Argentina - Balcarce Livestock Development Project

Document of The World Bank FOR OFFICIAL USE ONLY FLE C oP Y Report No. 3729 PROJECT PERFORMANCE AUDIT REPORT ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) December 23, 1981 Operations Evaluation Department 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. ABBREVIATIONS BNA Banco de la Nacion Argentina - State Bank of Argentina CB Banco Central - Central Bank CONADE Consejo Nacional de Desarollo - National Development Council CREA Consorcio Regional de Experimentacion Agropecuaria - Regional Agricultural Research Group ERR Economic Rate of Return FRR Financial Rate of Return INTA Instituto Nacional de Technologia Agricola - National Institute of Agricultural Technology JNC Junta Nacional de Carnes - National Meat Board FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) TABLE OF CONTENTS Page No. Preface .. . ..................................................... i Basic Data Sheet ......... ............................. ........... ii Highlights .................. *............ 0......................... iv PROJECT PERFORMANCE AUDIT MEMORANDUM I. SUMMARY ................................. 1 II. MAIN ISSUES ...6.................................. 6 A. Adoption of the Technological Package ........ 6 Profitability and Spread . 7 Livestock Prices: Period Average and Fluctuation ...................... . 9 Physical 'Impact of Project Investments .... 11 Amount of Investments Financed .................... 13 Rates of Return Achieved ......................... 14 Ranch Management and Adoption ..................... 15 B. The Interest Rate Issue ............................. 19 C. Disbursement Procedures ............................ 21 Tables: 1 Argentina Beef Prices, Domestic and Export .. 27 2 Production Coefficients ............................ 28 3 Average Physical Investments per Ranch .............. 29 4 Adoption of New Livestock Techniques ............ 30 Annex: Comments by Banco de la Nacion Argentina ................ 31 PROJECT COMPLETION REPORT I. BACKGROUND .............................. 35 Land Tenure ................ ............... 35 Land Extensive Agriculture ........................... 35 Beef Raising Features ......... .... . ... ....... 35 Project Area . ...................... ..... 36 The Loan ........ ............................ 36 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. TABLE OF CONTENTS (continued) Page No. II. FORMULATION .......................*.......**.......... 37 Origins of Project ......6 ....... .. ..... ... . ...... 37 The Technological Package .............. -... .......... 37 Financing .. .o. . ... .*......o.. ...........o........... 38 Benefits and Justification .............. .......... 39 III. IMPLEMENTATION .... .. ..... o_....ooo. ..... ....... 39 Project Cost -................... ..... .............. 40 Number of Ranches Assisted .......... ... 41 Project Financing ........... .... . ............ ....... 41 Indexation of Subloans .... . ......... .......... ..... 42 Disbursements ............. ......... ..... ............ 44 On-ranch Investment .. ...... .. ............ .. ..... . ...... 46 Delays in Project Execution . .......................... 47 IV. AGRICULTURAL IMPACT ........ ......... . ... .. ...... ... ..... 51 V. RATES OF RETURN ........... ... ............. 53 VI. INSTITUTIONAL PERFORMANCE AND DEVELOPMENT .............. 54 Project Unit ..............- o- . ......... 54 BNA ...................... 56 Drainage and Soil Studies ....... . ........... 59 VII. SPECIAL ISSUES ............... ....o.. ...o ... .... 60 Government Policies ................... ..... . 60 Replication ... ........... ....... .. ....... . 61 VIII. BANK PERFORMANCE ................oo.o..... .. .... ..... . 61 Map IBRD 1715 PROJECT PERFORMANCE AUDIT REPORT ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) PREFACE This is a performance audit of the Balcarce Livestock Development Project in Argentina for which Loan 505-AR was approved in July 1967 in the sum of US$15.3 million. The loan was closed as of July 31, 1980, and the unwithdrawn balance (US$300,000) was cancelled on December 10, 1980. The audit report consists of an audit memorandum prepared by the Operations Evaluation Department and a Project Completion Report (PCR) dated March 31, 1981. The P1CR was prepared by the Latin America and the Caribbean Regional Office on the basis of a country visit in November 1980. The audit memorandum is based on a review of the Appraisal Report (No. TO-540a) dated June 28, 1967, the President's Report (P-553) of June 29, 1967, the Loan and Project Agreements dated July 31, 1967 and the PCR; correspondence with the Borrower and internal Bank memoranda on project issues as contained in rele- vant Bank files have also been consulted and Bank staff associated with the project have been interviewed. Several academic studies relevant to the project were also consulted. The draft report was sent to the Borrower on August 18, 1981 and a telex has been received from Banco de la Nacion Argentina (attached as an annex to the audit memorandum) indicating they had no comments. On the basis of this abbreviated review process, the audit finds that the PCR covers adequately the project's principal achievements and shortcomings, and the PPAM generally agrees with the conclusions. In addition to summarizing the objectives and results of the project, the PPAM expands on several points because of their importance to this as well as other livestock projects.  - ii - PROJECT PERFORMANCE AUDIT BASIC DATA SHEET ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) KEY PROJECT DATA Appraisal Actual or Item Expectation Current Estimate Total Project Cost (Constant 1966 US$ million) 39.1 31.2 Underrun (%) - 22.8 Total Project Cost (Current US$ million) 39.1 45.7 Overrun (%) - 17.0 Loan Amount (US$ million) - 15.3 Disbursed - 15.0 Cancelled December 10, 1980 - 0.3 Repaid )May 31, 1981 - 7.9/a Outstanding )- 7.8. Date Physical Components Completed 07/72 05/80 Proportion Completed by Appraisal Target Date (%) 100 50.0 Proportion of Time Overrun (%) - 240.0 Economic Rate of Return (%) 27.0 14.0 Financial Rate of Return (%) 25.0-33.0L 10.0 Cumulative Estimated and Actual Disbursements (US$ million) FY68 FY69 FY70 FY71 FY72 FY73 FY74 Estimated 0.5 2.3 6.7 13.4 15.3L 15.3 15.3 Actual /e /e 0.1 1.7 2.9 4.6 5.7 Actual/Estimated (%) 4 3 1.4 13 19 30 37 FY75 FY76 FY77 FY78 FY79 FY80 Estimated 15.3 15.3 15.3 15.3 15.3 15.3 Actual 10.8 11.4 11.8 13.1 14.2 15.0 Actual/Estimated (%) 71 75 77 86 92 98 OTHER PROJECT DATA Original Actual or Item Plan Revisions Estimate Actual First Mention in Files or Timetable - 01/03/65 Government's AppLication - - 11/31/65 Negotiations - - 02/12/67 Board Approval 07/18/67 - 07/18/67 Loan Agreement Date 07/31/67 - 07/31/67 Effectiveness Date 10/30/67 - 10/30/67 Closing Date 07/31/72 - 07/31/80 Borrower Government of Argentina Executing Agency Central Bank Fiscal Year of Borrower January 1 - December 31 Follow-on Project Nam./f Agricultural Credit Project Loan Number 1564-AR Amount (US$ million) 60 Loan Agreement Date 06/29/78 /a Sold (and repaid) to third party US$1.1 million. Tb Borrower's obligation at that date US$7.8 million less exchange adjust- ments US$0.5 million equals US$7.3 million. /c 25% for partial ranch development and 33% for complete ranch development. 7 Original closing date, July 31, 1972. e US$20,000 disbursed by end FY68 and US$72,200 disbursed by FY69. 7T This loan ss cancelled (for the full amount) in April 1980. - iii - MISSION DATA Sent Month/ No. of No. of Man- Date of Item _y Year Weeks Persons weeks Report Identification /a Preparation FAO/CP 04/65 4 5 20 08/31/65 Appraisal IBRD 01/66 4 6 24 06/28/67 Reappraisal/b IBRD 06/70 4 5 20 09/15/70 Total 12 64 Supervision I IBRD 03/68 1 2 2 05/01/68 Supervision II IBRD 02/69 1 2 2 03/26/69 Supervision III IBRD 12/70 1 1 1 01/20/71 Supervision IV IBRD 06/71 1 3 3 07/14/71 Supervision V IBRD 02/72 1 1 1 04/18/72 Supervision VI IBRD 11/72 2 1 2 12/26/72 Supervision VII IBRD 06/73 1 1 1 06/19/73 Supervision VIII IBRD 11/73 1 1 1 11/29/73 Supervision IX IBRD 06/74 1 1 1 08/02/75 Supervision X IBRD 12/74 2 1 2 01/02/74 Supervision XI IBRD 10/75 2 1 2 11/21/75 Supervision XII IBRD 05/76 2 1 2 06/07/76 Supervision XIII IBRD 12/76 1 2 2 01/14/77 Supervision XIV IBRD 04/77 1 2 2 04/15/77 Supervision XV IBRD 11/77 1 3 3 11/30/77 Supervision XVI IBRD 07/78 2 1 2 08/29/78 Supervision XVII IBRD 12/78 1 1 1 01/05/79 Supervision XVIII IBRD 05/79 2 2 4 05/23/79 Supervision XIX IBRD 10/79 0.5 2 1.0 10/26/79 Supervision XX IBRD 03/80 0.5 1 0.5 04/04/80 25.0 35.5 Completion IBRD 11/80 2 1 2 02/01/81 Total 27.0 37.5 COUNTRY EXCHANGE RATES Name of Currency (Abbreviation) Peso ($a) Year: Exchange Rate: Appraisal Year Average US$1 = 1.90 Intervening Years Average US$1 = 233.26 Completion Year Average US$1 = 1,950.00 /a Based on a successful FAO/UNDP applied research project in pasture and animal production carried out in the project area. /b Basically in-depth review of project implementation, scheduled when project activities had been advancing slowly; however, in early 1970, lending accelerated and a reappraisal was no longer required (PPAM para. 52). - iv - PROJECT PERFORMANCE AUDIT REPORT ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) HIGHLIGHTS The project attempted to introduce advanced pasture production techniques to a limited number of ranchers in the Balcarce area of Buenos Aires Province. However, a strong demonstration effect was envisaged, which would increase livestock production throughout the area. Project funds were used for on-lending for pasture improvement, fencing, stock handling and watering facilities and purchase of breeding cattle as well as machinery for contractors who would provide services to ranches too small to justify pur- chase of machinery by themselves. The project also provided for the creation of a small group of specially trained technicians in the National Institute of Agricultural Technology (INTA) and the project also financed drainage and soil studies in the project area. Total project costs were estimated at US$39.1 million. The project had mixed success. Progress of the project was slower than anticipated at appraisal. The project financed investments on 1,147 ranches over 13 years, compared with the 700 ranches which were expected to initiate investments in the first 3 years, and although on-ranch invest- ments appear to have exceeded appraisal estimates in the major investment categories (pastures, fertilizer, cattle and fencing), the average on-farm investment was much lower than expected. A dispute between the Bank and Government over subloan lending terms at the project's initiation delayed implementation. Furthermore, project subloan approvals and project disburse- ments were strongly correlated with beef prices which oscillated considerably during the project period. The impact of the project on beef production and productivity was positive although much lower (40%) than predicted at appraisal. This, in addition to lower and more fluctuating beef prices and less intensive manage- ment on many ranches, made the profitability of the partially adopted tech- nological package much lower in the specific Argentine context than expected. The financial rate of return is now estimated at about 10%, compared with a rate of 25%-33% at appraisal, and the economic rate of return is estimated at 14%, compared with 27% at appraisal. The project had a very positive impact of INTA, an important side benefit (PPAM para. 9). Other points of interest are: - low profitability in the Argentine context of the technical package, resulting from an interaction between technological, economic and managerial factors, prevented the spread of the package to other farms (PPAM, paras. 2, 4 and 12-18); - v - - the slow and only partial adoption of the recommended technical package is similar to the situation experienced in other Bank- financed livestock development projects in Latin America (PPAM, para. 14); - livestock prices were an important contributor to the low rate of return achieved, but not as significant as the failure of project investments to have the physical impact expected. This failure in turn was caused by an interaction of numerous factors including Government price policy, beef/fertilizer price ratios, incomplete adoption of the technical package, etc. (PPAM, paras. 19 and 22-29); - there appears to be a bimodal distribution of ranch profitability and the rancher should either adopt all the technologies and spend most of his time on the farm or adopt no new technology and spend little time on the farm (PPAM, para. 38); - a significant number of small ranchers (by standards of this region) participated in the project, most of them, and more so than larger ranchers, stayed on their farm and therefore had an advantage in managing the new technology better; also, smaller ranchers performed better than larger ones (PPAM, paras. 39-42); - there was a continuous internal dispute about indexing of project subloans, with project staff normally recommending to fit the project into the existing (distorted) economic structure, while country staff generally suggested to use the project to improve this structure (PPAM, paras. 57 and 48-55); - lack of intergovernmental coordination proved to be one of the causes for implementation delays as well as for serious problems encountered under subsequent Bank lending operations to the agricul- tural sector (PCR, paras. 2.02 and 6.02); - the project unit found Bank project supervision inadequate: insuffi- cient guidance in setting up monitoring and evaluation systems, repetitious and time-consuming missions, too many changes in Bank staff who had to interview the same people on the same subjects repetitiously and their lack of Spanish fluency as a major drawback in communications (PCR, para. 8.04); and - the Bank reimbursed a lower percentage of ranch loans than agreed at appraisal due to long timelags and rapid depreciation of the Argentine peso. This issue is characteristic for agricultural credit projects where the Bank refinances (in foreign exchange) a fixed percentage of subloan disbursements (in local currency) using the exchange rate at the time of its disbursements and the audit suggests that operational guidelines be developed to cover this (PPAM, paras. 59-60). PROJECT PERFORMANCE AUDIT MEMORANDUM ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) I. SUMMARY 1. The Balcarce Livestock Development Project was financed by Loan 505-AR (US$15.3 million). It was prepared in August 1965, appraised in January/February 1966, signed in July 1967, and became effective on Octo- ber 30, 1967. The closing date was originally July 31, 1972, but was extended three times (PCR para. 3.02) ultimately to July 31, 1980, and the remaining loan balance of US$0.3 million was cancelled on December 10, 1980, more than five months after the closing date.l/ 2. The project was a pilot effort to induce ranchers in the Balcarce area, in the Province of Buenos Aires, to adopt improved technologies in pasture production and management and in animal health practices which had been introduced in a previous UNDP/FAO applied research program. The demon- stration effect of introducing these techniques to 700 participating breeding ranches was expected to gradually lead to a rise in livestock production, weaner calves in particular (PCR para. 2.04), on other ranches throughout the area./ The US$5 million per annum of credit expected to be granted from 1/ The closing date was June 30, 1980; a final disbursement was made on November 13, 1980 and the remaining balance was cancelled on December 10, 1980. For a full discussion of the issue regarding disbursements after the closing date, see Project Performance Audit Report, Ethiopia Addis Ababa Dairy Development Project (Credit 269-ET), OED Report under prepa- ration. 2/ The project objectives were described in the appraisal report (No. TO-540a, dated June 28, 1967, para. 34) as follows: "(a) to raise production and income of project farmers; (b) to augment the national breeding and fattening herd, primarily through increased production of heifers and feeder steers for sale to other farmers; (c) to demonstrate improved production techniques and investment opportunities to other farmers within and beyond the project area." Furthermore, additional benefits are described in the appraisal report (para. 67) as follows: "In addition to these direct production effects, the economy would gain from the demonstration of modern production techniques on project farms. Although this benefit is more intangible, it could make a substantial contribution to the long overdue breakthrough in Argentina-s static cattle industry." - 2 - the loan was insignificant compared with the total agricultural credit provided in Argentina or with the sector's economic importance (PCR para. 2.10). 3. Most project funds were to be used for pasture improvement, fencing, stock handling and watering facilities, and purchase of breeding cattle; credit to farm machinery contractors who would provide services to ranches too small to justify the purchase of machinery was also included. The project provided for the creation of a small specially trained group of technicians in the National Institute for Agricultural Technology (INTA). Under the guidance of a Project Director, this unit was to assist ranchers in preparing develop- ment plans and to supervise their execution. It was the first attempt in Argentina to provide such an agricultural program of technical assistance to ranchers (PPAM para. 9). The project also financed drainage and soil studies.! in different parts of the project area..2 Total project costs were estimated at US$39.1 million, including contributions by ranchers and contractors, the Banco de la Nacion Argentina (BNA), the only participating bank, and the Government. 4. The project had mixed success. At appraisal, it was predicted that ranch investments would be initiated on 700 ranches during the first three years, and completed after five years, exhausting project funds. Additional lending to continue the project was then expected. Instead, 1,174 ranches3/ were assisted by the project and actual on-ranch investments appear to have exceeded appraisal estimates in the four major categories - pastures, fertil- izer, cattle and fencing.4/ However, nearly 13 years were required for 1/ The studies provide a useful basis for land use and farm planning of the area and indicate that a major public investment program for main drainage canals, a concern at appraisal, is not warranted. For further details: PCR, paras. 6.11-6.13. 2/ Provision was made at appraisal for beef and meat marketing studies but because similar information was forthcoming from the National Meat Board (JNC), the Bank and the Government agreed that the project should not undertake these studies and the funds were shifted to other categories. Expenditures for telecommunications equipment and services for INTA, not originally planned, were included in the project after agreement betweeen the Government and the Bank. 3/ 173 ranchers received loans and initiated their investment plans, but subsequently cancelled their loans and withdrew; 1,174 ranchers completed their investment plans (PCR para. 3.04). 4/ The PCR in para. 3.12 also points to the fact that physical investment in improved pastures and fertilizer had reached 95% of appraisal esti- mates by August 1973; however, it should be remembered that the original closing date was set for July 1972 (PCR, para. 3.06 and 3.02). - 3 - disbursements to be completed. In addition, the package does not appear to have been widely adopted by other ranches making investments on their own1/ (the participating ranches constituted about 6% of the 20,000 ranches in the area). 5. The impact of the project on beef production and productivity was measured through two sample surveys carried out by the project unit (with assistance from Bank staff) one for 1972/73, and another for 1978/79. After nine years, the improved pastures on participating ranches had increased from 22% without the project to 46% with the project. The stocking rate, weaning rate, and beef production per hectare also all increased. Many participating ranchers also began to fatten steers of their own as well as to maintain a breeding herd, thus diversifying their output and spreading market risks. Nonetheless, the incremental physical output caused by ranch investments was still lower (40%) than that predicted at appraisal. For this reason, and because livestock prices averaged less than expected during the period 1975- 1980, both the economic and financial rates of return achieved were lower than expected at appraisal respectively, 14% vs. 27% and 10% vs. 25-33%..! It appears that the relatively low financial rate of return is a principal cause of the technological package's slow diffusion, brought about by unfavor- able Government policies and depressed international prices of meat. 6. Producer beef prices oscillated considerably during the project, declining just after appraisal (1967-1969), rising sharply to a historical peak (1970-1974), plummeting to a historical low (1975-1976) and then recover- ing (1977-1980). Project subloan approvals and project disbursements are strongly correlated with beef prices.3Y Livestock prices were determined 1/ This is the opinion of INTA and Bank staff. INTA's survey of the impact of project investments on participating ranches did not study operating procedures on non-participating ranches. Given that the spread of the technological package was a major aim of the project, a specific study of this spread would have been useful. 2/ The financial rate of return was estimated at 25% for partial ranch development (only improved pasture) and 33% for complete development (pasture improvement and complementary investments on the whole farm). Only 60 (or 9% of the project farmers) out of 700 were expected at appraisal to fall in the latter category. 3/ The PCR (para. 3.14) divides the project implementation period in five phases; these phases and related BNA disbursements are as follows: (i) 1968-1969 (slow start-up) US$372,000 or 1% (ii) 1970-1974 (the boom years) US$20.7 million or 71% (iii) 1975-1976. (the stagnating years) US$2.3 million or 8% (iv) 1977-1978 (signs of recovery) US$4.6 million or 16% (v) 1979-1980 (uncertain investment climate) US$1.1 million or 4% -4- principally by fluctuations in international beef markets, but Government policies, usually in an effort to control the cost of living to urban workers, also worked to depress livestock prices over much of the period (PPAM paras. 19-21).1/ 7. The ex-post evaluation shows, however, that most project farms continued to increase the area planted to improved pastures between 1972/73 and 1978/79, despite the severe beef price decline which took place between 1974 and 1978. This result suggests that most project ranchers found in- creased capitalization profitable even during this difficult period and despite what appears to be the moderate rate of return achieved. The negative interest rates on project subloans in these years might be a factor. Still, it is noteworthy that most project ranches utilized the technology increas- ingly intensively, while other ranches showed little tendency to adopt. Insufficient evidence is available to form a conclusive judgement on why some ranches have utilized the technology and others have not, but it appears that the availability of continuous competent on-ranch management may have been an important factor. Most larger ranchers in the project area are absent during most of the year and their traditional practice is to maintain herds on minimum investment and management outlays. There is some evidence that the less intensive techniques utilized on their ranches are nearly as profitable as the intensive techniques which were only partially adopted by the rancher. Also, the choice between the intensive and the extensive strategies was taken not solely on the basis of ranch profitability, but whether the rancher wanted and was financially able to live on the farm or in the city (PPAM paras. 34-42). 8. Although the number of annual loan approvals is strongly correlated with movements in the domestic producer price of beef, other problems associ- ated with the high and changing rate of inflation in Argentina and the inter- est rate to be charged to subborrowers, particularly the introduction of indexing, also affected the rate of implementation. The project lasted much longer than expected for several reasons: (i) Immediately after loan approval, the Government and the Bank dis- agreed on the interest rate to be charged subborrowers. Declining inflation in Argentina made the 16% originally negotiated appear too high to Government, which therefore delayed project implementation until late 1968. The originally established rate was never used and interest rates were revised four times between 1967 and 1969 (PCR para. 6.09). Only 1% of total BNA project disbursements occurred during this period, when incidentally beef prices were also quite low. I1 The region feels that Government policies were more important to deter- mine beef prices to the producer than international prices particularly when prices were on the upswing (see also PPAM paras. 20 and 45). - 5 - (ii) High inflation in Argentina after 1970 induced the Bank to again reopen the interest rate issue with the Government. The Bank requested indexing, as it had done at negotiations, and the Govern- ment at first refused. Loan demand was strong from 1970 through 1972, but it became clear that the project funds would not be totally committed by the original closing date. In addition, Government wanted to pursue a second project. The Bank indicated that it would not extend the closing date, nor discuss a second project, unless Government would accept indexing on project sub- loans. Government accepted indexing in 1973 (PCR para. 3.06). Indexing added an additional element of uncertainty to ranchers, a number of whom became worried when beef prices crashed after 1974. Approximately 10-15% of the subborrowers cancelled their loans during this periodl/ and few other borrowers./ came forward (PCR paras. 3.08-3.09 and 3.24). (iii) Rapid inflation in Argentina, combined with slow bureaucratic processes in the BNA and the Central Bank (CB), resulted in a request for reimbursement to the Bank only several months after the original peso funds for on-farm investment had been disbursed. Because the Government generally followed a policy of continuing exchange rate adjustment to internal inflation, Bank reimbursement usually was made at a substantially lower exchange rate. Thus, the Government bore a larger proportion of the total project costs than was originally agreed. Accordingly, the funds in Loan 505-AR assisted a larger number of loans, over a longer period than would otherwise have occurred (PCR paras. 3.10 and 3.28). 9. The project had a very positive impact on INTA, an important side effect. Prior to the project, INTA generally worked on a commodity basis, both in research and extension. As a result of the project, research efforts began to focus on the entire ranch as the production unit. The same approach was adopted in extension, and both technical and financial aspects were con- sidered for on-ranch development. This approach was innovative for INTA and subsequently came into use on a national level. Through the work of the 1/ It is difficult to disentangle the effects of indexing from those of beef price variation, but it appears that prices were significantly more important during the period that interest rates were negative. The indexing system adopted was imperfect so that subloan interest rates remained negative in real terms until July 1978, when a new system based on the General Wholesale Price Index was introduced (PCR para. 6.10). However, lending then virtually stopped, despite improving livestock prices after this date (PCR para. 3.10). 2/ Also, investment funds were frequently available from the BNA and other commercial banks at more attractive lending rates, causing ranchers to prefer these sources. - 6 - project unit, INTA also came into closer contact with ranchers, became con- cerned with more practical issues, and augmented and improved its extension staff (PCR para. 8.02). 10. Prior to closing this loan, a second loan (Loan 1521-AR)I/, was signed on June 29, 1978, to finance: (i) On-farm development investments for livestock and crop production in selected areas of the pampas (including the Balcarce project area) and Mesopotamia. (ii) Weed control in the extensive areas of the pampas. (iii) Land clearing in the northern provinces of the country. In part this project was to follow up the Balcarce project in extending credit and technical assistance to ranchers for the purpose of increasing livestock productivity in the country. The total project cost was estimated at US$162 million and the Bank loan was US$60 million. However, Government decided to cancel the loan in 1980. A separate Project Performance Audit for this project has been prepared.2/ II. MAIN ISSUES A. Adoption of the Technological Package 11. The technical package as implemented under the project resulted in lower profitability than expected at appraisal (PPAM, para. 5) and was not widely adopted by other ranchers contrary to a strong demonstration effect expected at appraisal (PPAM, paras. 2 and 4). The audit analyses below a series of interacting factors which led to this situation. It should be noted, however, that the audit does not and, in fact, cannot discuss the suitability of the technological package per se;.1/ it discusses the applica- tion of the proposed technology within the specific Argentine context and 1/ For further details, see also staff appraisal report, Argentina Agri- cultural Credit Project, Report No. 1918a-AR, dated April 25, 1978. 2/ Project Performance Audit Report: Argentina Agricultural Credit Proj- * ect, OED Report No. 3699, dated November 30, 1981. 3/ The technology has been intensively used under similar soil and climatic conditions in Australia and New Zealand; almost all livestock experts believe that the technology offers the potential for increasing livestock output in Argentina (provided there is a reasonable degree of economic stability and a satisfactory set of pricing and other policies). - 7 - circumstances prevailing during the implementation period. The PPAR points out why, given these adverse circumstances, the technology has not found greater use in Argentina to date, emphasizing the importance of prices, economic and management factors, an integral part of the prevailing production environment and their interaction with the use of the technology.1/ Under various headings, this will be discussed in more detail below. Profitability and Spread 12. The technological package financed by the project was introduced in Argentina on a trial basis in 1963. The initial results received substantial publicity in the Balcarce area and additional publicity took place after the project became effective in 1977. The project unit promoted the technology through direct contacts with ranchers and the Bank sent a well known New Zealand livestock expert to discuss the technology with Argentine ranchers in 1968. At various times during the project's life, the relative merits of the technological package were widely debated in the media. The technological package is therefore known by nearly all producers. After having been avail- able in Argentina for mtore than 15 years, and having been actively promoted for more than 11 years, however, it has not been widely adopted. Roughly 6% of the 20,000 ranchers in the Balcarce project area participated in the project; it appears that few other ranches have adopted the technology on their own. This view is held by Bank and INTA staff, based on impressionistic evidence. Some empirical evidence in support has also been developed (PPAM paras. 13 and 15). 13. In 1979, the Bank hired a consultant to assist INTA-s project unit with the design of the ex-post project evaluation.2/ This consultant also analyzed the technology's spread. Because the technological package centers on pasture improvement based on phosphatic fertilization of legume-grass mixtures, and since overall fertilizer consumption at the outset of the project was negligible in the Pampa region, he took fertilizer use as a proxy for the diffusion of the technology package. A comparison of phosphate fer- tilizer use for pastureS in the entire Pampa region with fertilizer financed by the project suggested that few ranches outside the project had adopted this 1/ The PPAR does not review how well the technology might do under different hypothetical scenarios; there is no evidence to support such futuristic judgements, which in addition are irrelevant to derive lessons from past experience. Neither is the PPAR a diatribe against improved pastures; it is an effort, however, to point to the importance of and the need to take more complete account of the local environment when trying to transfer a technology from one region to another. 2/ Supervision of Balcarce Livestock Project (Loan 505-AR) and Special Report on Status of Monitoring and Evaluation dated September 7, 1979 (further referred to as Special Supervision and Evaluation Report). - 8 - technology. Even if all the phosphate fertilizer utilized in the Pampa region were used on farms within the project area, not more than 20% of these farms (including the 6% within the project) could be applying fertilizer on a sig- nificant basis.!V The consultant concluded that "The spread effects of the project will be very difficult to measure but it appears at this time that project effects may have been small and more indirect than direct." 14. The slow and limited adoption of the recommended technological package in the Balcarce project is similar to the situation experienced in several other Bank-financed livestock development projects in Latin America, such as those in Uruguay, Brazil and Paraguay.2. In those projects, slow adoption was caused principally by the low profitability of the partially applied technology to producers. The low profitability resulted from an interaction between technological, economic, and managerial problems which were an integral part of the production environment in these countries, but which have only become clearer as the projects were implemented. 15. The consultant (previously mentioned in PPAM para. 13) undertook further analysis of the factors affecting diffusion of the technological package by relating the amount of fertilizer used in the Pampa region to the relative price of beef and fertilizer. Applying a linear function, he found that a trend term and the beef/fertilizer price-ratio accounted for 74% of the variance of total fertilizer use over the period 1967-79.1/ Thus, price 1/ In fact, the consultant pointed out that about half of the phosphatic fertilizer used outside the project does not correspond to the type of fertilizer promoted by the project. Moreover, the project area accounts for about half the area of Buenos Aires province, and a much smaller proportion of the total Pampa region. And it is known that project farms have purchased additional fertilizer with non-project financing, thus accounting themselves for part of the non-project fertilizer consumption. 2/ See PPAR - Uruguay Third and Fourth Livestock Development Projects, OED Report No. 1321, dated October 20, 1976; PPAR - Uruguay Fourth Livestock Development Project (Second Stage), OED Report No. 2572 dated June 29, 1979; PPAR - Brazil First and Interim Second Livestock Development Projects, OED Report No. 2402 dated April 5, 1979; and PPAR - Paraguay Third Livestock Credit Project, OED Report No. 1991 dated March 28, 1978. Similar but not necessarily completely identical experiences were also apparent in Ecuador and Honduras livestock projects. For further in- sights, see PPAR - Ecuador First and Second Livestock Development Proj- ects, OED Report No. 892 dated October 21, 1975; PPAR - Ecuador Third Livestock Development Project, OED Report No. 2799 dated December 28, 1979 and PPAR - Honduras First Livestock Development Project, OED Report No. 1920 dated February 21, 1978. 3/ The equation estimated is: y = -11.1 + 1.37t + 6.54 cattle/fertilizer price. No significance statistics were reported. -9- appears to be a major determinant. Further analysis suggested that farmers began to apply fertilizer at a price ratio threshold which was not reached in 1975-78, and was only barely surpassed in four more years (1967-69, 1974)..! The consultant concluded "that during much of the period economic conditions were not conducive to foster investments in pasture technology based on phosphate fertilizer." 16. This conclusion is consistent with both the variation in project disbursements, and also with other analyses by Bank staff and by academics. For example, in 1969, when the project was experiencing great difficulty with lending, a different livestock expert hired by the Bank to review the project reported that "the (current) financial rate of return feasible under the project does not justify investment...project management finds that under the new price input/output relationships, the estimate of the appraisal mission of a financial rate of return of plus 25% is currently only minus 15%.".2/ Shortly thereafter, livestock prices began to improve and project disburse- ments picked up noticeably, but then declined again after 1974 when indexing of subloans was initiated and livestock prices fell. 17. In another study,.3/ Gimenez Dixon estimated that the FRR should range between 5-14%, using 1968 prices. His estimate, although higher than that cited above, was significantly lower than the return estimated at appraisal. Gimenez Dixon also estimated that ranchers might have obtained a return of about 10% on their capital with relatively risk-free financial investments, so that the return on project investments (in the best situation) would only have barely compensated for additional risk. 18. Finally, the ex-post evaluation carried out by INTA's project unit,A/ as adjusted by the PCR, suggests that the FRR for project invest- ments was about 10% (the project's FRR is further discussed in PPAM, paras. 31 and 32). Livestock Prices: Period Average and Fluctuations 19. Although lower livestock prices, per se, were an important contribu- tor to the low rate of return achieved, they may not be as significant a 1/ For cattle/fertilizer price ratios, see PCR paras. 3.19 and 3.22. 2/ Bank memorandum dated June 20, 1969. 3/ Gimenez Dixon, Jorge Joaquin, "An Economic Analysis of Range Improvements in the Cattle Breeding Area of Buenos Aires Province," Ph.D dissertation, Michigan State University (East Lansing; 1969). 4/ Informacion Previa para la Evaluacion Final del Proyecto Balcarce, Referida al Encuestamiento, Resultados Fisicos Y Rentabilidad de las Inversiones, Balcarce, Junio 1980 (further referred to as INTA ex-post evaluation). - 10 - factor as the failure by project investments to have the physical impact expected.L/ Livestock prices and project physical impact will be examined separately. 20. During the project, substantial variations occurred in the beef price, both in international and in the domestic market. These movements are shown in Table 1. There is a high correlation between the international price and the domestic producer price, but Government policy acted to depress the domestic price almost continuously (PPAM, para. 45). 21. The appraisal report estimated the FRR using January 1966 producer prices (which stood at an index level of 112). This level was only 3% above the average producer price during the pre-appraisal period, 1960-65 (which stood at an index level of 109). Further, the appraisal report price was only 8% higher than the average price during the period of project implementation, 1968-80 (which stood at an index level of 104). The problem from the price side was not so much the average, as the fluctuation. Furthermore, project investments and project outputs were poorly coordinated with the beef price cycle.!/ The producer price of beef was highest during the period when the bulk of project investments was made, 1970-74, and then prices abruptly collapsed just as these investments were resulting in higher output on project ranches, 1975-78. Project ranches thus paid more for their inputs (breeding animals) and received less for their product (weaner calves and steers) than the average price during the period suggests. The relevant price statistics are: 1/ The region points to the fact that this is due to the incomplete adop- tion of the package which was influenced by non-farm factors such as international beef prices, Government policies and high inflation especially in fertilizer costs. CPS points to the fact that from 1974 through 1978, livestock prices were very low relative to fertilizer prices and very little fertilization of established pastures was carried out; this undoubtedly led to reduced productivity and thus reduced phy- sical output. These are valid remarks but the audit wishes to stress the importance of an interaction between various factors which determined the production environment (PPAM para. 14). 2/ Note that larger fixed financial commitments associated with the amorti- zation of ranch investments require heavier sales of cattle during periods of low prices, thus accentuating the cycle and working counter to the optimum ranch strategy. Thus, the existence of unpredictable price cycles probably reduces the attractiveness of the intensive technological package. - 11 - Beef Price Index, Period Averages 1970-74 1975-78 1968-78 Price Steers 125 78 102 Price Heifers 114 52 88 Source: Table 1 Physical Impact of Project Investments 22. The appraisal report assumed that 640 ranches would undergo partial development (improve pastures on 30% of their cattle area over three years) and that 60 ranches would undergo complete development (improve pastures on 50% of their cattle area over three years and other investments on the ranch). With complete development, the improvement in ranch technical parameters was expected to be both more rapid and ultimately larger than with partial devel- opment, as is shown in Table 2. 23. INTA's project unit carried out two sample surveys to determine the impact of project investments. In the first, ranches who had completed their investment program before June 1972 were surveyed with respect to production results in the fiscal year 1972/73. In the second, ranches whose investment program had been approved before June 1974 were surveyed with respect to production results in the fiscal year 1978/79. The second survey included ranches from the first survey, so that two post-investment observations are available for these. The average period between the initiation of ranch investments and the surveys are 3.5 years and 9.5 years for the ranches which were surveyed in both 1L972/73 and in 1978/79 (sample 1), and 7 years for the ranches surveyed only in 1978/79 (sample 2). 24. The average size!/ of participating ranches was close to projec- tions, if not slightly larger at project initiation and changed little during the project period. 1/ The average size does not reflect the considerable dispersion in the distribution of the ranches in the sample universe in 1978/79: % of Total Cattle Land Category Size % of Farms on all Project Farms I < 500 ha 39 12 II 500-1,500 ha 41 31 III > 1,500 ha 20 57 - 12 - 25. Participating ranches had 20-22 percent of their area already developed with improved pastures when they requested project financing. It had not been expected that ranches would have pastures prior to participation in the project. The participating ranches then improved, on average, an additional 25% of their ranch prior to 1978/79.11 Because the participating ranches had undertaken substantial pasture improvement prior to the project, the impact of project investments could not have been expected to follow either the "partial" or the "complete" development pattern, but a hybrid of the two. The "before development" level should have been close to that of the "partial" development pattern three years after investment (assuming that ranches had made their initial pasture investments sometime in the late 1960s), and the production level nine years after the project's investments should have been similar to that of the "complete" development pattern at the same stage. In fact, examination of the data on stocking and weaning rates, and on beef production per hectare, shows that the production impact has been less than projected even for "partial" development. 26. The stocking rate before development was projected at .80 adult animal units per hectare and was expected to increase to 1.20 and 1.80 animal units after nine years on partially and completely developed ranches, respec- tively. Instead, participating ranches in both samples (PPAM, para. 23) had stocking rates of .62 before development, and these rose to .86 and .94, respectively, seven and nine years after project investment was begun. 27. The appraisal report projected an increase in the weaning rate from .70 before development to .82 and .90 for the partially and completely devel- oped ranches. The actual increase shown on project ranches was from .74-.77 in the before development situation to .78 for both samples.2/ 28. Beef production before development was expected to be 65 kilos per hectare and was projected to rise to 172 kilos with partial development and to 298 kilos with complete development, both after nine years. Project ranches had higher beef production in the before project situation than was expected, 75-84 kilos per hectare, and increased their production much less, to 121 kilos per hectare after nine years. 1/ Precise data is not available, but the INTA ex-post evaluation, oR. cit., indicates the average project ranch improved about 240 ha with project financing, and another 100 ha with financing from other sources. This is in addition to the amount established prior to requesting project financing. 2/ INTA project agents felt that the pre-project weaning rates were over- estimated on project farms (PCR, para. 4.03), but Gimenez, oR. cit., found that weaning rates in the project area in 1968 averaged 70% on farms using traditional technology (ranging from 52% to 79%), and 80% on farms using modern technology (ranging from 68% to 87%). Given that most project subborrowers appear to have substantially used improved pasture before the project, the incremental impact of project investments on the weaning rate may have been small. - 13 - 29. The income stream from project ranches depends primarily on beef production per hectare, which is seen to have increased only 40 kilos per hectare instead of the roughly 100 kilos for the partially developed ranch. This shortfall sharply reduced the rate of return. Lower beef production per hectare resulted partly from the fact that both stocking rates and weaning rates have increased less in response to project investments than was ex- pected. Even more important, however, would appear to be the slower growth rate of animals. The appraisal report predicted that slaughter weights would increase by about 15% in response to ranch investments, and that project ranches, which had previously been engaged only in breeding, would produce about 20% of their income from 9-24 month slaughter steers (420 kilos live weight). The INTA ex-post evaluation reports no slaughter weights for the before development situation, but the data for animals sold in 1978/79 (Table 2) indicates, at least, that these have live weights significantly below those expected at appraisal, especially in the case of steers.1/ Amount of Investment Financed 30. The physical investments made per ranch and those projected at appraisal are shown in Table 3 2/. Somewhat fewer investments in each category were made than had been expected. These data are consistent with the financial records of the project as analyzed by Bank staff who found that the average on-farm investment financed by the project was about 60% of that planned at appraisal..!/ However, project ranches made additional invest- ments financed from non-project sources whose impact is also by necessity included in the ranch production data collected by INTA in 1978/79. The project unit found that these investments totalled about 21% of the project- financed investments;A/ it correctly included these investments when cal- culating the project rate of return. In sum, ranch investments were about 80% of those predicted at appraisal.2/ 1/ The generally lighter weight of the animals sold from project ranches suggests that the actual stocking rates in terms of animal units may not be fully comparable with those projected at appraisal. 2/ From INTA ex-post evaluation, op. cit., pages 35-37. 3/ Special Supervision and Evaluation Report, op. cit. Because the method of aggregating records used at the project level did not take inflation into account, a heroic effort was required to reconstruct project data. 4/ From INTA ex-post evaluation, op. cit., pages 63-64. 5/ Any on-farm investments made prior to the project are not included (PPAM, para. 31(ii)). - 14 - Rates of Return Achieved 31. Project investments were expected at appraisal to have a FRR of 25% (partial ranch development) and 33% (complete ranch development). The ex-post FRR estimated by the project unit was 14%. Because the 1980 price substan- tially exceeded the average price received by project ranches in the period 1974-1978 (PPAM paras. 19-21), the PCR adjusted the ex-post rate of return downwards to 10% (PCR para. 5.01). In addition to the adjustment for price, however, several other assumptions used in the INTA ex-post estimates seem optimistic with respect to the actual experience of project ranches: (i) The appraisal report assumed that production would increase over nine years before reaching equilibrium, with 50% of the final incremental output being achieved in year 6. In contrast, when prorating the production increase found in 1978/79 over previous years, the project unit assumed that production would reach its maximum level in five years, and that 50% of the project's impact would be achieved in year 3. The survey data collected by INTA provide some support for this assumption as beef production is nearly as high in year 3 as in year 9 on the farms sampled in 1972/73 and 1978/79, although this increase may reflect the impact of investments made prior to the project. (ii) The average participating ranch had already established a signifi- cant amount of improved pastures before obtaining project funds. It would appear that such investments were made after 1965, as the technology was not introduced before this date. Therefore, beef production may have been rising on participating ranches in the "before project" situation, rather than constant as assumed. If so, part of the incremental output achieved in the project period should correctly be attributed to prior investments. (iii) The appraisal report assumed that the pastures would have a produc- tive life of 12 years and estimated the rate of return over this period, while the project unit assumed that the pastures will have a productive life of 15 years, with constant productivity from years 5-15. The survey obtained no data on pasture longevity, but pas- tures are sometimes damaged by poor seeding, by lack of fertiliza- tion, by climatic problems, and by improper day-to-day management. The average pasture life in New Zealand is estimated at about 20 years, while that in Uruguay over the period 1966-1974 was markedly shorter, about eight years and shorter yet during the beef crisis, 1975-1978, when ranchers ceased to refertilize and overgrazed their pastures. Pasture duration in Argentina should lie closer to that of Uruguay, where management and the economic context are similar, than to that of New Zealand. The PCR also points out (PCR para. 7.05) that the apparent consumption of phosphate nutrients in Argentina declined from 21,000 m tons in 1972 to 4,000 m tons in 1975, suggesting that many Argentine producers stopped refertilizing during the beef crisis; severe rainfall and flooding in the project area in 1979 also damaged pastures. - 15 - 32. However, an upward adjustment to the FRR might offset the effect of the previously mentioned downward adjustments. Subborrowers enjoyed a large credit subsidy which res3ulted from the negative real rates of interest charged on subloans and at the same time, the producer was directly and indirectly providing a large subsidy to the consumer. The PCR estimates that the average real rate of interest for all subloans was about minus 20 percent per year. After indexing was introduced in 1973, the real rate of interest increased, but only to about minus 5 percent per year.1/ Given the absence of detailed information regarding these upward and downward adjustments, and given that the audit believes that they are likely offsetting, no further refinements have been made to the PCR's estimated FRR (10%). 33. The PCR estimated the economic rate of return (ERR) at 14%, higher than the FRR. This was done as an approximation again due to lack of precise information (PCR para. 5.04). The ERR is higher because export taxes were in force during much of the period 1969-80, the peso was frequently and signifi- cantly overvalued, domestic price controls on beef were sometimes in force, sales taxes on beef were frequently applied, and these factors depressed the financial producer price below that which would have prevailed in "free" markets. As recalculations would be complex and only likely to produce marginal changes, the audit accepts the PCR estimated ERR (14%). Ranch Management and Adoption 34. In the Uruguay Livestock Development Project, the quality of manage- ment was found to be an important determinant of the output achieved from application of a similar technological package.!/ A study by de Obschatko suggests that this may also be true in Argentina.i/ 1/ A new indexing system introduced in 1978 raised rates to 6%, but prac- tically no project loans were financed at this rate. 2/ "Considerable 'learning by doing' is required before improved pastures can be well managed; i.e., producers require substantial experience and understanding of the interaction between cattle and improved pastures before they are able to make the necessary day-to-day judgements which are required for good pasture management. And the greater the proportion of improved pastures on a ranch, the greater is the management sophisti- cation and dedication required to shift animals about so that all the improved pastures are adequately managed. The management approach, which improved pastures imply, is strikingly distinct from the extensive ranching traditionally practiced in Uruguay. It is not surprising that their complete adoption, for this reason is moving slowly." See Uruguay Third and Fourth Livestock Development Projects, an OED working paper for the project performance audit report, by Lovell S. Jarvis, University of California, Berkely, OED consultant, dated November 7, 1976 (para. 7.19). 3/ De Obschatko, Edith S., "Factores Limitantes a la Introduccion del Cambio Tecnologico en el Sector Agropecuario", M. S. dissertation, Escuela para Graduados en Ciencias Agropecuarias (Castelar: 1971). - 16 - 35. De Obschatko sought to determine the factors affecting the adoption of new livestock technologies. With the help of three staff members from INTA's Balcarce research station (where the project unit was based), she surveyed 71 farms in Ayacucho partido (county), a zone dedicated primarily to cattle breeding. This county lies at the center of the Balcarce project area an is highly similar to much of the rest of the project area. Her research was carried out in 1969-70, before the project had a significant impact, but as noted before (PPAM para. 25), a number of ranchers already had adopted some of the technologies promoted by the project.1!/ 36. De Obschatko chose 8 techniques applicable to cattle breeding, all of which had been studied at INTA and which were to be recommended in the project. The results of her survey are shown in Table 4. 37. De Obschatko then used regression analysis to study the relation between the degree of adoption, ranch profitability, and the rancher's per- sonal characteristics and habits, including age, education, place of principal residence, number of days spent on ranch each year, and total income. The results may be summarized as follows: (i) The degree of adoption, as shown by the adoption index, generally increased with farm size. (ii) The degree of adoption was positively associated with the rancher's education and the time spent off the ranch, which were taken as proxies for information and wealth. (iii) The average FRR obtained on the surveyed farms tended to increase with farm size. (iv) The number of days spent on the ranch by the owner tended to decline with farm size. 38. The regression analysis!/ also shows an important interaction bet- ween the FRR and the degree of adoption, the number of days spent on the farm, and farm size. Contrary to first appearances, the FRR increased with a simultaneous increase in the degree of adoption and the time spent on the farm, but decreased as each of these variables rose independently. Thus, the results implied a bimodal distribution of ranch profitability, with local 1/ See also Gimenez Dixon, op. cit., who studied similar cases. 2/ The study included quadratic terms in the regression analysis to test whether there was a positive interaction (on ranch profits) among tech- nologies as more technologies were adopted. The study found that such a positive interaction existed, i.e., adoption of the whole package produced better results than did piece-wise adoption. - 17 - maxima occurring where the farmer either adopts no new technologies and spends little time on the ranch, or adopts all the technologies and spend much time on the ranch. While this solution can only be taken as characteristic, the intensive technologies appear more profitable when adopted as a package and when the owner is continuously present and able to provide adequate management. 39. De Obschatko also suggested that small farmers, constrained by a lack of capital and/or information regarding the new technolgies and their appropriate management, appeared to consistently adopt less of the tech- nological package than would have been economically optimal, given that they were mainly resident owners, could manage the more intensive technologies correctly and thus gairp full benefit from them. By contrast, the large farmers appeared to suffer no capital and/or information constraint on adop- tion but many adopted more of the technological package than they could manage efficiently, given that they were largely absentee. Those larger and absentee ranches who adopted, therefore obtained low returns on the technologies utilized and found their overall rate of return reduced rather than increased. 40. The overall ranch FRR under the two polar cases (adopt and live permanently on ranch vs. do not adopt and do not live permanently on ranch) were highly similar, so that the choice between the intensive and the exten- sive strategies would not be taken on the basis of ranch profitability, but instead whether the rancher wanted and was financially able to live on the farm or in the city..L/ Because most large ranchers wanted to live in the city to exercise another profession, to supervise other investments, and to have access to better educational, medical, commercial, and social facilities, de Obschatko concluded that the technologies being considered would not be widely adopted in the project region unless permanently higher prices or improved management occurred. 41. The interaction between these three factors (price, management and technology) was important. With low livestock prices, which resulted from long standing Government policy, the benefits of intensive technologies were reduced, thereby providing less incentive for ranchers to provide improved on-ranch management. Without such management, the technology was only mar- ginally profitable, and intensive production did not appear as attractive as the extensive approach. 42. De Obschatko's results are only suggestive, but the results and interpretation seem consistent with much of the information which is available 1/ Real land price appreciation was high between 1950 and 1970, so that its inclusion in the overall rate of return skews the choice toward the extensive technology. Thus, if the rancher had the choice between investing in capital. intensive techniques or in the extensive techniques, de Obschatko concluded he would have been wiser economically to choose the latter. - 18 - regarding the Balcarce project's experience. In addition to the facts regard- ing the limited degree of adoption and the relatively low rate of prof it- ability on project investments, evidence from the INTA ex-post evaluation supports her hypothesis that the livestock technologies may be relatively more effective and profitable on ranches where the owner is continuously resident, and that such will often be the smaller ranches in Argentina: (i) A significant number of small ranchers (by standards of this region) participated in the project. Of project ranches, 39% had less than 500 ha (Category I), 41% had between 500 and 1500 ha (Category II), and 20% had more than 1500 ha (Category III)..!/ This distribution is identical to that in the 1960 census for 21 of the 41 counties in the project area (as cited in Gimenez, 1969). However, a rate of participation proportional to their representation in the region should be considered somewhat unusual, as the project did not con- centrate specifically on smaller ranches. Moreover, de Obschatko's data show that smaller ranchers in this region generally showed less inclination before the project to adopt the technologies being promoted, and the commercial banking system traditionally favored larger ranchers. The project therefore probably encouraged a significant change in their behavior and/or provided access to information and capital which was not previously available. (ii) Prior to the project, participating farms in the three farm cate- gories had improved pastures on respectively 12%, 17% and 24% of their cattle pasture area, i.e., larger farms had invested substan- tially more than had smaller farms. This result is consistent with that reported by de Obschatko on other farms in the region. However, the post-project evaluation shows that the small farms had fully caught up with the large farms; the average improved pasture area for the three categories were then 45%, 43% and 47%, respec- tively. (iii) Small farms had similar or slightly higher stocking rates, weaning rates, and beef production per hectare when compared with the larger farms prior to the project, and showed higher absolute and percent- age increases in technical coefficients (see the table below) in the post-project situation. 1/ Farmers in Categories I and II (80%) controlled 43% of the area on all project ranches, but 50% of the cattle on all project ranches (see also PPAM para. 24). - 19 - TECHNICAL COEFFICIENTS ON PROJECT RANCHES, BY RANCH SIZE Beef Production Stocking Rate Weaning Rate (kilos per hectare) BP AP BP AP BP AP Farm Size Category I 0.65 0.94 76 79 93 133 II 0.71 0.94 78 79 94 126 III 0.59 C.82 77 76 82 111 BP = Before Project AP = After Project Source: INTA ex-post evaluation. B. The Interest Rate Issue 43. The interest rate to be charged to subborrowers was an area of contention between the Bank and the Government throughout the life of the project, as in several other Latin American countries where the Bank assisted livestock projects.!! Sometimes there was no consensus among Bank staff on the same issue. 44. The interest rate issue arose because the Argentine livestock economy is afflicted by a number of important distortions, affecting both resource allocation and income distribution. Beef was a major food item whose price was an important determinant of the real income of urban consumers. Beef was also an important export good, whose price was thereby subject to sharp fluctuation because of variations in the relatively thin and unstable international markets. And beef was produced predominantly by large, rela- tively wealthy ranchers, a group which had vied for and lost political control to urban-industrial interests during the 1940's and 1950's. 1/ Serious differences of opinion about interest rates and monetary correc- tion occurred under the Brazil livestock projects; ultimately a sub- sidized Government Program was substituted and the Bank loan cancelled and prepaid; see PPAR - Brazil First and Interim Second Livestock Devel- opment Projects, OED Report 2402 dated April 5, 1979 (see especially paras. 8-18, and 24-26). A similar situation developed in Venezuela: the loan was cancelled but not prepaid; see PPAR - Venezuela Livestock and Agricultural Credit Projects, OED Report No. 1343 dated November 9, 1979. In addition, the following OED Reports also discuss the interest rate issue: PPAR - Uruguay Third and Fourth Livestock Development Projects, OED Report. No. 1321 dated October 20, 1976 (paras. 8.11-8.14); PPAR - Uruguay Fourth Livestock Development Project, OED Report 2572 dated June 29, 1979 (paras. 39-42). - 20 - 45. As a result of these factors, the Government historically used beef export taxes, domestic beef price ceilings, and overvaluation of the peso, to reduce beef prices to the urban consumer and thereby improve his real income, while also reducing the incomes of large ranchers. Government also used variations in export taxes and the domestic price ceilings, and in the real exchange rate, to reduce the impact of varying international beef prices, thereb attempting to alleviate the inflationary impact of such fluctua- tions./ Government recognized that these policies discouraged production, but politically the distributional issue was judged more important than the production issue. The Government had, however, made a concession to the ranchers by making short-term livestock credit available at subsidized inter- est rates, and thereby somewhat increased the production incentive for beef. But such credit went almost exclusively to large ranchers with good collateral and commercial-social connections, and much of it was diverted to non-live- stock ends.S Little long-term livestock credit was available. 46. The PCR (paras. 3.23 and 7.01-7.03) points out that Government policy toward the livestock sector varied significantly during the project, but frequently exacerbated rather than alleviated the domestic price fluctua- tions caused by external market forces. In 1965-69, Government maintained high export taxes on beef despite the decline in world prices during this period. When beef prices skyrocketed in 1970-73, the Government implemented direct restraints on producer beef prices, but permitted negative interest rates on livestock credits. Then, in 1973-76, when beef prices again de- clined, Government adopted a differential exchange rate system which penalized beef exports more than most other goods and imposed a domestic beef retail price ceiling. International prices remained low through 1978 and, although Government reduced export taxes in 1976, it subsequently allowed the peso to appreciate (taking into account the difference between domestic and interna- tional inflation), which restrained domestic beef prices. 47. In the Balcarce project, the Bank was principally concerned to introduce new livestock techniques, but also wanted to encourage the Govern- ment to adopt "sound" economic policies, by which it meant those policies which would reduce price distortions, reduce inflation, and improve income distribution. At times, these goals were in conflict. For example, the Bank wanted the Government to accept positive interest rates on subloans, 1/ Substantial discussion of international beef prices cycles and their impact on livestock sector profitability, and on other internal economic variables in countries which are important beef exporters, is contained in the two OED reports on Uruguay livestock development projects (OED Report Nos. 1321 and 2572, op. cit.). This material is relevant to the Argentine case as well, but is not repeated here. 2/ Project funds were granted on technical grounds and were carefully supervised so that these concerns were of no serious importance in relation to project financed subloans. - 21 - eliminating the "historical" credit subsidy. Indexing was seen as the proper approach because Argentina had a history of high, but variable inflation. Positive interest rates would have reduced the incentives for on-ranch invest- ment, but the Bank also wanted Government to forego beef export taxes and other beef price controls which it felt would increase beef prices, more than compensating for the higher interest rate. The combination of positive interest rates and highe:r beef prices was the Bank's policy emphasis. How- ever, increased beef prices would have had severe distributional and infla- tionary implications. The price increase resulting from the elimination of export taxes, for example, would have provided ranchers with a much greater income increase than did the credit subsidy, and would have caused a signifi- cant real income decrease for the average urban worker and an increase in beef prices affected inflation through its direct impact on the cost of living. Government was very concerned about both effects. 48. Although the Bank encouraged Government to provide producers higher prices throughout the project, most Bank attention was on the interest rate. Within the Bank, pressures were generated for a more liberal position on interest rates when implementation of the project appeared to be at stake, and a less liberal position when the project appeared to be disbursing well. For example, the identification mission argued that a negative interest rate on subloans would be acceptable because the Argentine producer was penalized by export taxes and fixed prices in the local market, thereby making investment in the livestock sector less attractive. The credit subsidy, it argued, somewhat offset this effect, particularly for those who wanted to improve their efficiency. This view was rather close to that of the Government, and was also the general view expressed by project staff throughout the project period. In contrast, country staff felt that an interest rate subsidy was an undesirable policy which should be strongly resisted. 49. The latter viewpoint prevailed at first and the Government was asked to accept indexing at negotiations. The Government refused, after which the Bank then accepted a 16% nominal rate on subloans which, given an average inflation rate of 25% in the period 1945-65, suggested an expected real interest rate of minus 10%. This rate was justified in the appraisal report as being similar to other commercial livestock loan rates in Argentina. 50. The project was negotiated in 1966, but not signed until 1967 because of a change of Government. Argentina's rate of inflation fell drama- tically in 1967-68, however, as a major stabilization effort was made. Inter- national beef prices also declined sharply during this period, assisting this effort. And when the peso was devalued in 1976, beef export taxes were increased to reduce the domestic impact; prices to the producer fell. The Wholesale Price Index rose about 10% on an annual basis from 1967 through 1969, so that the real interest rate on project subloans increased sharply, from minus 10% to a positive 6%. Since during the same period beef prices were low and falling, the Balcarce technology-credit package became less attractive. - 22 - 51. In early 1968, Government requested a reduction in the interest rate on project subloans, but it did not mention falling beef prices, empha- sizing instead that it was reducing the interest rates throughout the economy and wanted the project's subloans to remain consistent with this action. The Bank did not respond favorably, largely because the decision lay primarily with country staff, in contrast to the situation during negotiations. This shift is worth mentioning. 52. Because the Government was unwilling to accept the subloan terms, which on country grounds were believed preferable, compromise was reached during negotiations to permit successful negotiations. The project took precedence over the interest rate principles Nonetheless, once the loan was signed, economic principles became at least temporarily dominant when changes in the agreement were considered. When the Government's request for an interest rate reduction was received, country staff argued that interest rates in Argentina were finally becoming positive and that the opportunity to encourage improved resource allocation ought not to be turned down. They also feared that the decline in the inflation rate might only be temporary, so that any reduction accepted would imply a greater net subsidy over the longer run.L/ During this debate, the Government delayed project implementation. Eventually, concern with the project itself dominated again, and the Bank agreed to a series of nominal interest rate reductions, one in 1968 and two in 1969, but reserved the right to request increasing the interest rate if inflation should rise again. Throughout this period, some Bank staff con- tinued arguing for indexing. However, the Minister of Economy continuously refused, arguing that his stabilization program was based on a shift in expectations; introduction of indexing would counter this effort, as it would indicate that the Government expected significant inflation to continue. 53. Even after the first two nominal interest rate reductions on sub- loans, rancher interest in the project had remained low. In mid-1969, a Bank mission concluded that the project would have to be reappraised as it appeared that the project, within the then prevailing price and interest rate context, might be financially unviable (PPAM para. 16). This situation resulted in the last reduction in subloan interest rates, from 12% to 10%. International beef prices then began to increase rapidly, continuing their climb through 1973 and the increase was partly reflected in domestic producer prices. Inflation also rose, reaching nearly 50% in 1970. With beef prices high and real interest rates now extremely low, rancher interest in the project increased greatly. Project staff were relieved that the project had finally taken off, but country staff saw their previous concerns confirmed and argued for strong Bank action to ensure that the interest rates charged were positive. 1/ This opinion is based on Bank memoranda regarding these discussions. - 23 - 54. The original closing date was July 31, 1972. Although project lending had risen considerably, it appeared that funds would not be fully committed at this date. Moreover, after somewhat lukewarm interest in 1967- 69, the Government was eager to continue with the project and mentioned a second credit project which would also include funds for livestock develop- ment outside the pampas and for agricultural development. In this context, some staff argued that the Bank should not agree to an extension of the Balcarce project nor to a second project unless the Government accepted indexing. This position was accepted and communicated to Government. After some delay, the Government agreed to indexing. The indexing mechanism was introduced, the closing date extended, and preparation of a second project initiated. 55. During the period 1970-73, the Government utilized higher import taxes, domestic price controls, and peso overvaluation to restrain increases in domestic beef production. The Bank was not concerned with these issues because prices to producers were high, livestock investment was profitable, and interest rates were now expected to be positive. It might have been useful to discuss the desired response if and when beef prices would again decline and the technological package would become less attractive, i.e., whether a lower interest rate would be acceptable, whether the project should be cancelled in the event, or whether it should simply lie dormant while prices were low. 56. Viewed with hindsight, cyclical beef price fluctuations resulted in a cyclical demand for project funds and this probably could not have been avoided even by a different interest rate policy. In fact, because the indexing mechanism was imperfect (PCR, paras. 3.07 and 6.10), real interest rates remained negative through 1978. Because various policy mechanisms continued to depress the domestic price of beef throughout this period, this indexing system may have been betterl/ than one which would have charged producers 6% in real terms. Nonetheless, the system introduced was somewhat capricious, providing lower interest rates to ranchers who borrowed in one year than to those who borrowed in another, and was hardly instructive of how indexing ought to be carried out.Z 1/ By offsetting one distortion with another distortion. 2/ One justification used for the indexing mechanism actually adopted is that similar mechanisms had been accepted for use in Bank livestock projects in other Latin American countries. Thus, the decision to adopt in one country proved transitive. It is correct that different countries ought to be treated equally. But compromises in one country may have an unfortunate tendency to sneak into agreements elsewhere as well. - 24 - 57. Most of the issues involved in the indexing discussion are stated unusually well in several memos in the Bank's files, but two important issues were never fully clarified. First, projects staff wanted to fit the project into the existing (distorted) economic structure, while country staff gener- ally wanted to use the project to improve this structure, both through educa- tion and through lending leverage. Projects staff argued that the credit aspect was only incidental to the introduction of advanced ranching tech- niques, and that there was little prospect to persuade the Argentines to alter their interest rate policy for the sake of the project, and certainly not until the project had been proven and rancher credit demand had surged; they also argued that close project supervision could ensure that low interest rates did not result in badly spent funds. In contrast, country staff be- lieved that the interest rate subsidy caused allocational distortions and a distributionally undesirable transfer to wealthy ranchers. They argued that the Bank had pursued the same "positive" interest rate policy in other Latin American countries where livestock projects were also pilot projects and saw no reason to make an exception in Argentina. Although both positions are reasonable, the audit believes that the projects oriented argument was more effective. 58. Second, and related a more accurate estimate was needed of the actual financial and economic return to the project's technological package under different price conditions. A credit subsidy was probably attractive only if the technological package proved to be economically but not finan- cially attractive at the average prices, both international and domestic, over the whole beef price cycle. It appears this was the case. The credit subsidy might then have been specifically structured to offset this difference, as a "temporary" policy, with the Bank and Government then gradually seeking more optimal arrangement in the long run. The distributional and political problems associated with a beef price increase made it unreasonable to expect a major framework change in this area in the short run. C. Disbursement Procedures 59. The PCR (para. 3.10) points out that the BNA disbursed US$29.3 million for subloans, of which the Bank reimbursed only US$11.2 million, or 31% of the BNA subloans compared to the 50% agreed at appraisal. The lower amount reimbursed by the Bank is due to: (i) The long time lag, usually 6-8 months, between the time the BNA disbursed to the subborrower and the time the request for reim- bursement was submitted to the Bank by the Central Bank. (ii) Normal delays within the Bank in processing the reimbursement requests. (iii) The rapid depreciation of the Argentine peso during the project, because frequent devaluations were used to maintain parity given the high level of internal inflation. - 25 - The Bank reimbursed 50% of the peso amount previously disbursed by the BNA, converted at the exchange rate prevailing on the day of Bank reimbursement. 60. In the Fourth Uruguayan Livestock Project (second stage), a similar situation resulted in agreement between Government and the Bank that reim- bursements would be determined by the exchange rate prevailing on the day disbursements to the sub-borrower had been made, thus ensuring that the foreign exchange reimbursed to the Borrower would be similar to that agreed at appraisal. Alternatively, the percent of local disbursements to be reimbursed can be changed. The audLt su yests that operational guidelines be developed to cover this important issue.f 1/ This issue is characteristic for agricultural credit projects where the Bank refinances (in foreign exchange) a fixed percentage of loan disbursements (in local currency) to subborrowers using the exchange rate at the time of its disbursement. This point was recently also made in the Project Performance Audit Report on the Morroco Third Agricul- tural Credit Project, OED Report No. 3248 dated December 23, 1980. DFC lending, it seems, Is generally not as much affected by these differen- tials as on-lending is normally done in foreign exchange.  - 27 - Table 1 PROJECT PERFORMANCE AUDIT MEMORANDUM ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) Argentine Beef Prices, Domestic and Export (Indices, 1966 - 100) Steer Price, Heifer Price, Frozen Boneless Chilled Compensated Linier's Livestock Linier's Remate Export to EEC, quarters Year 11arket-/ Prices/b Market/c Feria/c FOB, Buenos AiresL4 FOB, Montevideo/e 1960 110 108 80 1961 90 99 74 1962 83 96 68 1963 98 104 61 1964 139 134 85 1965 132 125 106 106 1966 100 100 100 100 -- 100 1967 100 95 92 -- 83 80 1968 88 92 94 89 -- 80 1969 82 91 86 99 70 73 1970 109 112 100 91 75 75 1971 140 137 139 -- 96 1972 138 141 109 117 119 1973 137 133 118 133 136 1974 102 116 Ill -- 132 1975 69 84 62 51 66 1976 65 29 51 59 1977 86 65 51 85 1978 78 51 49 93 1979 112 90 1980 (115) 111 January 1966 112 (Appraisal) February 1980 113 (Project Unit Eval.) /a Linier-s market price per kilc of dressed weight, weighted by share of respective animal category. See Gustavo Noves, "Structure of the Argentine Beef Cattle Economy", Ph.D. Dissertation; Purdue 1972, Table 30. Deflation is by the official cost of living index. /b Livestock prices deflated by index of prices paid by farmers. See Lucio Reca, consultant-s report to the World Bank. /c Supervision Report, September 7, 1979, Table 7. The 1980 figure is estimated from the export price figure in colume 5 (frozen boneless export to EEC). /d PCR, Balcarce Livestock Development Project (Loan 505-AR). /e OED, based on data from National Meat Board of Uruguay (INAC). The nominal US$ figures were deflated by the U.S. cost of living index. - 28- Table 2 PROJECT PERFORMANCE AUDIT MEMORANDUM ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) Production Coefficients (as Appraised and as Realized on Ranches) Before After 3 YearsL. After 7 years After 9 years/b Project Value % Value % Value % Total Cattle Area AR - Partial Development 1,000 1,000 0 1,000 0 1,000 0 AR - Complete Development 1,000 1,000 0 1,000 0 1,000 0 Sample 2-7d 1,264 -- - 1,310 6 -- - Sample 1- 1,354 1,409 4 -- - 1,327 (-2) Area Improved - % of Ranch AR - Partial Development 0 300 /e 300 /e 300 /e AR - Complete Development 0 500 /e 500 /e 500 /e Sample 2- 279 -- - 609 109 -- - Sample 1- 267 490 84 -- - 612 129 Stocking Rate AR - Partial Development .80 1.00 25 1.20 50 1.20 50 AR - Comylete Development .80 1.20 50 1.80 125 1.80 125 Sample 2 .62 -- - .86 38 -- - Sample 1 .62 .86 39 -- - .94 52 Weaning Rate AR - Partial Development .70 73 4 82 17 82 17 AR - Complete Development .70 78 11 90 29 90 29 Sample 2- .77 -- - 78 1 -- - Sample 1-- .74 81 9 -- - 78 5 Beef Production per Hectare AR - Partial Development 65 71 9 152 134 172 165 AR - Complete Development 65 82 26 236 263 298 358 Sample 2-7 84 - - 115 37 -- - Sample 1-- 75 119 59 -- - 121 61 Partial Development Participating Ranches (7-9 years) (7.5-9.5 years) Slaughter Weights Weaner Calves 170 200 180 Fattening Cull Females 390 420 360 Bulls 650 650 550 9-24 month Steers -- 420 227 /a Participating ranches average 3.5 years. /b Participating ranches average 9.5 years. /c Ranches surveyed only in 1978/79. ) 7_ Ranches surveyed in 1972/73 and 1978/79. ) See PPAM para. 22 /e Infinite increase. - 29 - Table 3 PROJECT PERFORMANCE AUDIT MEMORANDUM ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) Average Physical Investment Per Ranch Appraisal Report Projection Partial Development Complete Development Project Ranches Area Improved (ha) 300 500 239 Fertilizer (kg) 45,000 (150/ha) 75,000 (150/ha) 42,977 (180/ha) Fencing (km) 8 10 5.2 Corrals (units) .33 .33 .25 Watering Points (units) 1.25 2.25 1.65 Cattle Heifers and Cows 100 300 80 Bulls 12 18 2.6 Tractors (units) .13 .50 .05 - 30 - Table 4 PROJECT PERFORMANCE AUDIT MEMORANDUM ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) Adoption of New Livestock Techniques in Ayachucuo Partido 1969/70 By Farm Size Survey I II III Average 200-1000 ha 1000-2000 ha 2201-750 ha Techniques % adopting --------% adopting------------ Permanent Pastures % producers adopting 50 33 64 93 % farm area covered 11 9 14 12 Fertilizer use 23 13 30 38 Rectal palpation 31 16 36 71 Early weaning 7 7 14 0 Vaccination against brucellosis 45 41 42 57 Seasonal breeding 18 16 29 14 Mineral supplements 39 35 57 36 Forage reserves 18 7 36 36 Global Adoption Index (average).La 3.05 2.04 4.14 5.07 Days Spent Each Year on Farm 212 245 191 132 FRR on total farm capital/b 2.5% 1.5% 3.7% 4.5% /a The global adoption index was designed to permit comparison among the ranchers. Two points were assigned for the adoption of each technique which required capital investment (permanent pastures, fertilizer use, and forage reserves), and one point for all others. Thus, the possible range of variation in the index was 0-11 on each farm. The figure shown is the average among ranchers. /b The rate of return was calculated as: (gross income - gross variable costs - depreciation)/capital, where capital includes invested capital plus land value. Land appreciation was not included in this return. For each of information on pre-investment output levels, she was forced to use a point estimate of the return rather than an internal rate of return. Source: de Obschatko, Edith, and Alain de Janvry, "Factores limitantes al cambio technologico en el sector agropecuario," Desarrollo Economico, Vol. II, 42-44 (July 1971/March 1972), Table 1. -31 - PROJECT PERFORMANCE AUDIT MEMORANDUM ANNEX 1 ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT (LOAN 505-AR) COMENTS FROM BORROWER Translation of incoming cable from Banco de la Naci6n Argentina, Buenos Aires to the World Bank, Washington, D.C. Date: October 30, 1981 Attention: Mr. Shiv Kapur, Director, Operations Evaluation Department I have no comments on the Project Performance Audit Report on Argentina Balcarce Livestock Development Project (Loan 505-AR). Regarding the Project Performance Audit Report on Argentina Agricultural Credit Project (Loan 1564-AR),1... . . . . . . . . . . . . ...... The above are strictly my own personal comments regarding BNA's partici- pation. They are not intended as any judgment on the interpretation of the policies of the national. authorities. Jose Manuel Ros Secretary of the Board of Directors 1/ For these observations see Project Performance Audit Report: Argentina Agricultural Credit Project (Loan 1564-AR), OED Report No. 3699 , dated November 30, 1981.  - 33- PROJECT COMPLETION REPORT ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT LOAN 505-AR March 31, 1981 Latin America and the Caribbean Projects Department Agricultural Division I  - 35 - PROJECT COMPLETION REPORT ARGENTINA BALCARCE LIVESTOCK DEVELOPMENT PROJECT LOAN 505-AR I. BACKGROUND 1.01 The livestock subsector comprises about one-half of total agricul- tural contribution to GDP and has increased its importance over crops during the last decade. Livestock output has increased by 21 percent in 1979 compared to the average of the 1974-78 period. However, livestock production has only increased by 12 percent in 1979, largely due to the cattle cycle, when compared to the average cattle slaughtered for the 10-year period 1967-78. Beef cattle account for about 30 percent of total agricultural production and about 20 per- cent of all merchandise exports. Land Tenure 1.02 About three-fourths of the land in Argentina is owner-operated. Although the area operated by tenants is only about 10 percent countrywide, it includes almost 20 percent of the land in the pampas. Sharecropping arrangements are infrequent in Argentina and other contractual arrangements are important only in the north. Land Extensive Agriculture 1.03 Argentina if; one of the major livestock producing and exporting nations of the world due largely to its abundance of fertile, highly productive land and relatively low population density. The agricultural sector contri- butes about 14 percent of GDP. Factor endowments of extensive land area (about 160 million ha under agricultural use) with excellent soils and low population pressure have accounted for the adoption of a traditional extensive pattern of production for beef, grains and oilseeds. Labor productivity in agriculture is almost as high as in the rest of the economy, with 15 percent of the country's employed labor producing about 14 percent of GDP. Though yields of the main agricultural export commodities are well below their potential under an intensified production system, and lower than those of the main beef and grain producing countries, Argentina's low production costs provide it with a comparative advantage in feed grains, wheat, oil-seeds and beef. Since the change of Government policies in mid-1976, grain and oilseed production has increased by about one-quarter reaching an annual output level of over 30 million m tons. Beef Raising Features 1.04 There are two clearly defined steps in beef production: breeding and fattening. One oE the major areas in the country for breeding cattle is in the Salado river basin in the eastern central part of the province of - 36 - Buenos Aires, which forms part of the project area. Other important cattle breeding areas are located in most of the semi-arid lands surrounding the humid pampas and in the open lands of the northwest and northeast of the country. Generally, in these areas, the breeding of livestock is, apparently, the best use of land. Fattening of feeder steers is mainly done in the pampas where beef cattle fattening is continually competing with growing crops, depending on the relative price ratio of beef and grains. Project Area 1.05 The Balcarce Project area includes 41 "partidos" of Buenos Aires province. The total area of the project is 16.5 million ha of which about half comprise the "Depressed Area" made up of two major river basins. About 20,000 ranchers with more than 50 head of cattle each populate the project area, which contains about 20 percent of the national cattle stock. The area has an ample supply of fertile land not directly effected by soil and water problems. It is near the domestic market of Metropolitan Buenos Aires with a population of 9.7 million and has easy access to additional urban markets of La Plata, Mar del Plata, and Bahia Blanca and to the meat-packing plants for slaughter and processing of beef for export. Small auction markets are owned and controlled by groups of farmers; terminal markets are well organized; livestock marketing, grading and price reporting are adequate. Infrastructure and communications are acceptable. 1.06 The soils are generally deficient in phosphorous. Up to 30 percent of the land along the Salado river is affected seriously by flooding during the winter and spring. There is also a problem of soil erosion due to wind and water and aggravated by the system of working the land in the hilly areas. 1.07 The mean temperature for the area is 140C to 160C, with the coldest month, July, averaging about 80C and the warmest month, January, averaging about 200C. Rainfall gradually decreases from east to west with an annual rainfall ranging from 900 mm in the west to 600 mm in the east. Rainfall accumulates in the spring and fall with occasional storms dropping 100 mm within 24 hours. The Loan 1.08 Loan 505-AR of US$15.3 million for livestock development was signed on July 31, 1967 and became effective on October 30, 1967. The project was designed as a concerted effort to induce ranchers in the project area to adopt improved technologies in pasture production, herd management and animal health practices which had been developed or identified in a previous UNDP/FAO research program. It was the first attempt in Argentina to provide an agricultural program of technical assistance to ranchers in the planning of investment and the supervision of herd management. Prior to closing this loan, a second loan (Loan 1521-AR) was signed on June 29, 1978 to finance: (a) on- farm development investments for livestock and crop production in selected areas of the pampas (including the Balcarce Project area) and mesopotamia; (b) weed control in the extensive area of the pampas, now heavily infested with weeds; and (c) land clearing in the northern provinces of the country. - 37 - In part, this project was to follow up the Balcarce Project in extending credit and technical assistance to ranchers for the purpose of increasing livestock productivity in the country. The total project cost was estimated at US$162 million and the Bank loan was US$60 million. Unfortunately, Government decided to cancel the loan in 1980. A separate project completion report is being prepared for this loan. II. FORMULATION Origins of Project 2.01 The rationale for the Balcarce Livestock Development Project is based on the apparently successful FAO research program in pasture and animal produc- tion carried out in the project area and referred to as the Sheep Reserve, which began pasture establishment in December 1963. Work on the Reserve has shown that a pasture of good quality, properly seeded and fertilized, with high carrying capacity can be grown in the area and, furthermore, by improved herd management and sheep health, higher lambing percentages can be reached. These conclusions were substantiated by a pasture and beef cattle development project financed by the Bank in Uruguay beginning in 1961. Land, soil and climate were similar to those found in the Balcarce Project and the Uruguay project was, in fact, a forerunner to this project. 2.02 In April 1965, the Ministry of Economy invited an FAO/CP mission to visit Argentina and begin project preparation. Unfortunately, the Ministry of Economy failed to inform adequately the principal Government agencies-- Secretariat of Agriculture, Agricultural Technological Institute (INTA) and the National Development Council (CONADE)--that the Government had requested the FAO/CP project preparation mission. Herein lies the beginning of lack of intragovernmental coordination which has proven to be one of the causes for delay of this project as well as for the serious problems encountered under subsequent Bank lending operations to the agricultural sector. The technical agencies were apparently not convinced of the need for improved technology in the Balcarce Project area and, hence, attempted to divert the preparation mission to the northern provinces. The identification of the project area was one of the first of several major issues during preparation of the project. Perhaps, as a position of leverage for the Ministry of Economy and the Bank, the Central Bank (CB) was brought into the picture as the executing agent of the project. This confused the issue of project organization and management as the loan should have been made to Government or directly to the National Bank (BNA) 1/. Furthermore, it soon became evident that after project signing CB had little or no interest in the project and this attitude prevailed throughout project implementation to the detriment of the project. The Technological Package 2.03 The technical package promoted by the project was finally agreed between the FAO/CP preparation mission and Government. The key input in the project was the application of phosphate fertilizer to newly seeded pastures 1/ Memo from A.D. Knox to L.J.C. Evans dated July 30, 1965. - 38 - with a combination of different pasture seeds proven to respond well in the area of the project. These newly seeded pastures were to be refertilized each year. Improved herd management was to be practiced through the use of the project investment in fencing, water points, corrals and cattle handling facilities, hay balers, tractors and other machinery. Animal health was to be improved with additional and seasonal vaccination of cattle. This technological package was to have a direct impact on carrying capacity of pastures, calving rates and production of beef per hectare. 2.04 It was hoped that the demonstration effect of introducing these techniques in a limited number of breeding ranches would lead to a rise in livestock production and weaner calves in particular. The project provided for the creation of a small specially trained group of technicians of INTA, under the guidance of a project director to assist ranchers in preparing development plans and to supervise their execution. The project also financed drainage and soil studies in different sites of the project area as well as beef and meat marketing studies. Most of the project funds were to be used for credits to ranchers for implementation of the technical package. The total cost of the project was estimated at US$39.1 million, including contributions by ranchers and contractors, BNA and the Government. 2.05 About 220,000 ha were to be improved on approximately 700 ranching units. Criteria for selection of participating farms by project staff was: (a) quality of physical farm resources; (b) ability and willingness of owner to implement required management techniques; and (c) creditworthiness. Eligibility was by type and quality of operation rather than by farm size, emphasizing the commercial and economic nature of the operation. No maximum or minimum farm or ranch size limit was imposed. During the first year's operation, about 120 ranches were expected to enter the project, followed by 240 the second year and 340 the third year for a total of 700 ranches over a five-year investment period. Financing 2.06 For on-ranch investment, the owners contributed approximately 20 percent of the estimated cost. For contractors' machinery, the contractors contributed 40 percent of the cost from their own financial resources. The remaining project costs were advanced to the ranchers and contractors as loans by BNA. In this manner, BNA financed 80 percent of the on-ranch investment and 60 percent of the contractors' investment. One-half of BNA's overall financing came from their legal reserves, since CB accepted the principle that this paper counted as satisfying their legal reserve require- ments. The other half of loans made by BNA were re-discounted by CB. 2.07 The terms and conditions of lending from BNA to ranchers included a 16 percent interest rate, repayable in nine years, including a four-year grace period. Contractors' loans carried a 16 percent interest rate, repayable in five years, including a one-year grace period. At the time of appraisal, farmers were paying about 15 percent annual interest for current commercial bank loans with repayment periods of up to five years. Inflation rates were clustering about 25 to 30 percent per 12-month period in the 1963-65 period. - 39- 2.08 The financing of technical services for project ranches were largely local cost (85 percent) and, hence, were financed by 1.5 percent, which would accrue to INTA, of the interest rate paid by sub-borrowers for project loan funds. The Bank provided US$290,000 (15 percent) of technical services, including vehicles and special equipment. However, the Bank loan would provide one-half the estimated cost of the drainage and soil studies and the marketing studies, and Government would provide the other half. Benefits and Justification 2.09 With project investment, ranches were to increase the stocking rate per hectare by 50 percent to 1.2 head per ha, to increase the effective calving rate by 17 percent to 82 percent and cattle sales by over 100 percent. Net ranch income after debt service was expected to increase by 160 percent after 10 years of development. The financial rate of return to the participating ranches was estimated to be about 25 percent. At the national level, annual, liveweight, beef production, after 10 years of development, was to increase by about 80,000 m tons (total national, liveweight, beef production averages about 4 million m tons per year). This increase was to be primarily in the form of calves sold for fattening to farmers and ranchers in other areas. If the fattened slaughter animals were taken into account, the project was to indirectly generate a total annual increase of 130,000 m tons liveweight per year equivalent to US$34 million in 1966 prices. 2.10 The credit provided under the Balcarce Project was considered instru- mental to introducing advanced ranching techniques on a pilot demonstration basis. The US$5 million per annum of credit expected to be granted from the loan is insignificant compared with the total agricultural credit or the sector's economic importance. In 1968, BNA alone made new agricultural loans of over US$300 million equivalent. Most of this lending was on a short-term basis, i.e., less than one year. Ranchers, who had sufficient collateral for loans, would often roll over short-term loans to finance capital goods. Total Argentine agricultural production in 1969 was US$3.2 billion and the beef producing sector alone had a gross output of over US$900 million. III. IMPLEMENTATION 3.01 The outstanding feature of project implementation is the long delay in disbursement of loan proceeds, which required about 13 years. The astounding rates of inflation, extreme shifts in Government livestock policies, the swing of the cattle cycle and a frequent change of interest rate levels and methods of calculating these rates inhibited project execution and strongly affected the performance of the project. To deal with this multitude of causes and effects on project performance in an orderly fashion, this chapter is divided into two sections: first, a statement of the facts, including the physical investment items, the changes in loan proceeds allocations, Closing Dates and the number of ranch development plans, subloan commitments and subloan disbursements; and, second, a critical historical review of beef prices, Government policies, interest rate problems and other factors causing the resulting project performance, divided into five chronological periods beginning in 1967 and ending in 1980. - 40- 3.02 The Closing Date on this loan was originally set for June 30, 1972 and was postponed three times, once to June 30, 1977, again to June 30, 1979 and finally to July 31, 1980. There have been three reallocations of loan proceeds to cover cost overruns for technical services and drainage and soil studies. The final accounting of loan proceeds as compared to the original allocations are as follows: Original Allocation Final of Loan Proceeds Disbursement October 17, 1967 November 13, 1980 ------ (In Current US$ '000)------- Subloans to ranchers and contractors 12,200 11,803 Technical services 260 1,427 Drainage and soil studies 1,220 1,720 Marketing studies 120 - Unallocated 1,500 - Cancelled - 350 Total: 15,300 15,300 The cost overruns in technical services are due to various factors of which the most important are: (a) the long duration of the project; (b) a revision in the method of Bank financing to cover 50 percent of operating costs, including salaries of project field agents; and (c) expenditure of about US$400,000 for telecommunications equipment and services, which was agreed between Government and the Bank. The beef marketing studies were considered less important over time and, eventually as unnecessary, as the National Meat Board (JNC) assumed this task. Project Cost 3.03 The actual project cost is difficult to determine because of the high rates of inflation of the peso which increased almost 3,000 times and US dollar inflation which increased over one and a half times during the 1967-80 period. The exchange rate of the peso to the dollar was continually under control by Government and did not, generally, reflect the difference in the respective inflation rates of the peso and dollar. For estimation of actual project cost, local currency figures were first converted into current US dollars using the official exchange rate, and, then, US dollars were deflated to 1966 prices using the US wholesale price index. Under these assumptions, the project cost is shown below: - 41 - Estimated at Appraisal Actual Actual (In US$ '000 of 1966 Prices) (In Current US$ '000) On-ranch investment 32,670 23,697 33,714 Contractors' machinery 1,380 987 1,638 Technical services 1,940 1,714 3,088 Drainage and soil studies 3,110 4,789 7,218 Total: 39,100 31,187 45,658 The average on-ranch investment was US$21,000 in 1966 prices or less than half of that estimated by the appraisal mission. Number of Ranches Assisted 3.04 The number of ranches assisted under the project was 1,174 as shown in the table below: Estimated at Appraisal Actual Actual (In US$ '000 in 1966 Prices) (In Current US$ '000) Number of ranches assisted 700 1,174 1,174 Total on-lending 26,140 20,233 29,265 On-lending per ranch 37 17 25 On an additional 173 ranches, project staff made ranch development plans and, generally, ranchers followed through with the planned investment, but later cancelled the outstanding balance of the subloan, particularly in the 1975 and 1976 period when low cattle prices coincided with substantial increases in the adjustment index for outstanding principal. Project Financing 3.05 The sources of project financing are shown in the following table: - 42 - Sub- BNA/ borrowers Government World Bank Total Cost Item US$ % US$ % US$ US$ % ------------(In US$ '000 of 1966 Prices) 1/-------- Cattle purchase 1,135 20 4,542 80 - - 5,677 100 Other on-farm investment 3,604 20 5,947 33 8,469 47 18,020 100 Sub-total on-farm: 4,739 20 10,489 44 8,469 36 23,697 100 Contractors' machinery 395 40 365 37 227 23 987 100 Technical services - - 1,063 62 651 38 1,714 100 Drainage and soil studies - - 3,113 65 1,676 35 4,789 100 Sub-total technical services: 395 5 4,541 61 2,554 34 7,490 100 Total financing: 5,134 16 15,030 48 11,023 36 31,187 100 1/ Local currency figures first converted to US current dollars using the official exchange rate, then, converted to 1966 US dollars using the US index of wholesale prices. The sub-borrowers' share of investment was only slightly lower than the estimates in the appraisal report (16 percent as compared to 18 percent) due to the reduc- tion in the on-ranch investment component. BNA/Government's participation increased from an estimated 35 percent to 48 percent due to the loss of value associated with devaluations and time-consuming reimbursement application procedures and cost overruns in the drainage and soil studies. Consequently, the Bank's share was 36 percent instead of the 39 percent originally envisaged. The Bank did not finance the purchase of breeding stock under the project. Because the Balcarce Project was unique in the planned approach toward live- stock development, borrowed project funds were not substituting other sources of financing. But other sources of funds did finance additional investment packages, to a limited extent, based on the Balcarce experience in the project area as well as outside the area. Indexation of Subloans 3.06 As already mentioned, the original Closing Date of the loan was July 31, 1972. On July 13, 1972, the Vice-President of the Bank wrote to CB advising that "it would be very difficult for the Bank to agree to an extension of the Closing Date" because of the negative rates of interest under the project. On July 31, 1972, CB sent a letter to the Bank requesting extension of the Closing Date for five years--two years for further loan commitment plus three for disbursement. About US$3.6 million was disbursed - 43 - out of US$15.3 million at that time. In regard to indexing, the letter only requested more information on "the index or indices which could be applied to give a more rational rate of interest." On August 14, 1972, the Bank sent a cable to CB threatening to stop disbursement if subloans were not indexed. Government responded with a willingness to work with the Bank on the introduc- tion of an indexing mechanism and the Bank agreed (cable dated August 29, 1972) to continue loan disbursement beyond the Closing Date, for subloan commitments made prior to July 31, 1972. On October 16, 1972, the Minister of Finance of Argentina expressed in a letter to the Bank President official acceptance of the principle of indexing for the Balcarce Livestock Development Project. In a letter of October 20, 1972, the Bank further informed CB that upon the introduction of a satisfactory indexing system, the Bank would postpone the Closing Date until July 31, 1977 as requested by the Government. On December 28, 1972, CB stated in a letter that an indexing system based on the CPI and on the liveweight price of cattle at Linier's market would begin on January 1, 1973. And on February 9, 1973, a Side Letter to the Loan Agreement confirming the above was submitted to the Bank and countersigned by the Acting Director of Country Programs Department. This Side Letter set the interest rate to sub-borrowers at 6 percent based on the monetary adjustment of the outstanding balance of the subloans. After considerable correspondence and ambiguity over the subloans committed by BNA prior to January 1, 1973 but not disbursed, the Bank finally agreed in a letter sent by the Vice-President on March 8, 1973 to CB that the Bank would reimburse CB for about US$96,000 (50 percent of subloans) not indexed in supplemental financing to sub-borrowers during January 1-March 15, 1973. After March 15, 1973, all subloans were indexed. But the Bank would not agree to reimburse CB for about US$1,075,000 (50 percent of subloans) of new subloans approved during January 1-March 15, 1973. 3.07 The consumer price index increased 142 times between September 1973 and July 1978, while the average price of beef cattle at Linier's market increased 480 times. The beef price at Linier's market largely declined or remained constant in nominal terms between September 1973 and June 1975 and, therefore, there was little actual adjustment of outstanding principal 1/. However, as major adjustment to principal occurred with beef price increases and accompanying inflation increases, subloan outstanding balances began being adjusted in 1975. 3.08 Ranchers were unaccustomed to the indexing system and BNA branch office personnel did not. understand how the indices were used to arrive at the individual adjustment coefficient, which undermined the confidence of the ranchers in the system. Furthermore, a rancher could not even fully cancel his outstanding balance, because complete adjustment could only be determined after the index numbers were available to BNA headquarters where the system was managed. This office at headquarters reviewed the calculations for each cancellation to assure that the correct method was followed by BNA branch officials. 1/ The reference index is based on the monthly percentage changes of the Consumer Price Index for Buenos Aires or of the average weighted price for live animals at Linier's market, whichever is the lower for the months. Therefore, the reference index is a composite index and, although cumulative, is always less than either of the two basic indices, CPI and Linier's market. - 44 - 3.09 Apart from the merits of indexing, it appears, in retrospect, to have been an inappropriate decision on the part of the Bank to impose an indexing system on a subsector of a country when the rest of that subsector activity as well as the entire economy was without indexing. The imposition of indexing in the project caused confusion and cancellations and set a tone of uncertainty for the continuation of the project. Disbursements 3.10 The number of on-ranch development plans by INTA and commitments and disbursements of subloans by BNA, as well as reimbursement of appropriate project cost by the Bank, over the period of the project implementation are shown in the table below: Subloans IBRD Reimbursement Programmed (INTA) Committed (BNA) Disbursed (BNA) for Subloans Year Per Year Accum. Per Year Accum. Per Year Accum. Per Year Accum. ----------(In Current US$ '000 Equivalent) ----------------- 1968 119 119 - - - - 1969 906 1,024 906 906 372 372 - - 1970 4,710 5,734 3,213 4,119 2,121 2,492 - - 1971 6,603 12,337 6,156 10,375 2,964 5,456 1,575 1,575 1972 8,381 20,717 5,783 16,058 4,528 9,984 1,282 2,857 1973 3,842 24,559 5,566 21,623 4,195 14,178 1,459 4,316 1974 5,036 29,595 6,703 28,327 6,871 21,048 2,517 6,833 1975 2,716 32,312 1,089 29,416 1,727 22,775 2,410 9,243 1976 706 33,018 593 30,009 546 23,321 157 9,400 1977 5,295 38,313 2,699 32,708 3,367 26,687 830 10,230 1978 2,629 40,941 3,027 35,734 1,433 28,120 712 10,942 1979 651 41,592 1,302 37,036 976 29,096 62 11,004 1980 204 41,796 184 37,220 169 29,265 186 11,190 Among other points, this table shows that BNA disbursed US$29.3 million, of which the Bank reimbursed only US$11.2 million, which was only 31 percent of BNA subloans compared to 50 percent estimated at appraisal. In addition to these subloans, BNA disbursed an additional US$2.15 million equivalent in unindexed subloans in 1973 on which the Bank refused to reimburse, because - 45- of delays on the part of Government to apply indexing. During most of the project, Argentina experienced high levels of inflation and, consequently, frequent devaluations of the peso. Those rapid devaluations had the undesirable effect on the disbursement system in that Government was not fully reimbursed for expenditures as agreed due to a six- to eight-month lag between the time BNA actually made the subloan to ranchers and the time that CB sought reimbursement of 50 percent from the Bank. For example, BNA gave a rancher a subloan of $a 2,000 when the exchange rate was $a 100 to US$1.00, but when CB sought reimbursement from the Bank, the exchange rate was $a 200 to US$1.00. Therefore, instead of receiving US$10 for subloans made, in fact, Government only received US$5. By the conclusion of project, BNA actually disbursed US$31.4 million equivalent in comparison to appraisal estimates of US$24.4 million. The chart below shows BNA project disbursements by year in 1979 US dollar equivalent. BNA Project Disbursement by Year for 1970-79 (In 1979 US$ and Index Number) 12,000 40,00 9,000 - YEARLY OISEURSEMENTS - 0)) ACCUMULATED DISBURSEMENTS I (LEFT SCALE) r 6,000 - - 20,000 0 3,000 \-10,000 0 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 - 46- 3.11 Because of the detailed planning of ranch development and the subsequent supervision of investment by project field agents, practically no diversion of funds occurred under the project. Cattle purchases were tightly controlled both by the field agents and by BNA personnel. Disburse- ments were made against proforma invoices and payment vouchers. On-ranch Investment 3.12 Actual on-ranch investment exceeded appraisal estimates in the four major investment items--pastures, fertilizer, cattle and fencing as shown below. Actual Appraisal As of As of Item Unit Estimate August 1973 End Project Pastures ha 222,000 210,900 265,637 Fertilizer m ton 33,000 32,010 56,886 Cattle head 90,760 70,793 106,612 Fencing km 5,720 4,061 6,156 Project investment goals were over 80 percent fulfilled by the end of 1973 and, in that sense, the project was quite successful in meeting its objectives within the estimated time period. Actual fertilizer application exceeded appraisal expectation by over 70 percent and pasture establishment and cattle purchases by about 20 percent. These four investment items account for about 82 percent of actual project investment. 3.13 The composition of project investment items varied somewhat in terms of value when comparing appraisal estimates with actual as shown below: Item Appraisal Estimate Actual - (Percentage Distribution) - Pastures 22 25 Fertilizer 12 12 Fencing 18 21 Cattle 34 23 Machinery 9 5 Water points 3 10 Stock handling facilities and sheds 2 4 Total: 100 100 - 47 - These variations reflect changes in prices rather than in volume. Pastures, fertilizer and fencing generally showed similar shares of project investment, with slightly higher participation of pastures and fencing. Cattle purchases were down relative to appraisal estimates while machinery, water points and stock-handling facilities and sheds were substantially more important in the composition of investment than originally estimated. Delays in Project Execution 3.14 The explanations of the delays in project execution are indeed complex. To understand the impact of different causes of delays on project performance, the project implementation period has been divided into five periods: (a) 1968/69 when US$372,000 equivalent or only 1 percent of total BNA disbursement made was disbursed to finance on-ranch development and contractors; (b) 1970-74 with US$20.7 million or 71 percent; (c) 1975/76 with US$2.3 million or 8 percent; (d) 1977/78 with US$4.6 million or 16 percent; and (e) 1979/80 with US$1.1 million or 4 percent. 3.15 The rate of sublending under the project over time has been uneven. For the first two to three years after effectiveness, very few subloans were made. However, in 1970-74, there was strong demand, which slackened during 1971 and increased again in 1972. During 1975 and 1976, there was very little demand and a number of producers cancelled their applications and/or their subloans. Since the beginning of 1977, demand again increased, but only to drop significantly in 1978-80. 3.16 Slow Start-up 1968/69. The common denominator throughout the project implementation period was the cattle cycle, which determined beef prices and price expectations of ranchers. The movement of internal beef prices during project execution is shown in the chart below. Liveweight Price of Beef Cattle at Linier's Market in Buenos Aires and of Feeder Steers at Local Aucitions in Project Area (Pesos per kg in 1979 Prices) 2,000 LINIER'S MARKET 1,500 14 1,000of AUCTIONS IN PROJECT AREA 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 - 48 - The price per kg of beef at the Linier's market in Buenos Aires was quite similar to prices paid for feeder steers at local auctions in the project area. The project began when world beef prices were on the decline. For example, for frozen boneless manufactured beef export to the EEC, fob price per m ton was US$2,100 in 1965, US$1,630 in 1967 and US$1,385 in 1969 when calculated in 1978 prices. The chart below shows the movement of export beef prices in relation to prices at the Linier's market and at local auctions in the project area. Indices of Liveweight Price of Beef Cattle at Linier's Market in Buenos Aims and of Feeder Steers at Local Auctions in Project Area and Price of Argentine Frozen Boneless Beef Exports to EEC (1979 = 100) 200 .. ARGENTINE FROZEN BONELESS so 7BEEF EXPORT TO EEC 0 AUCTIONS IN PROJECT AREA 2 LINIER'S MARKET -50- - 01 1 I 1 1 1 1 1 11 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 During this period, price expectations among feeder cattle raisers were low, which, in turn, created uncertainty and lack of investment among ranchers. Ranchers were not inclined to take risks by intensifying land use through an investment package which was not commercially proven at that time. 3.17 The tradition of absentee ranch owners in Argentina and their practice of maintaining herds on minimum investment outlays constituted a problem for the project. Despite efforts to promote the project, ranchers had little confidence in the relatively young INTA field agents who attempted to show them how to improve ranching practices. This lack of confidence of ranchers vis-a-vis INTA staff was never entirely overcome during project implementation and led Government to bring in the Regional Agricultural Research Group (CREA) staff on the repeater project. Furthermore, ranchers had fears that INTA involvement with their ranching operations affairs could lead to possible reassessment of taxes and litigation with the provincial government. With hindsight, it appears that, at least in the initial years, INTA maintained criteria that appeared to be excessively rigid regarding the selection of beneficiaries. This tone was set early in the project by the expatriate director of the project, who refused to accept many applicants, on grounds of financial, technical or managerial considerations. - 49 - 3.18 Continued haggling between Government and the Bank over the interest rate during the first two years of the project deflated much of the enthusiasm in BNA and CB for the project. In effect, the 16 percent interest rate originally established for the project was never used and the lower rate of 15 percent was finally agreed after strenuous debate. However, interest rates were revised four times during the period 1967-69 (see para 6.09 for historical review of project interest rates). 3.19 The Boom Years 1970-74. By 1970, beef herds were depleted and the cattle cycle reached the bottom of the trough. Prices, in turn, began to rise dramatically on the world market. Argentine export beef (fob) increased from US$1,385 per m ton in 1969 to US$1,500 in 1970, US$2,320 in 1972 and peaked to US$2,630 in 1973 in 1978 prices. During that period, ranchers began building up herds by retaining stocks. Beef slaughter which reached 14 million head of cattle in 1969 dropped to about 10 million from 1971-74. The chart below shows total beef production and beef exports during the life of the project. Total Beef Production, Beef Exports and Internal Consumption (In Millions of Heads of Cattle) 20,000 < 15,000- O TOTAL PRODUCTION 10,000 - 10,INTERNAL CONSUMPTION o 6,000- EXPORTS -J 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 Expectations were that high beef prices would remain for some time. The price ratio between prices of cattle and prices of fertilizer 1/ increased from a low of 1.7 in 1969 to 2.1 in 1970 and 2.8 in 1971. 1/ Steers, liveweight, 90 percent of Linier's market price and triple super phosphate. - 50 - 3.20 After successful efforts on the part of Government to hold the increase in the wholesale price index at 9 percent in 1968 and 6 percent in 1969, new pressures began being felt as the index began to climb to 14 percent in 1970, 40 percent in 1971, 77 percent in 1972 and 50 percent in 1973. Despite actions on the part of Government to increase interest rates to sub-borrowers, they were insufficient to cover the increase in the inflation rate, giving ranchers further incentive to take out long-term subloans under the project. Substantial negative interest rates were the result during this period throughout the economy. Government attempted to compensate somewhat for this leakage by increasing export taxes on beef up to 40 percent of the fob value. 3.21 Not the least important factor accounting for improvement in project performance was the rather successful experience of the technological package on several ranches of project beneficiaries. This demonstration effect, complemented by improved practices by INTA field agents, was an important reason for better acceptance of the project, given the encouraging investment climate prevailing at the time. 3.22 The Stagnating Years, 1975/76. In 1975, world market prices for beef plunged from US$2,630 per m ton in 1973 to US$1,000 in 1975/76 in 1978 prices. This drop in prices was even greater for the Argentine producers because the EEC virtually closed its doors on high quality table beef imported from Argentina. Ranchers began to sell off stock, reflected in the increase in cattle slaughter which rose to 12 million head in 1975 and 14 million in 1976. The price ratio between cattle and fertilizer dropped from 2.8 in 1971 to 0.1 in 1975 and 0.5 in 1976, indicative of the terrible cost-price squeeze in which producers found themselves. 3.23 Income to producers in the declining market was further reduced by the damaging agricultural policies of the Peron Government. Not only were high export taxes imposed but also differential exchange rates further decreased ranchers' incomes. By 1975, Government maintained an array of exchange rates which squeezed agriculture and favored industry. Furthermore, a price ceiling was introduced on beef sold to consumers. 3.24 Perhaps the most damaging factor to project progress was the indexing of subloans. While introduced in 1973, it had little effect until 1975 when the price of beef bottomed out and real adjustments hit sub-borrowers at quite high rates, depending on the month in which the subloan was taken. The confusion over the indexing system at the farm level and the fear of unfore- seen increases in the future in the face of unprofitable business in cattle caused significant cancellations of outstanding balances of subloans and caused ranchers not to participate in the project. The impact of the indexing system was all the more pronounced as it was the only one used by BNA and clearly the only one in use by the agricultural sector (para 3.09). 3.25 Counterpart financing of INTA field agents became a major issue as Government had to cut back on all expenditures due to the poor performance of agricultural exports which provided the source of finance for INTA. Only at the insistence of Bank supervision missions did Government continue some financing of the project unit. The lack of financing had a demoralizing impact on field agents. - 51 - 3.26 Signs of Recovery, 1977/78. World beef prices did not improve during 1977/78 for Argentine beef and remained stuck around US$1,000 per m ton (fob in 1978 prices). However, the military Government which took power in March 1976 launched a series of policies which improved real incomes of ranchers. Export taxes on beef were eliminated, the exchange rate was unified and the price ceiling on beef to consumers on the internal market was eliminated. However, ranchers still found themselves with excess cattle and preferred not to retain stock as at the beginning of the 1970-74 boom. Slaughter of cattle continued to increase in 1977 to 14.5 million head and in 1978 to 16 million, the beef largely absorbed by internal consumption which increased by 25 percent during 1977/78. Furthermore, grain production was attractive during this period and many farmers who fattened beef substituted crops for pastures. 3.27 Government also began protracted discussions on indexing or high nominal monetary corrections of subloans for all sectors in order to arrive at real rates of interest. It first sought to index all subloans to the wholesale price index, and such was the policy during 1977/78. This policy caused uncertainty and unfamiliarity on the part of ranchers and little borrowing occurred. 3.28 The project unit was unclear as to the balance of funds remaining to be committed because of the lag, often six to eight months between the time subloans were made by BNA and reimbursement was made by the Bank, with the undesirable effect that, under the circumstances of rapidly changing exchange rates between the peso and dollar, the Government was only partially reimbursed for cost incurred. Therefore, the project accounts showed unexpected dollar equivalents to be committed. Due to the lack of coordination between the project unit and BNA, the former assumed from time to time that project funds were fully committed and virtually stopped promotion of the project. 3.29 Uncertain Investment Climate, 1979/80. Despite the fact that world beef prices climbed from US$960 per m ton in 1978 to US$1,780 in 1979 and over US$2,200 in 1980 in 1978 prices, ranchers apparently did not borrow for invest- ment purposes. Real incomes of producers were reduced due to the Government policy of fixing in advance an overvalued exchange rate, which had an adverse impact on prices in pesos for beef sold. The overvalued exchange rate imme- diately affected prices to producers but ranch input suppliers were largely unresponsive to cheaper imports and maintained relatively high prices for machinery, equipment and materials. IV. AGRICULTURAL IMPACT 4.01 The impact of the project on beef production and productivity was measured by five variables: (a) the area of land in natural and established pasture; (b) the area in established pastures only; (c) the stocking rate; (d) the weaning rate; and (e) beef production per hectare. The results are derived from a sample survey carried out by the project unit in 1978/79 under the supervision of Bank staff. - 52 - 4.02 Generally, the area in natural and established pasture has not changed over the period of the project and land use of the project ranches has remained similar to what it was before the project. In the sub-areas where beef production is relatively more important, small shifts out of crops are attributed to flooding in 1978, the year of the sample survey. Improved pastures more than doubled over the period of the project with ranchers increasing improved pastures as a percentage of total pastures, from 22 per- cent without the project to nearly 50 percent with the project. The stocking rate increased nearly 40 percent to 0.93 animal unit per ha as compared to the appraisal estimate of 1.2 animal units per ha. 4.03 The average weaning rate of 78 percent remained, more or less, the same with the project and was about 4 percent below the appraisal estimate for full development. Ranches with less than 500 ha showed a better performance than larger sized units. There is a general feeling among project agents that "without" project coefficients were over-estimated by the original survey, taken at the time the plans were being prepared when ranchers had to recall by memory in most cases their experience in weaning rates. The beef production per hectare increased 37 percent with the project or an absolute increase of liveweight beef in excess of 30 kg of beef per ha per year. The annual average production of beef with the project was 115 kg per ha compared to 94 kg without the project. Those parts of the project area where beef production per hectare was the lowest before the project showed the highest increases, the highest being 40 kg per ha additional in that part of the project area with the poorest soils and greatest incidence of flooding. 4.04 An interesting feature of the productivity gains attributed to the project was the shift of ranchers to fatten steers of their own as well as to maintain a breeding herd to raise calves for fattening. Through improved pastures, herd management and better animal health practices, project ranchers were able to diversify their output to feeder and fattened steers so as to spread market risks. This practice was particularly evident in those parts of the project area where wheat or potatoes were an integral part of the farming systems. 4.05 At full development of the project in 1979/80, incremental beef production due to the project was about 30,000 m tons compared to a before project production of 80,000 m tons and national beef production of 6.1 mil- lion m tons. 4.06 Beef cattle raising is a land-extensive activity requiring little human labor. However, due to increased area of established pasture, fencing, physical improvements, better animal health care and more intensive herd management, the project required about 1,900 man-years of labor to carry out on-ranch tasks due to the project; annual labor employed due to the project corresponds roughly to one additional man-year for each of the 1,174 ranches included under the project. - 53 - V. RATES OF RETURN 5.01 Taking into account increases in tax payments of participating ranchers and also the cost of technical services to ranchers, the financial rate of return (FRR) of the project was estimated at appraisal at approximately 27 percent. This exercise was carried out in 1966 prices. The results of the sample survey applied to the entire project by the project unit indicate that the actual FRR for the project was about 14 percent. This exercise was carried out in 1980 prices. The FRR would, most likely, decline to about 10 percent if average annual prices of inputs and investment as well as prices of beef cattle over the period were applied, adjusted to 1979 prices, since beef prices were substantially depressed in the 1974-79 period and remained below the relative prices of inputs and investment items. 5.02 The assumptions used in calculating the FRR are based on the sample survey carried out by the project unit in 1979. Under the project at full development, ranchers increased beef production per hectare by 32 kg per annum. This 32 kg per ha per annum was disaggregated to 3 kg additional the first year, 6 kg the second, 10 kg each for the third and fourth years and 3 kg the fifth year after which production per hectare per year was assumed to remain constant. On the average, the pasture area per ranch accounted for about 82 percent of the total ranch area. At full development in 1979/80, the project accounted for about 30,000 m tons of additional beef production. The composition of cattle sales was based on different kinds of cattle sold under the project such as feeder steers, heifers, calves, cows and bulls. A weighted average price was derived based on the amount of different kinds of cattle sold at prices of February 1980. A 7-percent discount of that weighted value per kilogram of liveweight beef was assigned to cover marketing costs. 5.03 The additional costs considered in the re-evaluation were the direct investment costs of project items and variable costs which basically included additional expenditures in animal health and maintenance of machinery purchased and improvements undertaken by project beneficiaries. 5.04 The economic rate of return of the project has not been recalculated but is estimated to be similar to or slightly higher than the financial rate. This takes into consideration the economic cost implicit in the subsidized interest rates, the export tax levied at various times on project output, the accumulated tax to ranch beneficiaries over the life of the project equivalent to about 2.7 percent of total beef cattle sales (animal health taxes, municipal taxes, sales tax and a potential profit tax), and the variation in the value of the peso over the life of the project. - 54 - VI. INSTITUTIONAL PERFORMANCE AND DEVELOPMENT 6.01 Traditionally, Argentine ranchers financed their investments and operating expenditures through their own resources and through the commercial banking system, which includes BNA. The project continued to use that system, designating BNA as the initial commercial bank to handle project funds, but it was foreseen that other banks, such as provincial banks, would be permitted to participate in retailing and supervising project subloans. CB was named the project executing agency and authorized BNA to finance up to one-half of the subloans from their legal reserves. It rediscounted the other half and subsequently asked for reimbursement from the Bank loan. INTA was responsible for promoting the project and providing technical assistance in carrying out ranch development plans with subsequent supervision of project investment. First thoughts of the Bank were to have full-time services from a small project staff, but INTA insisted upon technical control of the project through its research station located in the project area. It was reasoned, and subsequently proven correct, that the project could provide INTA the opportunity to become more involved in the development field and to strengthen its work through close cooperation with credit agencies. 6.02 The effective coordination among these three institutions left a great deal to be desired. Government, through its Ministry of Economy, was not successful in creating an operational committee to take decisions on project implementation. It soon became evident after loan signing that CB was removed from project implementation and overshadowed by BNA with its extensive commercial operations. INTA was the institution most involved with the project, and the financing of that part of its operations related to the project became an important feature for receiving financial support from Government during difficult periods of Government liquidity. However, being a research and, to a limited extent, an extension institution, INTA knew little about banking practices and maintained a distance from BNA and barely had communication at all with CB. It appeared to Bank staff that the only time of the year that representatives of the three institutions met was during the visit of the supervision mission. 6.03 The project concept was based on thorough technical assistance to ranchers who received subloans. The first project director was an expatriate who insisted on approving credit only for ranchers who fulfilled a strict criteria of adequate soils, creditworthiness, infrastructure, herd size and willingness to follow investment plans. The project unit wanted to develop "demonstration" ranches over the first three to four years, which would prove the viability of the technological package to other ranchers in the project area. While this approach appears technically sound, it did not correspond to the scheduling of ranch development plans and project disbursement designed by the appraisal mission, which assumed that the experimental ranch at the Balcarce Research Station had already proved the technological package commer- cially viable. Project Unit 6.04 The project unit comprised a director, a deputy, two supervisors and 20 field agents. It was a practice of the project to give training to livestock and agricultural technicians, all having graduated from the - 55 - university as agronomists. Because of the solid training received by the agents from the project unit and the well-designed work they did in making ranch development plans, they soon received better offers from the private sector, which caused considerable turnover of staff. Also INTA maintained the practice of contracting agents for one to two years and not directly hiring them as permanent staff. On the average, there were about 18 project technicians per year throughout the project implementation period. They prepared 1,347 ranch development and contractor plans, which made an annual average of about seven plans per field agent per year in addition to super- vision of those plans under investment. 6.05 The technical assistance given by project agents was thorough and intensive. An average of three visits to the ranch was required before the development plan was ready to be presented to INTA and, subsequently, to BNA for approval and disbursement. The plan was for a three-year investment period and included a physical layout of the fields and a timetable for phasing into established pastures; purchase of breeding stock; installation of fences, water points, and cattle-handling facilities; and machinery. Herd build-up models and a cash flow for 10 years were prepared by the agents. All investments carried out on the ranch were confirmed by the agents, which entailed an intensive supervision on the investment. Agents were also required to promote the project through talks to groups of ranchers and visits to ranches. The ranchers' technical education was one of the principal features of the technical assistance in the view of the project unit. Nevertheless, the geographic coverage of the project may have been limited because of the unit's policy of providing the same high level of technical assistance throughout the life of each subproject. Even in the later stages of project implementation, the project unit continued to see the project as a demonstration for future cattle development. The following table gives an idea of the number of staff, including supervisory personnel, and their turnover during project implementation. No. of Staff No. of Field Withdrawing No. of Ranch Agents on No. of from Development Year Staff New Staff Project Unit Plans Made 1/ 1968 6 6 - 16 1969 15 12 3 33 1970 20 6 1 55 1971 26 8 2 194 1972 25 3 4 206 1973 25 4 4 193 1974 23 5 7 142 1975 21 2 4 110 1976 17 1 5 34 1977 16 3 4 171 1978 15 2 3 180 1979 12 - 3 8 1980 11 - 1 5 Annual Average: 18 4 3 103 1/ Including contractor plans and including subloans subsequently cancelled. - 56 - After the close of the project, field agents were given the opportunity to continue working with INTA as extension agents. BNA 6.06 BNA is the most important financial institution of the country's banking system. It has traditionally accounted for about half of the banking system's lending to the agricultural sector while its share of medium- and long-term credit has been around 80 percent of the total. Not only does BNA have 572 branch offices all over the country but it also has branches in neighboring countries and in the main financial centers in the world. The branch offices work under the supervision of 20 zonal managers who are responsible for proper development and operations of the branch offices. In total, BNA employs about 257,000 people, including about 3,900 specialist staff, mostly at branch offices. Its loan portfolio is in excess of half a billion US dollars equivalent, of which about 45 percent are loans to the agricultural sector. It has total assets of US$1.3 billion equivalent and a net worth of US$100 million equivalent and had shown satisfactory financial performance. 6.07 The line of credit for project beneficiaries was all but lost in the enormity of BNA activities. Due to the modest amounts of funds lent under the project, BNA officials felt the project had little importance in their total portfolio. One of the reasons for delays in reimbursement of CB by the Bank was the time it took BNA to filter out the project account which followed the monthly tabulation of all its transactions. There was virtually no disaggregation of project subloans at the BNA central office according to final use. The annual audit of BNA was a comprehensive exercise with little, if any, analysis done on project accounts. Accounts are kept separately by branches and are centralized at headquarters in Buenos Aires supported by an electronic data processing system. Internal auditing is conducted by inspectors, responsible to the Board of Directors, and external auditing is carried out by CB. Loan recovery performance is good, in general, and excellent for project subloans. 6.08 Because of its importance in lending to the agricultural sector, BNA has always had competing lines of credit to the project line of credit. While these lines, from time to time, offered lower interest rates than project funds, the terms of repayment, with the corresponding grace period, offered by project subloans were never matched; nor were the opportunities for livestock raisers to obtain major technical assistance matched by BNA competing lines. 6.09 The interest rate charged to project beneficiaries was in constant flux throughout the life of project execution, reflecting, in part, the changing currents in the financial structure of the national banking system. In 1973, all deposits of commercial banks were nationalized by Government and shifted to CB, which held a tight rein on interest rates and lending policy. In June 1977, CB, in accordance with the new financial banking law, released previously centralized deposits and withdrew practically all rediscounting facilities. In view of the high and fluctuating rates of inflation experienced in the country, interest rates to project beneficiaries were changed several times. The table below attempts to document the changes - 57 - in interest rates agreed by the Bank and Government, shows the corresponding rate of inflation at the time of the change, and, finally, shows how the interest rate earned was spread across the different components of the banking system* Interest Earned by Interest Participating Institutions Rates Infla- Fund for Date of Paid by tion 1/ Date of Exchange Side Letter Ranchers (WPI) Effectiveness BNA Risks INTA CB (%) July 31, 1967 16 21 Not applied 8.0 3.0 1.5 3.5 June 21, 1968 13 1 Nov. 6, 1968 5.5 3.C 1.5 3.0 September 4, 1969 12 4 April 22, 1969 5.5 2.5 1.5 2.5 -- 10 8 Oct. 2, 1969 5.5 2.5 1.5 0.5 December 8, 1971 14 48 Aug. 1, 1971 5.5 2.5 1.5 4.5 July 13, 1972 19 67 Feb. 1, 1972 8.0 3.0 3.0 5.0 February 9, 1973 6 2/ 42 March 16, 1973 3.0 1.9 1.1 - January 26, 1978 6 3/ 155 July 31, 1978 3.0 1.9 1.1 - 1/ Increase in the wholesale price index during the 12-month period preceding date of Effectiveness or the date of the side letter approving a change in interest rates. 2/ Principal adjusted monthly by the consumer price index or the price of cattle at Linier's market, whichever is the lower for the month. 3/ Principal adjusted monthly by the wholesale price index. 6.10 Interest rates charged under the project were positive in real terms through most of 1970, declining to negative levels thereafter until indexing of principal was introduced in 1973. The table below shows cash flows as well as nominal and real rates of interest for subloans granted each year between 1970 and 1978, under the assumption of repayment according to a nine-year amortization period, including four years' grace. Sub-borrowers' Debt Servicell in Nominal and Real Terms, and Nominal and Real Effective Rates of interest, for Subloans of $a 1,000, Granted 1970-1978 Subloans Granted 1970 1971 1972 1973 1974 1975 1976 1977 1978 Deflator2l (Year)Z/ Nom. Real Nom. Real Nom. Real Nom. Real Nom. Real Nom. Real Nom. Real Nom. Real Nom. Real 1970=100 Year of Payment 1970 - - - - - - - - - - - - - - - - - - 100.0 1971 100 72 - - - - - - - - - - - - - - - - 139.5 1972 100 41 100 57 - - - - - - - - - - - - - - 246.9 1973 190 51 190 72 190 127 - - - - - - - - - - - - 370.5 1974 390 88 190 60 190 105 190 158 - - - - - - - - - - 444.7 1975 352 27 390 42 190 36 190 54 68 23 - - - - - - - - 1,300.7 1976 347 4 396 7 445 14 245 12 352 20 223 37 - - - - - - 7,791.4 1977 326 2 389 3 452 6 515 10 1,078 25 683 46 108 43 - - - - 19,436.0 1978 263 1 326 1 389 2 452 4 9,018 84 1,319 36 208 34 88 36 - - 47,742.1 1979 - - 396 0 592 1 788 2 22,414 84 14,892 162 543 35 230 37 111 44 119,355.0 1980 - - - - 396 0 592 1 42,659 79 28,409 155 4,703 154 459 37 221 44 238,710.0 1981 - - - - - - 360 0 52,636 65 35,144 128 5,832 127 2,588 140 287 38 358,066.0 1982 - - - - - - - - 64,761 54 43,365 105 7,214 105 3,209 116 1,619 144 537,098.0 1983 - - - - - - - - - - 53,354 86 8,902 86 3,969 96 2,007 119 805,647.0 1984 - - - - - - - - - - - - 10,952 71 4,898 79 2,483 98 1,208,471.0 1985 - - - - - - - - - - - - - - 6,026 65 3,064 81 1,812,707.0 1986 - - - - - ,7 66 2,719,000.0 Total: 2 068 286 2 377 42 2 844 291 3 332 241 192 986 434 1776389 755 35 06 34 Effective Rate of Interest: 16 (31) 20 (39) 25 (41) 29 (48) 146 (14) 161 (5) 91 (8) 74 (9) 60 (9) 1/ For 1971-79, actual interest rates and adjustment indices were used. Repayment terms are not actual, but according to subloan agreements (nine years, including four years of grace), i.e., anticipated repayments are not cnasidered. Assumptions used for 1980-86: 1980: unindexed interest rate 98% (=to 1979), adjustment index rises 100%; 1981: unindexed rate 80%, adjustment index rises 30%; 1982-86: adjustment index rises 30% per annum. 2/ Mid-year (June) loan assumed. 1970-73: loans at flat nominal interest rate, subject to changes by BNA. 1974-78: 6% interest on balances adjusted by cost of living or beef price index. 1979 was not included since practically no new loans were granted under regulations effective from July 1978, implying adjustment by the index of wholesale prices. 3/ Index of wholesale prices, all items. 1970-79 actual. Assumed increases over previous year: 1980: 100%; 1981-86: 50%. January 16, 1981 - 59 - The year 1969 was not included due to the low level of disbursements in that year, which were limited to the last two quarters. For those subloans granted during the early years of the project on an unindexed basis, most if not all of the repayment. period is over. Although nominal interest rates on these subloans were raised to as much as 98 percent per annum in 1978, real interest rates remained negative. After indexing was introduced in 1973, the real rate of interest increased substantially, reaching nevertheless levels in the order of only minus 5 percent per annum. The only exception to this rule pertains to loans granted in 1974 which benefited from a large discrepancy between the increases in the prices for beef cattle on the one hand and consumer goods and services on the other, which were alternatively applied for adjustment cof principal. However, since this meant a. very substan- tial decline of cattle prices in real terms in 1975 and 1976, the ranchers did not benefit much from the implicit credit subsidy. Furthermore, in practice, over 100 indexed subloarns were prematurely repaid by borrowers unfamiliar with and fearful of the effects of indexing, thus decreasing the subsidy implied in this index mechanism. An estimate of the average real rate of interest for subloans under the project yields a figure in the order of minus 20 percent per annum. Drainage and Soil Studies 6.11 The purpose of the drainage and soil studies was to obtain physical resource information for improving pasture and crop management in the project areas, which is susceptible to flooding. The area studies covered about 10 million ha. The purposes of this project component were to: (a) carry out a reconnaissance study in the project area for the purposes of identifying the principal zones with drainage problems and of assigning priority to undertake semi-detailed drainage studies; and (b) carry out soil studies for the purpose of preparing maps of land classification, land use and land capability. 6.12 The terms of reference of the drainage and soil studies changed over the course of project implementation, but these changes generally were within the guidelines of the original plan and were made on a highly professional level. A major change in drainage design was to concentrate on micro-regional, small-scale drainage and soil conservation works instead of the major canal system that was originally envisaged by the Department of Soils in INTA. Of the 10 million ha, about half of the area was mapped at a scale of 1 to 50,000 and half at a scale of 1 to 100,000; about 300,000 ha are mapped at a scale of 1 to 30,000. 6.13 This information should permit various actions by Government to improve agriculture in the project area; specifically the studies could be used for the following purposes: (a) correction of flooding and inadequate drainage without causing dislocations to conservation of water in dams, lakes, catchment areas, streams and water ducts; - 60 - (b) maintenance and improvement of productivity of the soil according to crops, pasture, forest, and such; (c) determination of the tolerance levels against erosion; (d) adoption of farming and marketing systems compatible with soil conditions; and (e) rational land use planning and development, e.g., sub- division of land, colonization, industrial parks, urbanization, production credits, and land taxing system. VII. SPECIAL ISSUES Government Policies 7.01 Government policies toward the agricultural sector and, particularly, toward the beef subsector have fluctuated over the implementation period of the project. These policies, in addition to external market forces and an array of internal problems, resulted in an uneven implementation of the project. The project began at the bottom of the international cattle cycle. The impact of low international beef prices was further exaggerated by the Ongania Government's policy (1965-69) of squeezing agriculture by way of high export taxes of beef and a protected agricultural inputs supply industry. As beef prices on the world market began an explosive run from 1970 to 1973 of about 75 percent in 1979 prices, the Government set a reference price for prices paid to livestock producers on the Argentine market. This occurred in spite of the fact that Argentina does not export more than 30 percent of its beef output per annum. The Lanusse Government (1970-73) maintained high taxes on beef exports but permitted the practice of negative interest rates in the face of growing domestic inflation. This was a growth period for domestic producers and was reflected in the interest among cattle producers to begin investments on their ranches under the project. 7.02 In 1973, the Peron Government assumed control and immediately began anassaulton the agricultural sector as it carried out a development strategy similar to that adopted during the 1946-55 period. Despite the fact that beef prices were yet to peak on the world market, Government launched a series of disastrous policies toward the beef subsector. First, it put a price ceiling on beef to consumers; second, it adopted a differential exchange rate system which was more onerous for beef exports than for most other goods; third, it declared that certain weeks of the year no beef could be sold to consumers; and fourth, it nationalized some of the slaughterhouses which export beef. Those policies had an immediate adverse impact on beef producers. Ranchers began selling off stock, land values dropped, and a climate of uncertainty prevailed, at best, and one of fear of expropriation, at worse. 7.03 In March 1976, the Videla Government took political power on a platform of strengthening the balance of payments, restructuring the external debt and preventing hyper-inflation. Measures such as the lowering of export - 61 - taxes and raising of interest rates had an immediate effect on the balance of payments. Livestock producers responded strongly in 1977 by beginning capital replacement and increasing herds. But by mid-1978, interest rates became indexed to the wholesale price index while there appeared to be an apparent increasing overvaluation of the peso. These two factors dampened the demand for agricultural investment credit and created an attitude of "wait and see" for many livestock and crop producers. Replication 7.04 Thirteen years of experience under the project have shown that the technological package has generally been successful in terms of improved herd management, higher stocking rates, and better animal health. At full development, the project produced about 130,000 head of cattle per year compared to a national beef production of 3 million head. The national cattle stock in Argentina is about 60 million head and in Buenos Aires province, about 20 million. Total national stock has increased from 49 million in 1970/71 to 58 million in 1978/79, but there is no indication that this increase has significantly been influenced by the project. Beef cattle sales at the Linier's market, the most important in the country, actually declined over this period from 6.5 million head to 6.3 million head. 7.05 The use of phosphate fertilizers for pasture and crop development shows no significant trend on the national level. The apparent consumption of phosphate nutrients has fluctuated from 21,000 m tons in 1972 to 4,000 m tons in 1975 and 28,000 m tons in 1979. Because almost all phosphate ferti- lizers are imported and no subsidies exist on the price to ranchers, they, by and large, use only fertilizers when price expectations are strong. Finally, the project introduced baling and storage of hay to the project area on a commercial scale and this practice is now in use in the pampas. VIII. BANK PERFORMANCE 8.01 Generally, Bank performance during the preparation and disbursement stages of the project was satisfactory. The project unit and BNA staff focused on positive as well as negative features of Bank staff performance. Undoubtedly, the identification and preparation of the project depended heavily on the technical expertise of Bank staff and FAO staff to develop ranch management, herd improvement and pasture establishment in the project area. The production techniques applied at the Balcarce Research Station were those already in widespread use in New Zealand and Australia. Expatriate technicians were largely responsible for the adaptation of this package of transferred technology. Furthermore, the first project director was from New Zealand and be continued development of the demonstration ranch in Balcarce. Because of the quality of Argentine livestock experts, he was the only expatriate to work under the project. - 62 - 8.02 Prior to the project, INTA generally worked on a commodity basis both in research and extension. As a result of the project, this system changed and research efforts were based on consideration of the entire ranch as a production unit. This same approach was used in extension goals and in the preparation of ranch investment plans where both technical and financial aspects were considered. This method of on-ranch development was innovative for INTA and it became an integral part of its working system, not only in the project area, but also on a national level. 8.03 Within the scope of the project itself, the presence of Bank staff was a necessary condition, under the circumstances in Argentina, for coordination among the various national institutions involved in the implementation of the project. Unfortunately, virtually the only time INTA, BNA, CB and MINECON sat around the table to discuss project problems and possible solutions to those problems was during Bank supervision missions. The lack of close coordination among these institutions was one of the factors causing delay in project execution. Also the Bank presence in the project was the major factor in continuing counterpart financing to flow from Government to INTA, especially during the difficult period for the country in 1973-76. 8.04 While the project unit did a satisfactory job of undertaking a sample survey of project beneficiaries in 1978/79 as a basic for post-evaluation of the project, that unit complains that the Bank did not give adequate guidance in setting up an evaluation and monitoring system so that corrections could be made in project implementation and better information could have been provided for the post-evaluation. Also, the project unit states that many of the supervision missions were repetitious and time-consuming in that there were a variety of Bank staff who supervised the project from time to time and these individuals often sought to interview the same local people such as the rancher beneficiaries, branch officers of BNA, project agents and those carrying out local cattle auctions. Finally, the project unit found the lack of Spanish on the part of Bank staff to be a major drawback in communicating their problems to the Bank. ARGENTINA BALCARCE LIVESTOCK PROJECT ENTRE RIOS SANTA FE U R U GUA Y COR0OO8A 0BU ENOS AIRES - -- Province of Buenos Aires •••••99 Balarce Liv3stock Project Hills of Tandilia and Ventania1 Sand L,n 2 Lakes, Ls.o: n:ii9Iå" - d Swamp- ånj ilir:r-g- River" in rtrvs---: 1 Water:r.j i . I 'cfo' Drainjg Rainf li a- ---- /00,, rr 8Oo 10åcO 0 AT L LANDS WELL DRAINED FREQUENTLY SUBMERGED (Problematical SOUTH AMERICA lands where investigation on soils and Plains drainage has been recommended) Sierras and Depressions Piedmontes River basins Wet plains MCProjet 0 o0 100 MARCH 1966 IBRD-1715