Honduras - Livestock Development Project

Document of The World Bank ET R - 1 RPOR. D,:77K FOR OFFICIAL USE ONLY ONE v Report No. 19 20 RLE COPY PROJECT PERFORMANCE AUDIT REPORT HONDURAS FIRST LIVESTOCK DEVELOPMENT PROJECT (CREDIT 179-HO) February 21, 1978 Operations Evaluation Department 'IThis 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.  FOR OFFICIAL USE ONLY TABLE OF CONTENTS Page Preface Basic Data Sheet Highlights PROJECT PERFORMANCE AUDIT MEMORANDUM Introduction 1 Choice of Technology 2 Type of Rancher 4 Other Divergences :rrom the Appraisal Design 5 Demand for Credit 7 Relation to Agrarian Reform 8 Conclusions 11 Note on the Rate ol Return 13 PROJECT COMPLETION REPORT A. Introduction A l B. Project Objectives and Results Objectives A 1 Physical Results A 2 On-Farm Investment and Lending Program A 5 Financing and Disbursements A 7 Organization and Management A 9 C. Conclusions A10 Map This document has a restricted distribution and may be used by recipients only in the Performance of their official duties. lti contents may not otherwise be disclosed without World Bank authorization.  Project Performance Audit Report HONDURAS FIRST LIVESTOCK DEVELOPMENT PROJECT (Credit 179-HO) PREFACE Credit 179-HO, signed in March 1970 for US$2.6 million, was closed, fully disbursed, in December 1975. Its principal objectives were to raise the country's low level of beef consumption and support diversi- fication of agriculture, especially in production for export. It was followed by Credit 434-H0, signed in October 1973 for US$6.6 million, which helps finance similar activities under the Second Livestock Develop- ment Project. The livestock sub-loans of both these projects have mostly been directed toward medLum and large scale ranches and dairy farms. More recently, Credit 628-HO, signed in July 1976 for US$14.0 million for the Agricultural Credit Project, provides funds for on-lending to both Live- stock and other agricultural activities. The target population of the third project is different from that of the earlier two projects. This Project Performance Audit Report (PPAR) includes the proj- ect completion report (PCR) on Credit 179 issued by the Latin America and the Caribbean Regional Office in September 1975. The PCR concentrated on matters of implementation and disbursement, and, to the extent data then permitted, physical evidence of project impact. A number of subjects which are commonly addressed in PCRs of more recent date, for example on the preparation and appraisal period, Borrower and Bank performance, and the adjustments made in the repeater projects, were not discussed. Nor did the PCR update the rate of return estimates of appraisal, an exercise which would have required special field surveys. OED has added a brief note on the rate of return, but otherwise has not attempted to make up for these particular deficiencies; it lets the PCR cover to the extent it does the standard matter of project performance. A closer examination of some subjects will be appropriate at completion of the second or third project, for both of which more field data is expected to be available. OED elected at this time to focus its analysis on other aspects of the Honduran experience which are almost never discussed in depth in supervision reports, PCRs and other Bank evaluative documents but which are important to illuminate the way large-farm livestock projects work and whom they affect. The audit memorandum which precedes the PCR is a summary adapta- tion of the discussion of the special issues presented in a working paper prepared by a consultant and available in OED ("Case Study of a Livestock Development Program - Honduras"). The memorandum in effect bypasses the omissions of the PCR just mentioned and tries to answer the basic questions: who are these ranchers; how much did they need the credit; did they use it to finance technical change; how was it affected by the agrarian reform move- ment? What the memorandum concludes is that the world in which the daily drama of the livestock project was played out differed substantially from the expectations of the appraisal reports. To what extent the findings of the Honduran study apply to other countries has not been determined, though some of the trends discussed below have been detected in other large farmer livestock projects reviewed by OED. The valuable assistance provided by the Government of Honduras, and by its project staff in particular, during the audit mission in August/September 1976 is gratefully acknowledged. BASIC DATA SHEET HONDURAS: FIRST LIVESTOCK DEVELOPMENT PROJECT (CREDIT 179-HO) A. AmncuIts (US$ m1n) Original Exchange Current Dis- Principal Adjustment Principal bursed Outstanding Credit 179-HO 2.6 0.2 2.8 2.8 2.& - B. Project Data Original Plan Revisions Actual First Mention in Bank Files 21 APR 66 Government Application 06 AUG 68 Appraisal Mission FEB 69 Negotiations NOV 69 Board Approval 06 JAN 70 Loan Agreement 02 MAR 70 Loan Effectiveness 01 SEP 70 15 OCT 70 05 OCT 70 Percentage of Original Proj- ,!ct Actually Completed 100% Loan Closing 31 DEC 75 31 DEC 75 Total Costs US$5.2 mln US$4.8 m1n Financial Rate of Return 16-21% n.a. Economic Rate of Return 18% 13% 21 C. Mission Data Month, No. of No. of Date of Year Persons Weeks Manweeks Report Preparation MAR 67 1 1 1 13 APR 67 Appraisal FEB 69 4 4 16 16 DEC 69 Subtotal 17 Supervision i NOV 70 1 1 1 30 NOV 70 Supervision II MAY 71 1 1 1 18 JUN 71 Supervision III NOV 71 1 1 1 30 NOV 71 Supervision IV JUL 72 1 1 1 18 AUG 72 Supervision V /c MAR 73 2 1 2 03 APR 73 Supervision VI DEC 73 2 1 2 10 JAN 74 Supervision VII d DEC 74 1 1 1 22 JAN 75 Subtotal 9 Completion /d AUG 75 1 1 1 30 SEP 75 D. FoLlow-on Projects: Credit 434-HO, Second Livestock Development Project, for US$6.6 million, signed 29 OCT 73 and effective 18 JAN 74; and Credit 628-Ho, for US$14.0 million, signed 2 JUL 76. /a First maturity is due in 1980. T_ gee para. 35 of the Memorandum. Ic Together with three other members, this mission also appraised Credit 434-HO. /d These missions also supervised Credit 434-HO.  Project Performance Audit Report HONDURAS FIRST LIVESTOCK DEVELOPMENT PROJECT (Credit 179-HO) HIGHLIGHTS The first project, supported by a credit of US$2.6 million, was intended to create a mechanism for providing supervised long-term credit to Honduras' medium and large-scale ranchers for pasture and herd develop- ment. The 78 ranchers involved in the four-year commitment period (1970- 1973) was only 58% of the number of ranchers originally expected to be financed; but the area and number of cattle affected by the project were greater than planned and the impact on production and productivity was positive, although significantly less than intended. The institutional apparatus, which combines a new technical/promotional office in the Central Bank and private and public participating banks, has been successfully established. The project turned out to differ in important respects from expec- tations. Interviews with twenty-three participating ranchers (including three from the second project) revealed substantial divergences in rancher and banker behavior from the appraisal descriptions. At least half of the ranchers were absentee, were not motivated to accept the intensive, special- Lzed practices described in the ranch plans drawn up at the time of their applications, and had other sources of credit to finance many of the same investments - on shorter terms but without the paperwork and planning. Also, it seems that the lending program was used by some ranchers, and was seen by some agarian reform supporters, as a safeguard against land invasion and/ or expropriation that these ranchers feared from the accelerating reform activities. This project audit indicates that the real world in which credit and technical decisions were made by Honduran ranchers was vastly more complex and less development oriented than implied by the appraisal report. The following points may be of particular interest: Pasture and herd investments and management practices tended to resist the new technical orientation (PPAM, paras. 3 and 4) Owners were mostly absentee; many had urban investments that competed for their attention (PPAM, paras. 8-11) Structure of livestock sector was inhospitable to appraisal's design and distorted the impact of the ranch plans (PPAM, paras. 12-15) Many participating ranchers had alternate sources of credit; appraisal described ranchers who had none (PPAM, paras. 19-21) Some competition existed between the livestock project and both agrarian reform and intensive farming (PPAM, paras. 22-26, PCR para. 16) Supervision reporting did not address all relevant issues (PPAM, para. 16) Project Performance Audit Memorandum HONDURAS FIRST LIVESTOCK DEVELOPMENT PROJECT (Credit 179-HO) Introduction 1/ 1. The pur Tse of the special audit mission- was to use the Honduras livestock project= as a case study for investigation of four questions per- tinent to many of the Bank's livestock projects: (1) To what extent did the project result in the adoption of productivity-increasing changes in live- stock production methods by sub-borrowers? (2) Who were the project ranchers - where did they live, did they run their own ranches, did they have signi- ficant sources of income outside ranching? (3) To what extent did the credit made available to sub-borrowers represent a real addition to the credit and other capital to which they already had access - and to what extent did project credit substitute for credit and capital previously being used by sub-borrowers in their livestock operations? (4) Did the Honduran agrarian reform process have an impact on the decisions of ranchers to participate in the project, and in their choices of production techniques, and did the project in turn have an impact on agrarian reform? The main finding of the mission was that there were consistent and large divergences of sub-borrowers and their ranches from what the Bank had expected. The conclusions cannot be readily generalized to other countries, though it is clear that the conditions and trends were not peculiar to Honduras. 1/ Four weeks were spent in Honduras in August and September of 1976, three of which were taken up visiting project ranches. A week was spent in Tegucigalpa, where the project office is located, interviewing project staff, current and past government officials who had been involved with the project, and participating-bank management and technicians. Ranchers were visited with different project or participating-bank technicians - though interviews were usually not conducted in their presence. Twenty- three project ranchers were interviewed. Nineteen of the interviews took place at the ranch and all but three of the ranchers were from the first livestock project. Complete data could not be obtained on every case, so some of the comments below are made on the basis of only 20 (para. 3) or 17 (para. 20) project ranchers. Six small and medium non- project ranchers were also interviewed in the department of Atlantida. 2/ Though this evaluation covers the first livestock project most compre- hensively, it touches considerably on the second project (still dis- bursing) and the designing of the third (approved in June 1976). There was no hiatus between commitments under the first project and the second, and the project staff looked at sub-borrowing under the second project as a continuum of the first (a clause allowing for retroactive financing was introduced in the second project to make this possible). Thus some of the problems and issues that evolved only partially during the first project were better understood and worked on during the second and the elaboration of the third. The second project, moreover, added 179 sub- borrowers.to the 78 of the first project. With this total of 257 cases, it was possible to find patterns that would not be significant in the smaller group of 78, or that woild not show up at all. - 2 - 2. The four issues raised do not cover the whole of the project experience. They concern the sub-borrowers: the ranchers' decisions and activities. Another set of issues concern the lenders: in particular the creation and growth of the executing office in the Central Bank - the Proyecto Ganadero(PG) - and its capability to direct an increasing flow of long term livestock credit through the intermediary financial channels. That story of institution building, mostly a story of success, has not been examined in the audit memorandum (pieces of it are presented in the PCR) since earlier PPARs on Latin American livestock credit programs have explored the institutional dimension in depthl/; the Honduran exercise provided an opportunity to turn attention to the rancher. Choice of Technology 3. One group of divergences related to matters of technology, showing that production methods were less intensive than the methods of the appraisal design. First, with respect to pasture, stocking rates were considerably lower than expected - averaging 1.1 animals per hectare in the fifth year of project development instead of the projected two animals per hectare. Because of the casual and sometimes inconsistent classification of lands in pasture or coming into pasture, it is impossible to get accurate measures of actual carry- ing capacities and, therefore, to determine whether the lower ex post stocking rates reflected lower capacities or underutilization. There are suggestions, however, that the degree and type of pasture improvement financed was less intensive than anticipated, and that the carrying.capacities projected at ap- praisal were not achieved. The PCR data (PCR page A6) shows total hectarage of actual pasture improvement greater than the appraisal forecast (26,311 ha vs. 25,550 ha). It also shows a relatively greater increase in the area clas- sified as "new pasture" than in the area classified as improvement of "existing pastures". The first class refers to the substitution of new pasture varieties for traditional grasses and other natural grazing land; the second class refers to the renovation and regeneration of existing species. Investments in the first class are more costly than the second. Thus the PCR data would indicate that there was more intensification than expected. But the interviews throw a different light on the process. There it becomes clear that the average level of development of the pastures in the pre-project period was lower than implied in the aDpraisal design, and that the degree of improvements under the project - whetherof the first or second class - also fell short of design (though probably not of PG expectations). Almost one third of a sample of 20 project borrowers said they used their subproject funds to develop their ranches from a state of almost complete abandonment. The word does not imply that the land was virgin property being grazed for the first time, though that would sometimes have been the case. What it does imply is that much of the land brought under the influence of the project could only be called pasture in the most primitive of natural conditions, and improvement implies burning of bush and other clearing activities as well as establishment of some useful 1/ See Ecuador-First and Second Livestock Development Projects, Report No. 892 Uruguay-Third and Fourth Livestock Development Projects, Report No. 1321 Colombia-Second Livestock Development Project, Report No. 1344 Mexico-Third Livestock and Agriculture Development Project, Report No. 1573 - 3 - grasses, rather than the substantial upgrading of recognizable pastures. This is no surprise: there was hardly any improved pasture in Honduras to begin with. Project investments clearly resulted in an improvement, but the starting and end points were lower than appraisal had suggested and the -degree of improvement was less than anticipated. Sub-borrowers simply had less ambitious motives for taking credit. 4. Second, with respect to animal management, ranchers often reported that they sold their steers before they reached market weights, that they often sold reproducing stock to generate operating capital, that they milked their beef cattle and intended to continue to do so,1/ that they raised their dairy calves to maturity instead of selling them at weaning, and that they did not adopt seasonal breeding, or discontinued it. In each respect the rancher's action departed significantly from the management methods implied in the first appraisal report and brought out more explicitly in the second. In addition, expenditures on animal health seemed to be much lower than recommended levels. A sample of expenditures on salt and minerals showed that two thirds of the sampled borrowers were spending an average of 32% of the recommended levels on these items. Similarly, expenditures on veterinary products average only 36% of recommended levels for over half of the sampled ranchers. 5. Third, a tendency has developed for project funds to be shifted to the purchase of reproducing animals. With 60% of the second project credit funds committed, breeding-stock purchases have accounted for almost twice their expected share of ranch-development costs (i.e. excluding slaughterhouse credits) - 51% vs. 26%. An arithmetic error, behind the table on page A6 of the original PCR but corrected in this PPAR, implied a 70% increase over appraisal in the number of breeding stock purchased under the first project as well. The corrected figures now show a slight decrease (6,781 vs. 6,805), but there is reason2/to believe the actual figures may understate the number of breeding stock ultimately purchased, and that the tendency to shift to animal purchase began during the period of the first project. Moreover, one of the top project officers asserted that the shift was a fact, and partly explained it by the fall in the price of heifers. I/ The appraisal models show milking continuing on non-dairy farms only up through the fifth year. Ranchers interviewed in year 4 and 5 had no intention of stopping milking. Ranchers interviewed in year 6 had not stopped and had no intention to. The relative improvement of milk prices vis-a-vis beef prices could explain the persistence of milking activities, but it was clear from the interviews that the behavior mainly reflected the influence of cultural rather than price phenomena. 2/ "Actual"figures often are aggregated from loan applications and ranch plans, and would understate the true numbers purchased if ranchers sub- sequently responded to a decline in heifer prices by buying more animals than they had proposed to buy. Changes of this sort were not secret, but could be missed in the aggregation of data for report requirements. More- over it is known that the PG had to accede to rancher requests to include more cattle in order to make the loan. 6. All these findings suggest that increases in the beef and dairy production of project ranches were less a result of the animal productivity improvements expected to result from the project than from extension of a modest pasture and paddock establishment package to primitive grazing lands and of the expansion of the herd through purchase. The Bank loan, meant to give ranchers the financial backing necessary to give up the old husbandry practices and wait until the new practices yielded their return, did not play this facilitating role to the extent hoped for. Type of Rancher 7. Another group of divergences from Bank expectations was related to the size of the subloans and the type of borrower. Ranch size and loan size-turned out to be significantly larger than expected under both first and second projects. Under the first project, for example, average cost of investments per ranch was 79% greater than expected. About half of the ranches were larger than 500 hectares, while the ranches of the principal models in the appraisal report were 150 and 400 hectares. The average ranch size has fallen in the second project - the PG reports only 22% are more than 500 hectares - but it is still somewhat above original expectations. Though the PCR attributes the greater average costs of ranch development to the larger-than-expected ranches, it turns out that loan size was markedly larger than anticipated even for ranches that were roughly the size of the models (a discrepancy that cannot all be explained by inflation). For dairy/ fattening ranches in the range of the model size (150 hectares), average investment costs were 130% greater than the model. For beef ranches close to the model size (400 hectares), costs were 27% greater. 8. Also in contrast to the Bank's assumptions at appraisal,.- a large majority of project ranchers (73%) did not live on the ranch; half of them spent less than half-time on the ranch; and about half had significant in- comes outside ranching. A large minority had urban investments and/or full- time urban employment, and almost half of the borrowers lived in the country's three principal cities - Tegucigalpa, San Pedro Sula, or La Ceiba. 9. These divergences from the appraisal design can be explained in various ways. The more extensive production methods frequently went along with the absentee ranching. Yet most of the productivity-increasing prac- tices recommended by the project were management-intensive. It would be difficult for absent owners to supervise them. At the same time, the com- petent, well-paid ranch managers projected at appraisal were not common on project ranches. A sample of operating costs for project ranches showed foremen receiving only about 60% of the annual salaries stipulated in the design. 10. For sub-borrowers whose urban investment activities took consid- erable time, the opportunity cost of managing the ranch more closely was high. In addition, the ranch management decisions of these other-activity sub-borrowers seemed to be guided by expectations and time horizons that 1/ The characteristics of the participating rancher were not described in the appraisal report, or in the lending agreements. Nevertheless, the preparation report and Bank files make it clear that residency was thought to be important and, at least up to appraisal, had been included as a condition of on-lending. - 5 - were quite different from the Bank's conception of a project rancher almost completely dependent on his ranch for income. 11. To a certain extent, the divergence from the appraisal design was also caused by the focus of the project on the largest ranchers. Though appraisal and other documents justified this choice on the grounds that the largest were the most efficient, the evaluation suggests that the largest were actually the most extensive and least technically oriented of the project ranches. In addition, the proportion of total property on project ranches in improved pasture decreases consistently as ranch size increases. The same phenomenon was found for the value of investment per hectare of land in pasture on project ranches - i.e., a consistent decrease as ranch size increases. Project technicians reported as well that the larger ranchers were the most extensive. Since larger ranches were more characteristic of the marginal zones, one could have anticipated an inverse correlation between ranch size and intensive practices. However, even within the same zones the inverse correlation persists. Clearly, this outcome was heightened by the fact that project ranches turned out to be even larger than the Bank expected. Other Divergences from the Appraisal Design 12. Other reasons for the divergence from the appraisal report are directly attributable to the design of the ranch development plan. Mainly, the original plans were designed to accommodate fairly intensive investment and low or negative returns in the early years, followed by substantial returns in the later years. The income shortfall in the early years was to be made up with short-term credit for the purchase of feeder steers for fattening, the sale of which would provide operating cash. For various reasons, however, there is not much economic incentive to fatten steers in Honduras. The practice of purchasing feeder steers has never been wide- spread and did not Increase much even with the encouragement of the project. As a result, the project's operating-capital credit was barely used. Only 27% of the borrowers availed themselves of this credit; only 40% of the short-term credit projected at appraisal was actually committed. Of this amount, 46% was never disbursed. Thus the value of operating capital credit actually used was 21% of that expected at appraisal. 13. The shortfall in the use of short-term credit for feeder steers had serious implications for project ranches. In the appraisal design, the sale of feeder steers was projected to account for roughly one half of the operating income of project ranchers. If feeder steer sales were not a feasible alternative for many rancher-applicants, as noted above, then they would have to obtain their operating cash from other activities. The results were that: (i) borrowers with sources of income outside livestock were often the only applicants who were able to withstand the lean early years of investment - as was pointed out by the project office in recommending several applicants for financing. This type of borrower, however, was not necessarily the most likely to undertake intensive manage- ment techniques; and (ii) borrowers obtained operating cash by engaging in just those practices that the project was supposed to eliminate - mainly the milking of beef cattle, the selling of steers before they reached market weight, and the selling of reproducing stock. Thus,though the Bank's ap- praisal design was intended for a rancher who would devote full time to - 6 - the ranch and live mainly off its income, its financial rigor and its fail- ure to incorporate certain features of the Honduran livestock sector often resulted in the encouragement of just the opposite type of behavior. 14. Implicit in the livestock project was the assumption that special- ized beef and dairy production were more profitable than the dual purpose system that prevailed in Honduras. Thus the project recommended practices that required a shift to the specialized system - e.g., the cessation of milking of beef cattle, the sale of calves on dairy farms at weaning (this feature was emphasized more in the second and third projects than in the first), the transition from year-round to seasonal breeding. Similarly, the project assumed that intensive management was more profitable than extensive management. Yet project ranchers consistently demonstrated a preference for diversifying rather than specializing their activities - combining beef and dairy, livestock and agriculture, livestock and urban investments. 15. The persistence of extensive, unspecialized production methods suggests that they may have been more profitable in the Honduran context than the production mode promoted by the Bank project. Prices paid by Honduran packers are the lowest in Central America - 32% less than the average for the other beef-exporting countries of the region. Corres- pondingly, Honduran slaughter weights are also the lowest - 29% less than the average for the other countries. No price premium is paid for quality, and premiums for weight increments are said to be insufficient. The difference in price paid per pound for the increment from a 700-800/lb. steer to an 800-900 lb. steer has averaged only 3.2% over the last several years, and was actually somewhat higher in the second half of the 1960s. The monopsony power of the meat-packing industry seems to have contributed to this situation. Although ranchers can often obtain better prices by exporting their animals on the hoof to Guatemala and Nicaragua, this prac- tice was made illegal some years ago as a result of pressures on the government by the owners and trade unions of the seven export meat packers. 16.* Another contributing cause of the shift from appraisal expectations was the pressure exerted on the project office to keep up the projected rate of sub-loan commitments. Supervision reports and other project implementa- tion documents concerned themselves mainly with the rate of commitment of project funds, apparently without recognizing that urgency gave priority to selection criteria which might promote subprojects and subborrowers quite divergent from the original plan. The participating banks also turned out to be reluctant to lend to borrowers without urban real estate and non-ranch income - a reluctance that was caused in part by the agrarian reform (see para. 22.). 17. This discussion refers only to the divergence between what ranchers did and what the set of appraisal assumptions implied they would do. It - 7 - does not imply a divergence between what ranchers did and what, from their view or that of the PG, made sense for them to do in the years of project implementation, when operational decisions had to be made. Some shift toward specialization has been reported, and the PG properly takes credit for having promoted those first steps. The PG notes, and OED agrees, that the series of dry spells that were recorded since 1972, the rise in the ratio of milk prices to beef prices starting in early 1974, the fact that the selection of participating ranches was ultimately determined by the participating banks and not the PG, and other extenuating circumstances all make the behavior seem reasonable, and the slow but measurable progress toward apprai- sal objectives more impressive. The PG also notes that the processes of technification and specialization have both been started and can be expected to continue, and OED has no reason to dispute that assertion. 18. However, the project actually implemented, though of benefit to Honduras, turned out to be significantly different in terms of its quantity and time achievements from that originally proposed, for reasons which in retrospect reflect misapecification of rancher and banker behavior in the project design. Demand for Credit 19. A considerable amount of investment credit for livestock was already available to ranchers through the banking system, though the apprai- sal report said that "the larger and more efficient ranchers (had) received little (investment) credit". In 1970, the year before project funds started disbursing, new medium-term loans of the banking system for investment in livestock amounted to more than five times the value of project sub-loan commitments in 1972, thae year when commitments were greatest. This invest- ment credit for livestock was granted for three-to-seven-year periods, and accounted for about 64% of total livestock credit in the banking system; the rest was for operating capital. Project credit never amounted to a significant proportion of total investment credit for livestock; annual project commitments in the 1971-74 period ranged from between 5% to 10% of total new loans for livestock investment credit../ What can be said is that project credit extended the average term of the loans, and introduced a whole new credit concept in Honduras, including the preparation of ranch development plans, the financing of the full development of the ranches, and the supervision of the actual investment process. 20. It was also stated at appraisal that the Bank loan would represent the first time that a government-sponsored program in Honduras would make 'Long-term investment credit available to large ranchers on terms and condi- tions appropriate for the development of their ranches. The state development 1/ The figures for total livestock investment credits can be misleading. Some of it would have been diverted to other purposes, and is better labeled credit to people who owned livestock. - 8 - bank (BNF), it was said, provided substantial amounts of livestock credit but restricted its application to small and medium farmers. Without access to BNF credit and without long-term credit facilities at private banks, it was said, large ranchers were being left out. The Bank loan was meant to fill this alleged gap. It turns out, however, that many project ranchers had ample access to BNF credit, both before and during their participation in the Bank project. Out of a sample of 17 ranchers interviewed, 15 or 88% reported having had substantial BNF credits for livestock. Since the BNF accounts for 34% of total livestock investment credit in the banking system, this finding suggests that large ranchers had more than a proportionate share of BNF credit. Again, the point should be emphasized that the terms and packaging of the project credit offered the critical new feature. 21. Nevertheless, with livestock credit so available before the Bank project, many ranchers had to be persuaded to switch from their other lines of credit which they were used to taking in smaller amounts and without the supervision required on Bank projects. This availability of other livestock investment credit, and the resulting reluctance of some ranchers to abandon the old type of credit for the new, had the effect of limiting the number of applications. There were more of these anyway than the project could have handled (215 were presented in the first project period), but the collateral requirements of the participating banks eliminated well over half (78 were approved; some of those rejected were picked up by other banks in subsequent years). Thus the two factors combined to produce a relative scarcity of acceptable applicants. The project office, pressured to keep loan commit- ments moving apace in the face of rancher reluctance, was not in a position to take direct action to increase the tempo of lending, since that depended on decisions of the participating banks. Nevertheless, there was much con- sultation between the PG and the banks, and the PG did not resist the tendency of the bankers to fall back on larger loans and some sub-borrowers who were not especially capable of or interested in introducing productivity improvements.1/ Relation to Agrarian Reform 22. The agrarian reform taking place in Honduras during the implementa- tion of the livestock program had two implications for the program, in addi- tion to the reluctance of participating banks cited above. First, the threat of invasion and expropriation was an important element in the decision of many sub-borrowers - in the months before and after the decisive decree No. 8 in December 1972 - to develop their ranches, or to develop them in the ways that they did. The new law exempted properties that were already being worked, and, particularly pertinent to potential sub-borrowers, properties 1/ The project directors have also pointed out that many smaller ranchers were excluded either because their tenure titles were not clarified or because the BNF already held first mortgages on the property, precluding the use of title to secure a project loan as long as BNF stayed outside the project.(it entered in 1975). - 9 - that had obtained financing to undertake farm development.1/ Pasture was the quickest way to get considerable amounts of land under cultivation (as opposed to crops). This point was made .by various project ranchers and technicians. To the extent that such motivations were important, they help explain why project ranches showed less intensive and specialized production than had been planned. 23. The second implication of the agrarian reform for the project was that the project and the Bank came to be identified, in some Government circles, with the forces of rural inequity in the countryside - as supporting the class that was the reform's most organized and vocal adversary. The project, it was said, was being used by landowners to evade expropriation. To the extent the reform forced the landowners to make a significant and sustained improvement to their property, the effects for Honduras were welcome and the action had the effect of compliance whatever the motive. But the desire to avoid the effects of the law was also present. It shows up in those cases where a minimurt amount of infrastructure was financed, enough to meet the law but not the spirit of the farm plan. It shows up on those occasions when PG officers supported the cattleman's association in asking for better treatment of expropriable cattlemen. The project director, who in the past had been e7ecutive secretary of the association, was cooperating with the committee2= set up by Government to recommend changes to the regulations for the December 1972 legislation. These amendments, accepted in April 1973, substantially broadened the class of exceptions. The project director's role was clearly in the interests of the project, and OED does not consider these activities reprehensible. However, wrong impressions about the Bank's interests in land reform had been created. 24. At the time, of the second appraisal mission in February-March 1973 the direction to be taken by the agrarian movement promoted by the change in government in 1972 had begun to clarify and it was recognized by the mission that a collision between the objectives of reform and rancher community was possible. Within the reform group in government extensive cattle ranching had been defined as a major target of the reform process - particularly in 1/ The exemption that the new law provided to project borrowers had its greatest impact on the second project portfolio. Nevertheless, since 47% of the sub-loans of the first project had been committed before October 1972, and 87% before January 1973, the rapid pick-up in the rate of lending in the last quarter of 1972 must be partly related to the new legislation. Throughout 1971 and 1972 ranchers were reacting to the considerable threat of reform and land invasion that started to accele- rate in 1971. 2/ The members of the committee were the Ministers of Finance and Communi- cation, the President of the Central Bank, the Director of INA, the Secretary of the Cattlemen's Association (FENAGH) and two legal advisers. The committee called on various persons, among them the project director, representing different interests of the agricultural sector. - 10 - the more populous areas of the country where project ranchers were located. The appraisal mission was aware of the reform objectives - it interviewed the heads of the government's agrarian reform institute as well as of cam- pesino groups. In fact, the appraisal report described the reform as "the single most important issue facing Honduras". The design of the project followed the intention of the law and was intended to avoid a conflict with objectives of the land reform by limiting the development of large, extensive ranching operations to marginal zones of the country that attracted little campesino agitation and by intensifying dairy and feeding operations that could meet the spirit of the reform legislation in the more fertile areas of the country. What was missing in the appraisal analysis was, first, an awareness that the first project was falling short of a similar set of objectives, exposing it to the criticism of the reformers, and, second, that any finance for ranchers could be interpreted as Bank support for the cattle- men's position. Subsequent supervision reports mentioned the impact of reform on project ranches, but not the impact of the project on reform. 25. Honduran critics also note that the first two livestock projects promoted livestock development in the most fertile and flat lands of the country, often the lands most suited to intensive agriculture. The land reform legislation did not explicitly call for an end to livestock activi- ties on these fertile lands. The 1972 decree insisted only that such ranching property be committed to pasture. The 1975 decree introduced lower limits on carrying capacities, and these were set high enough to limit livestock enterprise to dairy and other intensive operations. Since that shift has not occurred at the desired rate, the project gives the appearance in some areas of competing with the agricultural thrust of the land reform objec- tives, although it should be noted that the reform and the policy commenced only in late 1972, well after approval of the first livestock loan in 1969. 26. Rural population densities were also higher in the areas of con- centrated project development, as were organized peasant demands for land. There was a strong statistical association between the counties where proj- ect ranches were located and those counties where peasant groups had success- fully claimed land under agrarian reform legislation. 27. Concern about livestock and lands apt for agriculture was not exclusive to Honduras. In the early 1970s, Bank papers on livestock lending stated that livestock projects should not be located in areas suited to intensive agriculture and/or where rural population was concen- trated. The intention of the second appraisal mission - to support extensive type improvements only in the marginal lands - was not reflected in the par- ticipant banks' lending policies, at least until 1975. The ranches of the second project, like the first, came to be located in those areas of Honduras that were most suited for intensive agriculture. - 11 - 28. Clearly the land issues are contentious and subject to interpreta- tions. The preceding views to which the OED mission was exposed in Honduras indicate tension between the objectives of the first and second livestock projects and of land reform in the middle 1970s. The objectives of part of the reform group called for the reduction, not the improvement, of the exten- sive ranching sector in regions where peasant pressure is high and the land is suitable to intensive cropping. Project lending of the sort that was most characteristic in these regions was, from this perspective, unwelcome. The present view of Bank staff and the PG is that the project has neverthe- less played a constructive role in helping some ranchers improve their properties to the point where they contribute to national output. If that improvement process has been slower than expected, perhaps the origi- nal expectations were too ambitious. Recent experience does not undermine the case for this type of credit operation, and, with time, the results will support the objective of efficient land use to which all groups subscribe. Conclusions 29. The first livestock development project was instrumental in increasing beef production on project farms (PCR, page a4). On the whole it must be viewed as a worthwhile effort, even though production, as well as productivity improvements, fell short of appraisal estimates. However, production increases were obtained in a way that diverged in important respects from the way envisaged at appraisal. 30. The answers to the four questions posed at the beginning of this memorandum are highly intertwined. The rate of adoption of productivity- increasing methods of livestock operation was seen to be influenced by who the sub-borrowers were, whether they already had access to livestock credit before the project, and how their investment decisions were affected by the agrarian reform. In general, the divergence from the appraisal design toward more extensive, traditional and unspecialized production was caused by two independent factors: first, the borrower selection process and the type of borrower selected; and second, the fact that the specialized inten- sive design of the project may have been actually less profitable than traditional practices. 31. The incidence of divergence from the appraisal design was high: it would include at least half the borrowers under the first and second livestock projects. For some particular divergences - such as residence off the ranch, low stocking rates, and milking of beef cattle - the diver- gent population would represent about 75% of the sub-borrowers. However, between 25 and 50%,of the project ranches did follow the path the project expected them to and there were many project ranches where significant improvements in productivity were achieved. - 12 - 32. Could the project have reached more of the kinds of ranchers it hoped to reach - namely, the one-quarter to one-half of the project ranchers who were like the kind of producers originally aimed at by the Bank? The tendency of private participating banks to lend to large, prosperous farmers and ranchers that have urban properties and other sources of income, and give low priority to medium and small farmers, has been observed in several Bank- financed projects involving on-lending operations without maximum farm size restrictions. To forestall this tendency in more recent projects the Bank has arranged for improved security loans, chattel mortgaging, state guaran- tee, land registration, collective guarantee, etc.. Ceilings on either farm size or loan size have also been incorporated in the newer loans and credits. As for the many medium-sized ranchers not able or afraid to take loans of the size and amortization period typical of the Bank project,1/ a slower, stag- gered development plan for their ranches, financed by a series of smaller loans, on shorter terms, might have been more acceptable than the still common approach envisaging full-scale ranch development financed out of a single large loan with a long amortization. (The completion of the series under the framework of the comprehensive ranch plan would be agreed to in advance.) Although large sub-loans and long amortization periods have gen- erally been considered a contribution to economies where this type of credit is scarce, the findings of the audit mission suggest that this feature may have been counterproductive to the productivity goals for some groups of potential applicants in Honduras. 33. These observations suggest that two basic project design issues - the kind of livestock operation that is most profitable under prevailing circumstances, and whether in some regions livestock itself makes sense when compared to other agricultural alternatives - were not well enough explored before or during appraisal of the first two Honduran livestock projects. As discussed above, actual ranch developments show consistent divergences from the ranch development models in the project appraisal report. The third project, with its dual purpose model and its inclusion of some credit for agriculture and agrarian reform beneficiaries, shifts the lending program further in the right direction. 1/ This reluctance, noted by the evaluation mission, seems to have existed despite an increase in legal interest ceilings during disbursement of the first project, which temporarily left project interest rates two percentage points below other lending rates (9% vs. 11%). To these ranchers, the small interest advantage of project credit seemed over- shadowed by the risks they perceived in the taking of large amount of credit for long periods. - 13 - 34. The question has been raised whether actual achievements can be legitimately compared with appraisal models, since ranch models cannot be expected to be blueprints, forecasts, or binding on project managers and ranchers. Actual ranch developments rarely match appraisal models: market conditions, prices, techniques and other factors continue to evolve after appraisal; and variations around averages may hinder comparisons. The answer offered here is that if ranch models are to be a credible basis for analyzing the technical, economic and financial merits of projects, they must broadly approximate expected rancher behavior. In this case, actual behavior on at least half the participating farms diverged significantly from the premises of the project - that ranchers who took the loans would be resident on the farms, that they would be intent on technical improve- ment, that they would specialize their enterprises, and that they had no substitute source of finance. In OED's view the models will improve if they can be made to more closely reflect actual behavior; if non-residency is a fact of life the project should be designed to avoid its pernicious effects. 1/ Note on the Rate of Return - 35. The appraisal report on the first livestock project estimated an economic rate of return of 18%. The PCR discusses the meager evidence available on physical impact and shows beef and milk production falling behind the benefit build-up expected in the first five years (60-70% of the appraisal level in year 5). The PCR does not re-estimate the rate of return but concludes that "the benefits projected at appraisal were fully achieved" (PCR para. 19). The technical coefficients have been discussed in qualitative terms by an OED mission during interviews with the ranchers, and their comments suggest that productivity improvements as well as produc- tion have fallen short of appraisal estimates and that the rates of return expected at appraisal cannot be sustained. With that general finding before us, and in view of the PCR data on production levels, we estimate very approx- imately that a rate of return on the Honduran project would fall in the range of 10-15%, giving a best point estimate of 13%. The pattern of ranch investments and management observed in this project is similar to the exten- sive operations described in an OED evaluation of another large farmer live- stock project in Latin America for which new calculations were possible (see the PPAR on the Mexico Third Livestock and Agriculture Development Project, Report No. 1573). The shift from intensive to extensive operations was shown to result there in a fall of the economic rate of return of beef cattle opera- tions from an appraisal range around 25% to a range around 13%. OED suspects that the last estimate is broadly applicable to large ranchers who invested in extensive management operations in Mexico and Central America during the period of rising beef prices in the early 1970s. 1/ The PCR, issued in September 1975, did not include a new estimate of the rate of return. The discussion in para. 35 has been jointly formulated by the Regional and OED staff on the basis of the information available to the Bank, and the assumptions which follow from that information about technical changes and their impact on economic returns.  HONDURAS - FIRST LIVESTOCK DEVELOPMENT PROJECT (Credit 179-HO) 1/ PROJECT COMPLETION REPORT A. Introduction 1. The Project f:inanced partially with Credit 179-HO (US$2.6 million) was the first credit made by the Bank group to Honduras in the agricultural sector. It was also the country's first project in the agriculture-livestock sector to provide long-term private commercial bank loans to ranchers on terms and conditions appropriate for ranch development. The participating banks (PBs) made sizeable contributions to the Project in the form of capital for on-ranch investments and operating capital, plus providing technical and administrative services. The Project aimed at the improvement of the beef cattle industry, with particular emphasis on production for export. It was conceived as a country-wide operation, but with most of the activity con- centrated in the Atlantic Coast, which had the greatestconcentration of suitable ranches and most favorable resources for development. 2. The Borrower was the Republic of Honduras, represented by the Ministry of' Econom and Finance and the program was carried out by the Central Bank (CB) which channeled funds to ranchersthrough six participating commercial banks. In accordance with IDA agreements, CB established a Livestock Project Account to administer and disburse the proceeds of the Credit. Project operations were conducted under the direction and supervision of a Project Commission, estab- lished by the Government, and a Project Director (PD), appointed by the Ccmission, with IDA's approval. Technical staff from the PBs were seconded to work under the Project Director. Sub-loans to ranchers under the Project lending program were made by PBs only after the PD had approved the respective ranch dvelopment plans. B. Project Objectives and Results Objectives 3. The principal objectives of the Project were to assist Honduras in pursuing its national agriculture policy which calls for raising the country's extreme3Vlow l3vel of beef consumption (3.7 kg per capita, 1967) and help it diversify its agriculture, especially in production for export. The Project 1/ Issued September 30, 1975. Corrections on pages a3 and a6 made on August 17, 1977. - A 2 - was not designed to have an impact on income distribution. Project objectives were to be achieved by developing about 135 beef ranches.and dairy farms, which would increase national livestock production by about 15%. Physical Results 4. The Project itself and by virtue of its multiplying effect on ranches not-covered by the program has had a significant impact on livestock production in Honduras, but it is difficult to determine precisely how much credit can go to the project exclusively since other factors have also had an effect on national production increases. Herd expansions, greater milk production, higher per capita beef consumption, and increased exports were also influenced by such factors as changes in purchasing power, price policies, and liberalization of US beef import quotas (most of the beef is exported to the US). The following table shows the increases recorded in the past five years. 1969 1974 Increase $ Cattle population (head, TOO) 1,559 1,690 131 8 Beef production, MT 28,91j48 36, 149 7,201 25 Beef exportation, MT 16,319 21,202 4,883 30 Beef consumption 12,629 14,947 2,318 18 Per capita consumption kg/year 5.20 5.47 0.27 5 Source: Direcci6n General de Estadistica y Censos. Consejo Superior Planificaci6n Econ6mica. 5. Although the Project was designed to finance about 135 ranches, only 781/ were actually developed. The reason for this difference was that the PBs started lending operations by dealing only with their old clients who could offer sufficient collateral for the sub-loans. This meant that only the larger ranchers were being financed. The program was new in Honduras, and PBs bad never previously done any long-term lending for on-ranch investments. Further- more, they considered it too expensive and too risky to lend to small farmers. Besides political and agrarian uncertainties made PBs very cautious in select- ing their clients. 6. Although the number of ranches financed was only about 58% of that expected, the total area and the number of cattle involved under the Project were greater than planned at appraisal, thus meeting Project objectives, as shownin the table below: 1/ 78 ranches but 79 sub-loans. One was a repeater. -A 3 - Situation Before Development Expected at Appraisal Actual Av. Size Total Area Av. Size Total Area Ranches under the Project No. (ha) (ha) No. (ha) ha 1. Dairy steer/fattening 50 150 7,500 45 352 15,8h0 2. Beef breeding/fatteniLng 75 400 30,000 27 1,517 4o,959 3. Beef breeding 10 15,000 6 1,450 8,700 4. Total 135 389 52,500 78 80 65,l.99 5., Area in pasture (ha) 381 51,t50 328 56,508 A survey of 23 ranches that had the most reliable informaticn showed that average yearly beef production per farm was increased 100%, compared with production before the Project, and milk was about 43% higher, as shown in the table below: Surv of 23 ranches Before Actual Appraisal Edtimate 135 ranches the at Time of Before 5 year of Project Survey Increase % Development Development Increase % Beef animals sold (No.) 36 73 37 103 54 152 98 181 Lwt (kg)V 12,600 25,550 12,950 103 18,900 53,200 34,300 181 Milk (its) 82,670 118,370 35,700 43 9,600 15,900 6,300 65 Effective calving rate (%) 57 68 11 19 51 70 19 37 Total Cattle (No.) 329 602 273 83 393 628 235 6o Animal Unite (AU)3/ 246 376 130 53 315 479 164 $2 1/ Average Lxt/head 350 kg. 2/ AU.Cattle over 8 months old. -A5 - Beef pradeotion, which included purchased fattening steers in the ranches sarveyed, was lower than expected at appraisal because a two-year drought had made forage scarce for rattening. The small amount of working capital utilized (para 10) is also reflected in the lower production figures achieved. The increase in the effective calving rate was not as spectacular as was expected at appraisal, but the eatimate was actually too optimistic. Differences in beef and milk production may also be partially due to the difference between the types of ranches actually financed and the type expected at appraisal, as shown below: Ranch Type Appraisal Actual Survey Results No. No. No. 7 Dairy steer fattening 50 37 45 58 10 43 Beef breeding/fattening 75 56 27 35 8 35 Beefbreeding 10 7 6 7 5 22 Total 135 100 78 100 23 100 O-farm Investment and Lending Program 7. As forecast, most of the ranches participating in the Project were located in the Atlantic Zone in the SulJa Valley and La Ceiba area: No.of Zone Loans % Amount US$'000 $ Atlantic (North) 46 58 1,777 54 Central 24 31 1,062 32 Pacific (South) 9 11 462 14 Total 79 100 3,301 100 On-farm investments by categories of investment were very much as expected except in construction and machinery where actual investments were 284% and -1 !)8% respectively of expected values, and breeding cattle purchases were 23% below expectations. These differences might be a result of the increase in machinery and construction costs and the fact that more cattle handling facilities and housing were financed. In the case of cattle purchases, even though more animals were acquired than planned, the price of heifers was lower than expected and purchase of frozen semen partially replaced purchase of bulls. The following table shows that new pastures were established at triple the rate estimated at appraisal, while improvement of existing pastures was 24% below expectations. This data should be interpreted carefully, however, since technicians differed in their opinions as to what constituted a new or an improved pasture in on-ranch investment plans. - A 6 - Project Expenditure Appraisal Estimate Actual No.US$O00 No. US$'000 7 Pasture ImVrovement 25 550 806.25 20 26 311 1 013.7 24 New Pasture (ha) - - 94L.1 I - Existing Pasture (ha) 22,250 - - 16,910 492.6 - Fencing 2 575 328.75 8 2 352.5 334.6 8 New (kn) ([? - "- 44275 -60.6 Repair (ka) 1,800 - - 952 74.0 - hter supply (units) 1,155 476.25 12 606.5 284.5 Constructions 416 186.90 5 1,413 531.1 13 Machinery 352 313.75 8 20k 494.2 12 Breeding Stock (Heads) 6 805 1,913.10 47 6 781 1,483.9 36 Cows/Heifers - - 171 936. Balls 1,055 - - 610 528.3 - Frozen semen - - - 3,820 18.8 - Total - 4,025.0 100 - 4,160.8 100 8. There was no limit in regard to farm size or amount of sub-loan per farmer under the Project, and distribution and amounts of loans were as follows: Range of Sub-loans SubL_Toans Amount (US$) No. US$I000 7 9,000 - 12,500 1 1 10.3 - 12,501 - 25,000 21 27 407.9 12 25,001 - 37,500 19 2 579.0 18 37,501 - 50,000 17 22 719.7 22 50,001 - 75,000 11 14 642.4 19 75,001 - 100,000 10 12 942.1 29 Total 79 100 3,301.4 100 Average 41.8 Most of the loans, 7k%,were under US$50,000, but h84 of the funds went to sub- loans over US$50,000. Average sub-loan size was about US$42,000, which was considerable higher than the appraisal forecast of US$27,000, even when allowance is made for input price increases from 1969 to 1973, about 12%. -A7 - Fliancin, ard 'Wibrsements 9. Implementatia of the Project in regard to number of ranch plans initiated each year followqd closely the pattern anticipated at appraisal, but ranch investments were made in four years instead of five as planned. Ranch Plans Initiated 1970 1971 172 1973 197k Total Appraisal Estimate 40 65 30 - - 135 Actual 25 44 10 - 79 Total Ranch Investments (US$'000) Appraisal Estimate 758.3 1,450 1,168.6 50k. 144.0 4,025 Actual 485.6 1,412.0 1,832.0 431.4 - 4,161 10. Total Project cost was about 9% lower than planned because of the little use made of working capital (para. 6); thus, total ranch development cost was about US$0.5 million less than expected. Technical services were about 24% higher, partially because of the two assistants CB assigned to help the Project Director (Fara. 17). A smmaray is given below: Appraisal Estimate Actual --- ---US$1000------- Total capital development 4,025 4,161 Working capital (fattening steers) 1 ,030 378 Total on-ranch development 5,055 4,539 Technical services 176 218 Total Project cost 5,231 4,757 11 Project fund.- (US$3.3 million) were fully committed in April 1973 and totally disbursed in January 1975, almost a year before the closing date. To cope with sub-loan demand, it was,necessary in June 1974 to reallocate funds to Category I, long-term loans, from Category II, Technical Services. Ranchers contributed about 20% of on-ranch investment and PBs contributed 25% of sub-loans out of their own funds, while IDA reimbursed the remaining 75% through CB. The following table shows how Project funds were disbursed. - A 8 - Original Amount Transfer Total Long-term Loans ID Credit 2,450.0 + 29.6 2.179.6 Participating Banks 816.0 9J 825.9 Total 3,266.0 39.5 3,305.5 Technical Services IDI Credit 140.0 - 29.6 110.4 Central Bank 36.0 +E D.1107.1 Total 176.0 + 41.5 217.5 Total - - 3,523.0 12. Three PBs, known as the El Ahorro Group (El Ahorro Hondareffo Bank, Occidente Bank and Bancahsa), were the most active in the lending program, accounting for 62% pf the total, while two larger banks, the Atlaintida (associated with Chase Manhattan) and the Honduras Bank (associated with First National), participated to a lesser extent. The following table provides specifies: Number of Bank Sub-loans Committed (US$'000) AtlAntida 14 463.7 Ahorro Hondureflo 25 1,16.7 BANCAHSA 18 608.2 Occidente 7 28t.4 Financiera 6 181.5 Honduiras 619.9 79 3,301J4 - A 9 - 33. Terms and interest rates for the Credit and sub-loans were as foIlows: Number of Years Total Grace Term Interest % IDA to Government 10 50 0.75 Goveriment to CB 10 50 4.5 CB to PBs 3-5 8-12 5 PBs to ranchers 3-5 8-12 9 The Government of Honduras donated the funds withdrawn from Category II ('US$1"0,000) to CB to pay for technical assistance services. Organization and Management J'&. The Government of Honduras, by Decree No. 865 of December 18, 1969, established the Project Commission, whose members were: the Minister of Finance and Commerce (President), the Minister of Natural Resources, the President of the Central Bank, the General Secretary of the National Planning Board, the President of the National Development Bank, three representatives of the Participating Banks, a representative of the Cattlemen's Association, and a representative of the National Agrarian Institute (INA). The Project DLrector acted as Secretary of the Commission. This body met regularly every other month without major problem, although, in some instances, the regular mtobers were represented by authorized alternates. The Commission, however, did not make manW significant contribution to the implementation or coordination of the Project. 1 . The Project Eirector, an expatriate, was employed by IDA, seconded to BC and approved by the Commission. He took office on October 26, 1969 after a two weeks briefing in the Bank at Washington. Then, in November 1969, a mission from the Bank, including the Project Director in Ecuador, assisted him in organizational matters. By December 21, 1969, when the Commission held its first meeting, seven PBs had already signed the Project Administration Agreement and several sub-loans applications were ready for a;proval. It was very difficult to get this first project started because CB and PB officers did not know much-about it (legal documents and the appraisal report had to be translated into Spanish) and because, after the June 1969 war against El Salvador, the economical and political situation of the country absorbed most of the Government's (the Commission) attention. - A 10 - 16. Deterioration of the Honduran economy after emigration of the Salvadoran labor force and retirement of Honduras from the Central American Common Market (July 1969) yas aggravated in 1971 by uncertainties caused by the campaign to elect a president (1971), the first election to be held since 1961. Agrarian unrest, land invasion by campesinos, and incapacity of the Government to solve land tenure problems made an unfavorable climate for Project execution to the extent that PBs suspended lending operations on several occasions, fearing that their investments would be lost if ranchers' lands were invaded by campesinos or expropriated by INA. This problem was partially solved when CB provided (October 1972) a guarantee to cover loans made by PBs. 17. Although IDA did not require it, CB appointed two assistants to the Project Director, who proved to be very useful in Project execution. One of them became Project Director when the post was left vacant in July 1972. Six PBs were active in the lending program and they seconded livestock technicians to the Project Director. These technicians were trained by the Project Director on the job. Field trips to the neighboring countries of the US and Mexico were also made and a special short course was held, with the cooperation of the Interamerican Institute for Agriculture Science (GAS). Most of these men became efficient and hard-working professionals, motivated by a higher salary, more responsibilities, and better working facilities than they had had before. Nevertheless, PBs did not have an adequate salary increase policy and turn-over of these technicians became a problem. 18. Administrative problems arose when the program was first implemented, but they were solved when the President of CB gave full support to the Project and the technical unit offices were moved out of the CB building. Only trans- portation for the PD remained a problem until a late stage of Project execution when a vehicle was finally assigned to him, as called for in the appointment contract. In the meantime, he used his own car for Project matters. C. Conclusions 19. Project implementation and execution were highly satisfactory, and, although difficult to quantify, the benefits projected at appraisal were fully achieved. 20. A Second Project is under execution to continue the lending program. ICATKc 0 ORTIRN REG10N S.n Pdro Sula - ) ø T.S.citalpae-> *• 5 - HONDURAS -(L-vestoc Developmeni Project PiciFic Ok SOurE REGoN Locotion of Projed Forms - First Lvestock Projec CR 179 HO c3 AJø• N)Second Livestock Projec CR .34. HO L San Pedro Suto Abolloir 0 8.nch.r Interie~.d 0n kRch (19) £ Iancher Intervewed off Banch (4)> -- - - 1.210nal sondar1.a