation of the soils in the area. 1.03 The first stage of the program was supported by Loan 502-CO made in 1967 for the Atlantico Irrigation Project to develop 3,900 ha under irrigation and to protect a further 6,000 ha from flooding. Experience with irrigation development under the first stage had been negative and therefore the second stage of the program, which constituted the Second Atlantico Development Project, was to extend by 11,000 ha the area protected from flooding, to a total of 17,000 ha suitable for growing two crops each year by dry farming. For the purpose of economic analysis of the second stage, it was projected that: (a) 12,400 ha would be used to grow two crops a year: sorghum, maize and soybeans in the first semester, and sesame, maize, soybeans, cotton, groundnuts, and tomatoes would be grown in the second semester; at a later stage, beans, tobacco, onions, melons, and various fruits and vegetables might also be included; and (b) 4,600 ha would be developed as dairy farms that would initially have field crops, forage and pasture crops, but gradually the field crops would be phased out as the dairy enterprises reached ful:l development. 1.04 The proposa:Ls for the second stage of development were based on a project preparation report and background material prepared by INCORA with the assistance of consultants and on the findings of a Bank appraisal mission which visited Colombia in October 1970. These proposals were questioned extensively wiithin the Bank and Colombia; therefore, a further Bank mission visited Colombia in February 1972 to update the appraisal report, which was, however, not materially changed from the initial proposals and the appraisal was finalized as Report No. PA-84b, issued on June 2, 1972. This was shortly after the Operations Evaluation Division (OED) of the Programming and Budgeting Department had issued Report No. Z-18 of May 25, 1972 concerning "Bank Operations in Colombia: An Evaluation." The OED report (pages 136-142) questioned the technical assumptions underlying the proposal for the second stage. It pointed out that: "Another upward revision in the minimum size family unit may be required in the dry farming part, where most project beneficiaries are located, in order to reach net family incomes projected at appraisal, unless a more profitable cropping pattern can be effectively developed and implemented", and concluded "that the Bank  -2- financing for this project (program) should have been confined to the provision of flood control, some drainage, and a minimum of supplementary irrigation, and should have followed, rather than anticipated, an effective program of land acquisition and distribution to the landless peasants". During presentation of the Second Atlantico Development Project to the Board of Directors on June 29, 1972, the difference of conclusions between the appraisal report and the recent OED report was raised by several Directors; however, the staff assured the Board that the technical assumptions in the appraisal report were well founded. 1.05 The Project Performance Audit Report (No. 1841) (PPAR) on the first stage (Loan 502-CO) which was issued on December 28, 1977 concluded that, because of improper project preparation and appraisal and delays in project implementation, the economic results of the project (first stage), were unsatisfactory, but that the project had a positive impact on farming incomes and living standards of project beneficiaries. 1.06 In April and October 1981, OED missions visited Colombia to prepare an Impact Evaluation Report (IER) on the first-stage project (Loan 502-CO) because: (a) it was the first Bank-financed irrigation and land settlement project in Colombia; (b) the implementation time overrun of 57% had been mostly because of technical problems; and (c) the economic and social impact of the project had been uncertain at the time the PPAR was prepared. The IER on the first stage was prepared just as the loan for the second-stage project (Loan 849-CO) was about to be closed. Therefore, the conclusions of the IER in Report No. 3959 distributed on June 21, 1982 are of direct relevance to an evaluation of the second-stage project, in particular since the IER report treated the areas of loans 502-CO and 849-CO as one unit. 1.07 In addition to the assessment's of the PPAR (para 1.05), the IER drew particular attention to the following points: (a) information and data provided by the consultants were not sufficiently reviewed by the Bank or adequately checked for accuracy, completeness and consistency; (b) the Bank's requirement that consultants be responsible for (first stage) project management was not justified; (c) the failure of the production cooperative, promoted by the consultants, severely dampened the cooperative spirit of settlers and demonstrated the risk of (i) introducing systems modeled after those of other countries and not adapted to local socio-economic conditions; and (ii) developing a settlement scheme that left little initiative to settlers and which was not based on a concept of "self-help"; and (d) settlers, who were initially considered to be the poorest segment of the population, turned out to be a privileged group compared to squatters who entered the project area.  - 3 - 1.08 The Colombian authorities accepted the general conclusions of the IER, but drew attention to the following issues, which in fact relate to both stages of the project.: (a) the Bank should have taken a more comprehensive view of regional development, rather than focus on a relatively restrictred project area; (b) as the initial feasibility studies were evidently inadequate, why did the Bank not redesign the project? (c) the Bank should have insisted on the provision of adequate counterpart financing and included in the project the costs of operation and maintenance until the whole scheme was in full operation; (d) the Bank should facilitate transfers between loan categories during project execution so as to make it possible to amend the project Ln the event of changes in the conditions that were the basis for the loan; (e) the technical problems of draining the project area have still not been solved; (f) the failure (i) to establish an experimental farm; (ii) to complete the interconnection with the national electrical grid; and (iii) to overcome the resistance of the local farmers to take part in any form of cooperative marketing are all points that should receive attention by the Bank; and (g) although the PPAR and IER comment favorably on the positive impact of the agrarian reform work of INCORA, and report that living standards for many families are now better than before the projects, the social situation in the project area is becoming daily more critical, and serious problems are expected in the short term, unless Government provides a final solution to the recurrent flooding problem suffered by the area. This has reduced the farmers to a subsistence existence, swollen INCORA's portfolio of bad debts, and made it impossible to recover the investment made in the civil works. 1.09 The present PCR sets out to respond to some of the questions raised in the previous paragraphs and to bring out lessons to be learned by both the Bank and the borrower, particularly from implementation of the second stage of the Atlantico development program. Details contained in the reports on the first-stage project (appraisal, PPAR and IER) are not repeated in this report, which should be read in conjunction with the earlier reports for a comprehensive overview of the program.  -4- II. DESCRIPTION OF THE PROJECT 2.01 The project comprised the second stage of a flood control and drainage program for an area bounded by the Magdalena River, the Canal del Dique, the Guajaro Reservoir and the Manati Hills in the Atlantico Department of Colombia (Map IBRD 3245R3). The project was to complete the development for settlement and dry farming of about 17,000 ha of cultivable land, of which about 12,400 ha was expected to be used for crop farming and about 4,600 ha for dairy farming. The 17,000 ha included 6,000 ha protected from flooding under Loan 502-CO and for which project benefits had been included in the economic evaluation for Loan 502-CO. The project consisted of the following components, which were to be completed by the end of 1975: (a) design and construction of five main drains (51 km), 31 lateral drains (60 km), and sublateral ditches and waterways to complete the drainage for about 17,000 ha, of which 6,000 ha had been drained during the first stage (Loan 502-CO); (b) design and construction or improvement of about 77 km of roads, including four secondary roads and a number of tertiary roads, to provide access to the farmland of the project; (c) acquisition by INCORA of approximately 17,000 ha; land clearance, rough leveling, initial deep plowing and then allocation of 8 ha each to about 1,800 farmers by 1977, of whom 740 farmers had already been settled on 5,600 ha in the first stage; INCORA was to treat saline spots that developed after the final allotments had been made; (d) design and construction of five depots, each of about 1,000 m ton capacity, for the storage of agricultural produce and supplies of farm inputs; (e) expansion of INCORA's headquarters at El Limon in the project area; (f) (i) purchase of vehicles and equipment for the operation and maintenance of facilities constructed under the project; and (ii) purchase of farm machinery, including about 100 tractors, for use by about nine private contractors on project lands and to be supervised by a board formed by INCORA and the Central Cooperative for Agrarian Reform (CECORA); (g) (i) the preparation of an agricultural research and extension program for the project area in close collaboration with the Colombian Agriculture and Livestock Institute (ICA); and (ii) evaluation of irrigation trials on the Manati pilot farm which was to have been established under Loan 502-CO;  (h) preparation of a training program for the 1,800 farmers to be settled within the project area in close collaboration with the National Training Service (SENA); and (i) training for staff of INCORA, totalling 12 man-years, in the fields of operation and maintenance of flood control and drainage systems, mechanized dry farming and project administration. 2.02 The description of the project in paragraph 2.01 combines details given in the Loan Agreement 849-CO'(Schedule 1 and Schedule 2) and in Appraisal Report PA-84b (paragraphs 5.01 to 5.07; and Annexes 6, 8 and 10) because the "Description of the Project" in Schedule 2 of the Loan Agreement,when read alone, does not clearly define the project, particularly concerning: (a) the participation of private contractors; (b) the carrying out of irrigation trials; and (c) the extension of the consultancy service contract. 2.03 The loan for the second-stage project was approved a considerable time before the first stage (Loan 502-CO) was completed. During the period in which both loans were effective, it was found that there were funds available in Loan 502-GO to construct two of the five main drains and 62 km of the 77 km of roads (Loan 849-CO Category I) and to purchase all the equipment (Loan 849-CO Category II) included in the second-stage project. Also during this period, in 1974 the National Planning Department of Colombia (DNP) established the policy that on-farm land improvement, such as clearing, leveling, initial plowing and digging sub-lateral or tertiary drains, should be financed through loans to be taken out by individuals or groups of farmers. The Agricultural Financing Fund (FFAP) in the Bank of the Republic (BOR) had resources available for this purpose which were supposed to be channelled to INCORA's beneficiaries, as well as to other farmers, through Caja Agraria. Funds from Loan 1357-CO were available for this purpose from 1977 onwards. 2.04 The appraisal estimates of project costs and financing are summarized in Table 1 and the initial allocation of the proceeds of Loan 849-CO and the projected disbursements schedule are given in Table 2. In February 1977, the amount of the loan was reduced by US$2.8 million for the reasons given in paragraph 2.03 and explained in more detail in the following section. The actual costs and disbursements are also shown in Tables 1 and 2. III. PROJECT IMPLEMENTATION Civil Works 3.01 From November 1972, when Loan 849-CO became effective, until the end of 1975, civil works continued to be financed by Loan 502-CO. Subsequently, contracts were financed from Loan 849-CO. Construction proceeded much slower than expected, in part because INCORA did not group all civil works into no more than four contracts, as required by the Loan Agreement. I/ The Bank agreed to this change to encourage the development of 1/ Compliance with Loan conditions is summarized in paragraphs 6.02 to 6.04.  - 6 - small local firms, but unfortunately, on two consecutive occasions, contractors failed to complete their tasks and work was suspended, each time for over a year, until the necessary legal steps had been taken to cancel the contracts and select new contractors. INCORA's capacity to supervise the contractors was severely limited for the following reasons: (a) frequent changes in Project Manager, without the prior agreement of the Bank to the nomination of the new manager; (b) failure of INCORA to continue to employ consultants to assist in the design, procurement and supervision of construction of the civil works; and (c) INCORA staff being housed principally in Barranquilla because of insufficient housing and social facilities in the project area. 3.02 In 1976, the Colombian Institute for Hydrology, Meteorology and Land Improvement, (HIMAT) received nationwide responsibility for the construction, operation and maintenance of all civil works within irrigation and drainage areas. The majority of the works eventually financed from Loan 849-CO were constructed between 1977 and 1979. These totalled 40.2 km of drains and 23.7 km of roads. Land Acquisition, Settlement and Development 3.03 During the second-stage project, INCORA purchased 2,068 ha which, together with the land purchased in the first stage (6,524 ha) and 5,026 ha reclaimed from former swamplands, but not purchased, brought the total available for allocation to 13,618 ha (Tables 3 and 4) although not all of this area was suitable for farming. During the first stage, INCORA had settled 740 "parceleros," or settlers, on 5,600 ha. At the conclusion of the second stage there was a total of 1,252 settlers on 11,262 ha or 70% of the 1,800 expected at appraisal. In addition, in 1982 there were some 660 unauthorized "squatters" on 1,285 ha, each farming about 2 ha for the production of subsistence crops from the relatively fertile soils of several seasonal swamp areas. In the original design, these low zones were to serve as flood retention areas from which surplus water would eventually be evacuated through the pumping station "because it was found too costly to lead all drain water to the main system" (Appraisal Report, Annex 6). At the conclusion of the second stage of the development program for the project area, about 9,000 ha remained in about 400 private farms unaffected by agrarian reform (average 23 ha/farm). The distribution of land ownership among these farmers is relatively equitable, with few having more than 100 ha. 3.04 INCORA did not carry out the planned initial bush clearing, rough leveling and deep plowing prior to land allocation, and only about 2,100 km of tertiary drains were dug out of the estimated requirement for 32,000 km. This was because the Government decided after loan signature that these works should be the direct responsiblity of the settlers (para 2.03). The Bank repeatedly objected to this decision b cause:  - 7 - (a) much of this work could be done only by machinery contractors, who would not be interested in 8-ha units; (b) many individual settlers already were in arrears with payments due to Caja Agraria; (c) following the failure of the Caribbean Agricultural and Livestock Cooperative (CAC), there was even less interest by settlers in group action; (d) in order to make use of funds from the FFAP (para 2.03), Caja Agraria would require a lien on the settlers' land and this is prohibited by the Agrarian Reform Law; and (e) subsistence farmers tend to avoid risk. Lack of clear demonstration that tertiary works would be beneficial resulted in refusal by farmers to incur further debts to carry out such works. 3.05 The appraisal foresaw to convert extensive crop farming and livestock farming (on 58% of the area) into intensive double crop farming and intensive dairy farming, the latter on a reduced area (27%) (see Table 5). Poor climatic conditions and soils, however, presented an unacceptable level of risk, resulting in a steady decline in the area under cultivation (see Table 6). By 1981, extensive livestock farming occupied 86% of the area and susbsistence crops 11%. Storage Depots 3.06 The proposed construction of five farm produce and input storage depots was cancelled in 1977, following the collapse of CAC and because private traders were adequately handling the limited agricultural output from the project area. Vehicles and Equipment 3.07 All vehicles and equipment planned for the second stage of the program (Loan 849-CO) were purchased with funds from Loan 502-CO which was for the first-stage project. Agricultural Research and Extension 3.08 As part of the first stage project, a pilot farm of about 320 ha was to be established between Manati and the Guajaro reservoir; 190 ha was to be for crops and 130 ha for pastures for dairy production. The farm was to provide facilities to study supplemental irrigation with water from the Guajaro reservoir and from the drainage system using low-lift mobile pumps. The farm was not established during the first stage, because ICA was unwilling to assume additional responsibilities while it had serious financial problems in operating existing research stations throughout Colombia, including the sub-station at Santa Lucia in the project area.  - 8 - During the second stage, even the Santa Lucia sub-station was closed by ICA as part of its retrenchment program. On several occasions INCORA tried to prepare with ICA a realistic agricultural research and extension program, but this was not possible because of the financial problems of ICA and because the Manati pilot farm had not been established during the first stage project. The appraisal report (para 6.04) states that "INCORA would be responsible for agricultural research and extension in the project area, with technical backstopping from ICA," but there was no financial provision for these activities in the cost estimates of the project, other than the equivalent of US$30,000 of local costs for unspecified irrigation studies on the Manati pilot farm. Farmers Training Program 3.09 The second stage project did not provide funds for training farmers, but the Loan Agreement required ICA to draw up a program in collaboration with SENA. Although INCORA attempted to meet this obligation, uncertainties about financing the activity and the technical recommendations to put forward made it very difficult to fulfill. Throughout the execution of the project, SENA staff conducted short training courses for farmers, but these inevitably did not constitute a coherent program and the impact of the courses was not evaluated. Training INCORA Staff 3.10 During the first stage project, training of INCORA staff had been carried out in Mexico, Peru, Israel and Iran and it was planned that, during the second stage, training would be in similar locations, as well as in Colombia. Early in the execution of the second stage the Government severely restricted foreign travel by civil servants. INCORA therefore proposed that the loan funds available for training purposes be used for all types of local instruction for INCORA staff throughout Colombia. The Bank agreed to this change and courses were carried out greatly in excess of the equivalent of 12 man-years of instruction, but full details of the courses and an evaluation of their effectiveness is not available. Only 52% of the amount available for training had been disbursed at the closing date. Consultant Services 3.11 At appraisal it was expected that INCORA would renew the contract with consultants who were assisting in the development of the engineering and agricultural aspects of the first stage. However, INCORA was not prepared to extend the contract beyond June 1975, because it considered that the technical advice during the first stage had not been fully satisfactory and, in particular, the formation of production cooperatives, promoted by the consultants, had been ill advised. These views have subsequently received support in the IER which concluded that, for the first stage, the Bank's requirement for consultants to be responsible for project management was notjustified. However, as noted in paragraph 3.01 (b), the lack of experienced civil engineering consultants was one factor contributing to the erratic and slow completion of the drainage system during the second-stage project.  IV. ORGANIZATION AND MANAGEMENT 4.01 The Atlantico regional office of INCORA was responsible for managing the whole project and also the Repelon irrigation scheme to the West of the Guajaro reservoir and some other areas to the North of the project area. The periodic reports and studies prepared by the office were generally on a regional basis so that it is almost impossible to isolate the costs and benefits attributable to Loans 502-CO and 849-CO. Within the project area, INCORA staff was not able to obtain technical or quantitative information from individual private farmers, occupying 9,000 ha, who continued to believe that such enquiries were the initial stage of expropriation proceedings. 4.02 The project was implemented while the future role of INCORA and official policies for agrarian reform and settlement were under continuing but inconclusive debate. Funds for the institution were curtailed and this caused particular problems for the Atlantico project, which had a lower priority than many other INCORA projects. The project also suffered from frequent changes in management; there were five managers in the period 1973-81. 4.03 At the height of its operations in the Atlantico region, INCORA had over 300 employees, including laborers, and even in 1977, by which time HIMAT had taken over responsibility for all civil works construction and maintenance, INCORA slill employed 288 persons for the stage 1 and 2 project areas, and for the Repelon scheme. In 1981, total recurrent expenditures of INCORA and HIMAT for the project area of Loan 849-CO, excluding meteorological services, were about Col$ 20.0 million ((US$364,000) and Col$ 31.8 million (US',578,000), respectively. For the 21,547 ha occupied by settlers, squatters and other farmers (Table 4), the HIMAT expenditures for operation and maintenance were equivalent to Col$ 1,476/ha or about US$27/ha. This cost Should be compared with the official annual drainage charge in 1982, which was equivalent to about US$6/ha although in fact no charges have been collected yet because farmers refuse to pay for what they consider to be an unsatisfactory service. The main operational expenditures of HIMAT are for the operation of the drainage pumping station at Boquitas, the removal of aquatic weeds and silt from the drainage ditches, and roads maintenance. Restrict:ions on HIMAT's operational budget have lead to inadequate maintenance and a serious deterioration in the condition of the drainage system. Until adequate funds are provided to HIMAT for operations and maintenance expenditures, it is improbable that there will be any improvement in the collection of drainage charges. 4.04 Under the project, but not financed from Loan 849-CO, at full development INCORA was to provide Col$ 58 million of production credit per season and, in addition, the Government was to make arrangements to provide dairy farmers in the project area with long-term credit totalling about Col$ 47 million. The supervised credit operation was not satisfactory throughout the implementation of the second stage. An important reason for this failure, as has been noted in the IER, was that as late as 1981 a number of previous CAC members were still not eligible for agricultural credit because of debts they had cont:racted, but never reimbursed to Caja Agraria after CAC  - 10 - collapsed. Settlers were supposed to sell their harvests through CAC, which would thus be in a position to recover its advances for the inputs. Instead, settlers either sold their harvests individually or left the cooperative before repaying their debts. As a result, CAC collapsed in 1977. The appraisal mission did not study this matter sufficiently to design a workable solution (para 6.06). During supervision, INCORA was asked to analyze the credit status of each settler, but by the time this review was completed in 1979, Caja Agraria was not prepared to extend new loans in the project area using its own funds and the Government was not prepared to provide additional resources for this purpose. 4.05 External audit reports were received from the Comptroller General of the Republic for each year up to 1980, but they were invariably extremely late. From 1977 the opinions were adverse, principally because there was no clear partition of the assets of the project between INCORA and HIMAT. There were, however, other material accounting weaknesses that had not been resolved when the loan closed, but which are being addressed with the help of consultants financed from Loan 1163-CO. V. FINANCIAL AND ECONOMIC RETURNS 5.01 Estimation of the physical production changes and economic and financial returns directly or indirectly attributable to Loan 849-CO is complicated by a number of factors. First, a large part of the area has benefitted from civil works undertaken under Loan 502-CO, 50% of which was disbursed after Loan 849-CO was initiated. Second, the project area has received a large, but unquantified, windfall gain as a result of the paving of two major access highways from the North. Third, the stimulus to invest in dairy operations since removal of price control on milk in 1979 must be judged as very considerable. The combination of these factors leaves a wide margin for interpreting the "without project" situation over the period 1973-92. Financial Returns 5.02 As of the date of project initiation (1973), INCORA, through Caja Agraria, was engaged in an active credit program to support commercial cropping activities within the project area. In 1972, 928 INCORA settlers had 2,750 ha under commercial crops and about 1,000 ha in subsistence crops. In addition, 450 private farmers, located on better soils, had 4,250 ha in commercial crops. No information is available on credit and financial performance of the private farmers, but their interest in pursuing risky dry land farming is attributed to high potential profits to be obtained from cotton during the early 1970s. At the outset of Loan 849-CO, the incentive for INCORA settlers to engage in extremely risky dryland cropping operations was attributable to the close involvement of the consultants in the management of CAC, supported by a poorly managed credit program. The farmers were acting as laborers for the cooperative, with their wages being guaranteed while cultivating crops at a loss. The debts they incurred through production credit were accumulated on their behalf by CAC. At the date of project start-up, settlers were already in arrears on credit by  - 11 - approximately Col$ 40 million (US$2 million, 1972 equivalent). By 1977, outstanding credit to both private farmers and settlers in the project area of Loan 849-CO over the period 1973-77 amounted to Col$ 74 million, of which Col$ 44 million was in default as of the end of 1977. These figures do not include credit granted to settlers who had abandoned approximately 2,700 ha between 1972 and 1976, many of them leaving for Venezuela. 5.03 There is no record of total credit default within the project area. INCORA records show new credit issued to December 1981 of Col$ 80 million with repayment of principal and interest amounting to Col$ 37 million. With the dissolution of the bankrupt CAC cooperative in 1977, it must be assumed that all outstanding credit at that time was written off. Further, even where repayments have been made, annual inflation rates in the order of 25%, restructuring of part of the outstanding debt in 1979 with a 1% interest rate and numerous extensions to the repayment dates, indicate very substantial financial losses to the Government credit institutions. 5.04 At appraisal it was projected that net annual farm income for INCORA settlers would increase from US$900 in 1973, to a range of US$2,700 to US$4,300 in 1982. These figures are expressed in 1981 dollars and are before tax but net of project charges (US$350 per year) and debt service. Since in reality settlers did not pay project charges and in the majority of cases did not repay their debts, their financial returns immediately before the project and until the collapse of the CAC in 1977 were probably considerably higher than official estimates. After 1977 easy credit was no longer available through the INCORA-CAC-Caja Agraria channel (paras 5.02 and 5.03), but settlers were able to obtain livestock credit through Banco Ganadero, the Livestock Credit Bank, which has a lien on cattle purchased with its loans. This further accelerated the conversion from crop farming to livestock farming, but under better controlled credit. On the assumption that credit recovery will be substantially higher in the future, the financial status of the average settler is projected to be only little better than in the "before project" situation and for 1982, before tax, net farm income after project charges and debt service is projected to range between US$1,000 to US$1,400 (see Tables 6, 8 and 9). 5.05 In absence of the civil works provided under Loan 849-CO, about 2,000 ha less land would have been reclaimed and 250 fewer settlers would have received an allotment. However, without the project, the average return to the 1,000 rather than 1,250 settlers families between 1982-92, would probably have been within 90% of the "with project" range. Gross incomes have varied widely over the 1972-81 period due to drought, flood losses and price fluctuations (see Table 7). 5.06 The situation for squatters on private and INCORA lands was even less favorable than for settlers. Their number increased from 400 to 660 over the 10-year period, but the net value of subsistence crops per family in 1981 dollars remained roughly constant at US$300 to US$600, but with wide variability.  - 12 - 5.07 The number of private farmers in the project area declined from 450 to 400 between 1972 and 1981. By 1979, the excessive risk of crop farming had resulted in a 100% conversion from mixed farming to livestock farming (see Table 6). No projection was made at appraisal of improvement in income for private farmers in the area affected by drainage and flood control works. Their incomes in 1973 are assumed to be of the same order of magnitude as the settlers at that date (para 5.04). For the period 1982-92, net farm returns from their livestock operations are projected at US$2,000 to US$2,500, compared with US$1,300 to US$1,600 "without" the project (Table 9). 5.08 Although there are difficulties in comparing estimates made in 1971 and 1982 because of changing price assumptions and conversion of nominal pesos and dollars to constant values, some indications of the changed expectations for the project between appraisal and completion can be gauged from the following: at appraisal the annual gross value of production was expected to increase from US$4.2 million (1981 dollars) in 1972 to US$23.4 million by 1990, with the net value increasing from US$1.7 million to US$8.9 million. At completion these projections for 1990 were US$5.9 million for the gross value and US$2.4 million for the net value of production, or 27% of the appraisal estimate. (Table 10). 5.09 The expected increases in production without the project are based on the assumption that credit would have been available through Banco Ganadero and that livestock output would have expanded in response to improved highway access to the project boundary, a general improvement of technology and favorable prices for milk. Economic Return 5.10 The estimated economic rate of return has been based on the following assumptions: (a) project costs and benefits are reflected by market prices, i.e., a standard conversion factor of one; (b) all family labor was valued at the prevailing market rate (Table 11); (c) all expenditures under Loan 502-CO were taken as sunk costs. Thus, contrary to normal procedure, even expenditures under Loan 502-CO incurred over the period 1973-76, which could directly or indirectly have affected agricultural production conditions on the Loan 849-CO project area, were excluded from the cost stream. However, operation and maintenance costs for all infrastructure financed by Loan 502-CO, outside the irrigation area, was included as a cost to the second-stage project; (d) both costs and benefits for the 6,000 ha, which had already been developed under Loan 502-CO, were included in the calculations (para 2.01);  - 13 - (e) INCORA land acquisition expenditures were excluded, but the project-imputed operational costs of INCORA and HIMAT for administration and maintenance expenses were included as costs; and (f) project life was taken as 20 years with incremental herd value added as a benefit in year 20. (Table 11). 5.11 Based on these assumptions, the benefit stream is negative in all years, except year 20. Expressed in current prices, the net present value (NPV) is minus Col$ 95 million (US$14 million) at a 12% discount rate or, when expressed in 1981 currency, minus Col$ 170 million (US$23 million). The only component affecting net benefits to any significant extent is the cost attributed to family labor; even with this component valued at zero, the NPV remains negative. 5.12 The poor economic performance must be attributed to: (a) the substantial shortfall in high value crop production; (b) the relatively satisfactory performance of livestock farming in the project area which is expected to have prevailed even in the absence of the project; and (c) the excessive costs of administration, operation and maintenance for civil works which, apart from roads, are believed to have had limited impact on livestock and subsistence cropping operations, jointly accounting for over 95% of production from the project area. 5.13 For the year 1981 the total cost of administration, operation and maintenance by INCORA and HIMAT was estimated at Col$ 51.8 million (para 4.03). For the same year, the net incremental benefits attributed to the project are projected as Col$ 24.5 million (Table 10). Some possibility for improvement does exist. In the first place, government's recurrent expenditures can be reduced and in the second place, the present coefficients of animal production are so low (Table 12) that a reasonable potential for increased production exists (para 7.10). VI. PERFORMANCE OF THE BORROWER AND THE BANK Performance of the Borrower 6.01 The Government, which was the Borrower, did not provide INCORA with the funds necessary to carry out the project as designed and, furthermore, did not authorize expenditures necessary for the regular maintenance (a) by INCORA and later by HIMAT, of the roads and drains in the project area; (b) by the local Government authority responsible for access roads to the project area (Loan Agreement: 4.01 (c)); and (c) the electricity authority which was to have completed the interconnection between the national grid and the pumping stations.  - 14 - Compliance with Loan Conditions 6.02 The appraisal report drew attention to the following principal points on which agreement was reached during loan negotiations. Reference is made below to the relevant paragraphs in the appraisal report (AR), the Loan Agreement (LA) and the present report (PCR): (a) INCORA would let a maximum of four contracts for drains and would promptly issue import licenses for imported equipment and materials (AR, 5.1; LA, Schedule 4, (i); and PCR, 3.01); (b) within one year of the effective date of the loan, INCORA would prepare and submit to the Bank for review, programs on: (i) agricultural research and extension; and (ii) training of Atlantico Stage II farmers and would then implement such programs (AR, 6.06; LA, 3.02(b) and Schedule 2 (f); and PCR, 3.07 and 3.08); (c) Government would submit to the Bank for comment a training schedule for INCORA personnel, giving details of individual training assignments before undertaking work (AR, 5.07; LA, Schedule 2 (g); and PCR, 3.09); (d) Government would levy a charge sufficient to cover operation and maintenance costs of the project and as much of the capital expenditures as feasible (AR, 6.12; LA, 4.03; and PCR, 4.03); and (e) Government would make adequate funds available to provide long-term credit to establish dairy farms and seasonal credit for settlers ("parceleros") (AR, 6.10; LA, 4.04 (ii); and PCR, 4.04). 6.03 Other important loan conditions were as follows: (a) before making any appointment to the position of Project Manager, INCORA would inform the Bank of names, experience and qualifications and afford the Bank reasonable opportunity to comment (AR, 6.01; LA, 3.02 (f); and PCR, 3.01 (a) and 4.02); (b) Government would make available to INCORA promptly, as needed, all funds required to carry out the project, including the proceeds of the loan (AR, 5.11; LA, 3.02 (a); and PCR, 4.02 and 4.03); (c) Government would cause INCORA to operate and maintain adequately the flood control and drainage facilities, roads, buildings and equipment (AR, 6.01; LA, 4.01 (b); and PCR, 4.03); and (d) INCORA would have its project account audited and, not later than four months after the end of each fiscal year, would furnish the Bank certified copies of the audited accounts, as well as the auditor's report (AR, 5.14; LA, 3.02 (k); and PCR, 4.05).  - 15 - 6.04 As has been indicated in the corresponding paragraphs of this report, the Borrower and INCORA were not in full compliance with any of the principal agreements reached during negotiations and recorded as loan conditions. Performance of the Bank 6.05 The civil engineering specifications proposed by the appraisal mission, except for drains M and I, were appropriate to the economic potential of the area, but the agronomic proposals were unrealistic, although they appeared to justify the expenditures on civil works. More attention should have been given at appraisal to the integration of the Atlantico 2 development program into the development activities in the region. During implementation it was soon realized that livestock production would continue to be the predominant activity in the area, but the Bank did not explain effectively to INCORA that (a) if settlers were to depend on livestock, then land allotments would have to be considerably more than 8 ha/family; and (b) squatters must be prohibited from sowing crops in depressions which were designed to serve as temporary water storage areas. 6.06 Consistent with many other early irrigation and drainage projects financed by the Bank, at appraisal attention was paid only to civil works. The design of this project did not relate the project to a regional development plan (para 1.08 (a)) and did not pay sufficient attention to the complementary institutions and services (credit and inputs; research,training and extension; and storage and marketing) that would be critical to the successful implementation of the drainage and land improvement project. 6.07 All these problems were well known by 1977 when US$2.8 million of the loan was cancelled, but the Bank did not take this opportunity to require INCORA to redesign the project (para 1.08 (b)) because, at that time, the Government provided insufficient political or financial support for INCORA, which showed little interest in the Atlantico 2 project. VII. SPECIAL ISSUES AND LESSONS LEARNED Farming System 7.01 The farming system planned at appraisal to be established throughout the project area, depended principally on double-cropping (para 1.03 and Table 5). This proposal was based on an insufficient understanding of the local physical and social environment. The soils, the climate and the history of the communities of former fishermen have together presented a particularly difficult combination of circumstances throughout the implementation of both stages of the Atlantico development program. The soils of the area are of marine origin with salinity and alkalinity increasing with depth, as well as localized saline surface patches distributed throughout the area. The soils most suitable for crop production are located in depressions that were formerly swamps and which are still liable to seasonal flooding (para 7.10).  - 16 - Rainfall distribution does not favor double-cropping without irrigation because there is a long severe dry season and, even in the normally wet season, serious droughts have occurred (Chart). Rainfall is often intense and may be localized over a part of the project area where it can cause heavy damage, unless the fields are graded and drained to prevent the crops from being waterlogged. 7.02 Execution of on-farm improvement works, such as land leveling and construction of tertiary ditches, should not be delegated to the responsibility of individual farmers (para 3.04). Most of them will avoid the risk of engaging debts for long range benefits, which are beyond their customary horizon of providing immediate security. Therefore, such works should form part of the project facilities to be executed by the authorities. The arguments in favor of this are that it provides the means to achieve as rapidly as possible the economic goal of full production, as well as the financial goal of farmers being able to pay their operation and maintenance charges at an early stage. 7.03 As was realized at appraisal, farm mechanization is essential in the area to prepare the land rapidly in the short period during which the soils are cultivable. However, private contractors were not prepared to continue land preparation as they were not paid promptly through CAC. CAC itself was not capable of maintaining and operating the machinery pool established under both projects, mainly because of its financial collapse resulting from its members failing to pay what they owed CAC and, hence, Caja Agraria (para 4.04). Other constraints were late or inadequate land preparation, unreliability of seed quality, and unavailability of fertilizers and plant protection products in sufficient quantity. Between 1972 and 1982 there was a steady decline in the area under cultivation by settlers and private farmers (Table 6), who were increasingly unwilling and unable to continue with crop farming, which represented to them an unacceptable level of risk. 7.04 Even when the Bank Loan does not include provision of agricultural credit, the general credit situation of project beneficiaries should be investigated during appraisal (paras 5.02 and 6.06). 7.05 It proved to be a failure to introduce a foreign type of cooperative. This experiment has alienated farmers to participate in any further cooperative arrangements (para 5.02). Drainage System 7.06 No provision was made for adequate allocations for operation and maintenance during construction and until the time when incremental project benefits start to accrue to the farmers (para 6.01). Colombian regulations prevented such provisions. 7.07 Reliance upon diesel power for large pumping plants requires a sophisticated organization for maintenance and fuel supply. Electric power companies are usually safer and cheaper providers of energy. The Bank should include in its loan agreements conditions to assure timely connection to the national grid (para 1.08).  - 17 - 7.08 In the feasibility study of the second-stage project, based on a 60 mm storm in 24 hours affecting the whole area, it was proposed that the nominal installed capacity of the Boquitas pumping station should be increased from 5 m3/sec to 20 m3/sec to serve the whole area. This would have required a major modification to drain 'M' to increase its discharge capacity as well as installation of additional pumps at Boquitas. The appraisal mission considered,that a more economical solution would be to retain the installed pumping station capacity of 5 m3/sec and to use four depressions as retention areas from which water would evaporate in addition to the large Boquitas swamp, which would serve as a buffer reservoir from which the water would eventually be evacuated through the pumping station. Between 1970 and 1979 the 60 mm design storm was exceeded on 22 occasions and in November 1979, after 171 mm had fallen in two days, there was a breach in the western interceptor which led to serious flooding causing extensive crop damage and civil disorders among the settlers. Before the flood diversion works were constructed during the first-stage project, about 8,000 ha were inundated annually. At the conclusion of the second stage, the area regularly flooded is 1,500 ha to 2,000 ha, except following the storm in 1979. The areas still liable to seasonal flooding are: (a) close to the main drain 'M' upstream of the Boquitas reservoir; and (b) small depressions scattered throughout the project area. The Boquitas pumps were operating at about 70% of their nominal capacity of 5 m3/sec in 1982, indicating that they were badly worn. However, these pumps were still capable of withdrawing water faster than it can flow down drains I and M for which both the initial design and routine maintenance has been inadequate. 7.09 After the 1979 floods, HIMAT engaged consultants to review the whole drainage system. The Bank agreed to extend the loan closing date until September 1981 to permit the study to be completed. Using a runoff coefficient of 60%, the consultants recommended an additional main drain and a second drainage pumping station at a total cost of US$(1981)6.4 million. In effect, the consultants reverted to the recommendations of the feasibility study. On the basis of information on rainfall and volume of water pumped from 1979 to 1981 (Table 13), it is estimated that the monthly maximum corrected runoff removed was about 13% of rainfall volume. This apparently low percentage was sufficient, under all but exceptional circumstances, to prevent prolonged flooding because of the massive storage capacity of the scattered low depressions which resulted from the lack of land improvement works. Under the original plan for field crops, land leveling was supposed to eliminate such spots. With the present land use for grazing, however, they are beneficial in providing additional moisture to support pastures for the dry season. It is expected that the project area will continue to be used predominantly for livestock, which benefit from dry season grazing in the seasonally flooded depressions. It is concluded, therefore, that the runoff coefficient used by the consultants was unrealistically high, since it is most unlikely that it will ever be agronomically feasible to devote the majority of the project area to crops as envisaged at appraisal.  - 18 - 7.10 During the preparation of this report, the Bank provided suggestions to HIMAT for low cost solutions for the various drainage problems. These consist mainly of increasing the capacities of the main drains M and I to match the maximum possible capacity of the Boquitas pumping station, in which the pumps will have to be renewed. A small sub-pumping station should be installed near Manati to drain the 800 ha of best soils located around drain IV. Furthermore, INCORA's cost of administration should be reduced by limiting its staff to only those who are required to improving agricultural production in the area. Since livestock farming is the predominant activity and since its present level of productivity is extremely low, it is clear that INCORA should focus its extension effort in this field (para 5.13).  Annex 1 Translation of incoming telex - 19 - Page 1 Bogota, May 1.6, 1983 455 Mr. Shiv S. Kapur Director Operations Evaluation Department The World Bank Washington, D.C. Subject: Completion Reports on Palmira Water Supply and Sewerage Project (IBRD Loan 738-CO) and Second Atlantico Development Project (IBRD Loan 849-co) Dear Mr. Kapur: I am pleased to inform you that the referenced documents can generally be said to present the essential conditions that prevailed during execution of the projects in question; as a result, I regard these documents as satisfactory and not in need of any amendment. Despite the foregoing, and with a view to improving the execution of future projects, I feel that it is important to bear in mind the conclusions of the reports, namely that conditions, goals, period of execution, adminis- tration and evaluation be made more realistic by employing more pragmatic assumptions. As a consequence, the National Government is prepared to extend to the Bank whatever assistance it requires to ensure better execution of future projects and so secure increased social benefits for the regions in question. Finally, I feel that the greatest importance should be attached to comments in this regard by the project executing agencies, as they are the bodies best placed to assess the results, both technical and financial, of the projects and the lessons to be drawm from them. Regards, Jorge Serpa Erazo Director General of Public Credit Ministry of Finance  - 20 - Annex 1 Page 2 INCORA Ministry of Agriculture Ref.: 06927 Bogota, June 3, 1983 Mr. Shiv S. Kapur Director, Operations Evaluation Department The World Bank Washington, D.C. Subject: Project Completion Report for Loan 849-CO Dear Mr. Kapur: The Project Completion Report for Loan 849-CO prepared by the Bank's Operations Evaluation Department is in general satisfactory, although I would note the following: 1. BACKGROUND From the standpoint of development, Loan 849-CO was neither sufficient nor adequate for local conditions and the beneficiaries. The loan should have included: - construction of schools and health posts - farming and/or livestock credit, i.e. in the form of an integrated development project. 2. FEASIBILITY STUDY The consultants were overly optimistic as regards project feasibility and the results were expected within a very short time frame. 3. ORGANIZATION AND MANAGEMENT Para. 4.03 states that INCORA employed 288 persons for the stage 1 and 2 project areas, and for the Repelon scheme. That figure should be correct_d to 110, which is the correct number. Para. 7.10 states that "INCORA's cost of administration should be reduced..."  - 21 - Annex 1 Page 3 The idea is not to reduce, but rather to transfer employees from other areas to stage 2 in order to increase the number of livestock technicians, as stockraising is the predominant activity. In addition, it should be noted that the number of employees in Atlantico remained at an average of 100 throughout the last five years. 4. CREDIT OPERATIONS In late 1982 INCORA's Boar,d of Directors approved a Col$26 million portfolio operation, benefiting 423 borrowers, who remained eligible for additional credits, which they are in fact receiving. 5. MAINTENANCE AND CONSERVATION The maintenance and conservation of the drainage works changed very significantly during the last four months of 1982 and the early months of 1983. HIMAT has worked resolutely to make the necessary resources available to normalize drainage service in the area and thereby avoid flooding. The inhabitants of the region are very satisfied with HIMAT's performance and have said so in writing. As the 1979 floods were caused by rainfall that exceeded all forecasts, the government authorities working in the area cannot be blamed. 6. BENEFITS OF THE LOAN Despite all the difficulties that arose during the implementation of Loan 849-CO we must stress the transformation that has taken place in the target area. It is true that the projected parameters or indicators have not been achieved, but direct and indirect benefits have been felt in all respects. Conditions are now favorable for the credit operations and it is safe to say that there is no threat of flooding in the area. Yours truly, /s/ Natividad Daza Casilimas Deputy Manager for Planning   - 23 - COLOMBIA Table 1 SECOND ATLANTICO DEVELOPMENT PROJECT (Loan 849-CO) PROJECT COMPLETION REPORT Cost Table Appraisal Estimate Actual (1972) Col$ Million (1972) Col$ Million Local Foreign Total Local Foreign Total Civil works 29.8 19.7 49.5 9.4 6.3 15.7 Land acquisition 27.0 - 27.0 1.0 - 1.0 Equipment 0.9 27.1 28.0 - 0.5 0.5 Consultants, training 3.6 13.4 17.0 0.3 - 0.3 Administration, engineering 9.6 - 9.6 23.4 - 23.4 0 & M during construction - - - a/ 15.5 6.6 b/ 22.1 Subtotal 70.9 60.2 131.1 49.6 13.4 63.0 Price contingencies 3.4 2.9 6.3 n.a. c/ n.a. n.a. Physical contingencies 10.6 9.0 19.6 n.a. n.a. n.a. Dairy farm investments 46.6 - 46.6 12.5 - 12.5 Total Project Cost 131.5 72.1 203.6 d/ 62.1 13.4 75.5 Actual I of Appr. (1981 USS Million) e/ (1981 US$ Million)e/ Estim. Local Foreign Total Local Foreign Total Civil works 4.03 2.66 6.69 1.14 0.76 1.90 28 Land acquisition 3.69 - 3.69 0.12 - 0.12 3 Equipment 0.11 3.69 3.80 - 0.06 0.06 2 Consultants, training 0.49 1.83 2.32 0.43 - 0.43 19 Administration, engineering 1.32 - 1.32 2.79 - 2.79 211 0 & M during construction - - - 1.91 0.82 b/ 2.73 n.a Subtotal 9.64 8.18 17.82 6.39 1.64 8.03 n.a Price contingencies 0.46 0.40 0.86 n.a. n.a. n.a. n.a Physical contingencies 1.46 1.23 2.69 n.a. n.a. n.a. n.a Dairy farm investments 6.35 - 6.35 1.57 - 1.57 25 Total Project Cost 17.91 9.81 27.72 7.96 1.64 9.60 35 Project Financing Bank (excluding interest during 11.73 1.31 11 construction) f/ Borrover 9.64 6.72 70 Beneficiary 6.35 1.57 25 TOTAL PROJECT 27.72 9.60 35 ia 0 & M costs during construction were not foreseen during appraisal b/ Estimated. c/ n.a. - not applicable. d/ Col$203.6 million corresponds to US$9.69 million in 1972 dollars. e/ 1972 dollars multiplied by 2.86 to express in 1981 dollars. fi Interest during construction estimated US$0.9 million; actual US$0.302 million. October 20, 1982  - 26 - COLOMBIA Table 4 SECOND ATLANTICO DEVELOPMENT PROJECT (Loan 849-CO) PROJECT COMPLETION REPORT Land Distribution and Tenure in 1981 Ha INCORA Land Former swamps 5,026 Purchase of land 8,592 Gross area INCORA 13,618 Area in Suan township 702 Gross area INCORA Loan 849-CO 12,916 Right of way 369 Net area INCORA Loan 849-CO 12,547 Private Lands Townships 138 Class VIII and erosion lands 5,126 Net private land 9,000 Gross private land 14,264 Gross area INCORA Loan 849-CO 12,916 Gross area private land 14,264 Gross area Loan 849-CO 27,180 Gross irrigation area Loan 502-CO 4,400 Gross area Loan 502-CO and 849-CO 31,580 Type of Farmers N