Papua New Guinea - Smallholder Livestock Credit Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 3997 PROJECT PERFORMANCE AUDIT REPORT PAPUA NEW GU [NEA - SMALLHOLDER LIVESTOCK CREDIT PROJECT (CREDIT 348-PAP) June 23, 1982 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 CLUA Clan Land Usage Agreement DPI Department of Primary Industry IAESR Institute of Applied Economics and Social Research OPCR Overview of Project Completion Report PCR Project Completion Report PNG Papua New Guinea PNGDB Papua New Guinea Development Bank CURRENCY EXCHANGE RATES Name of Currency (Abbreviation) Kina (K) Year: Exchange Rate: Appraisal Year Average US$1 = A$0.84 Intervening Years Average US$1 = KO.72 Completion Year Average US$1 = KO.64 Fiscal Year of Borrower July 1 - June 30 FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT PAPUA NEW GUINEA - SMALLHOLDER LIVESTOCK CREDIT PROJECT (CREDIT 348-PAP) TABLE OF CONTENTS Page No. Preface ................................................................... i Basic Data Sheet ............................................ .. . ii Highlights ............. ........................................ iv PROJECT PERFORMANCE AUDIT MEMORANDUM I. PROJECT SUMMARY ...................................... . 1 Formulation and Design .............................. 1 Implementation .... ....................... ..... 2 Impact .................................. 3 Bank Performance ... .................. .... 3 II. MAIN ISSUES ....................... 4 A. Smallholder Cattle Component ... ............. 4 B. Village Land Tenure, Social Relations, and the Cattle Subprojects. .................................. 8 C. Extension and Farmer Training.........................9 D. Beef Prices, Imports, and Domestic Production Trends .. . . . . . . . . . . . . . . . . . . . . ..1 E. Bank Performance. ...................................... 12 F. Conclusions ........ . .................................. 14 Tables 1 - Cattle Numbers, 1956-78 ............................ 15 2 - Smallholder Cattle Loans, 1967/69 - 1977/78 ............ 16 3 - Growth of PNG Cattle Stocks by Period ...... . .17 4 - On-farm Technical Achievements ......................... 18 5 - Financial Data Concerning the Cattle Projects ......... 19 6 - Real Cattle Prices in PNG .......................... 20 7 - Beef Consumption 1971-78 ........................... 21 8 - Cattle Slaughterings at Licensed Premises, 1971/72 - 1977/78 ................................ 22 OVERVIEW OF PROJECT COMPLETION REPORT Introduction ...... ..................... 25 Project Implementation o ..................................... 25 Agricultural Impact ........................................ 28 Institutional Performance and Development .................. 32 Bank Performance ................................................... 33 Conclusions ..................................................... 34 Lessons Learned ............................................ 35 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. Table 1 - Project Implementation, Number of Farmer Beneficiaries, Total Value of Subloans and Technical Services and Average Value of Subloans, Appraisal Estimates Compared to Actual Results ....................................... 37 COMPLETION REPORT1/ Summary and Conclusions .................................... 41 Introduction ............................................... 43 Background ................................................. 43 Project Formulation ........................................ 44 Basic Project Description ............ . ......... . ..... . 45 Detailed Project Performance ............................... 45 Institutional Performance and Development ................. 50 World Bank Performance .......... ....... ....... ........ 55 International Loans to PNG ................................. 56 Recommendations ........................... 0................. 56 Annexes I-X2/ Map IBRD 3922 1/ As the full PCR contains over 200 pages, only the summary is provided here. The full report is available from OED. 2/ Available in OED. PROJECT PERFORMANCE AUDIT REPORT PAPUA NEW GUINEA - SMALLHOLDER LIVESTOCK CREDIT PROJECT (CREDIT 348-PAP) PREFACE This is a performance audit of the Smallholder Livestock Credit Project in Papua New Guinea for which Credit 348-PAP in the amount of US$5.0 million was approved in November 1972. The project was closed in June 1980, extended from the original closing date of December 1976. This audit report consists of an audit memorandum prepared by the Operations Evaluation Department (OED), a Project Completion Report (PCR) received in the Bank on March 24, 1981, and an Overview of the PCR (OPCR) dated May 6, 1981. The PCR was prepared by the Borrower and the OPCR was prepared by the East Asia and Pacific Regional Office. The audit memorandum is based on a review of the Appraisal Report (PA-144a) dated October 16, 1972, the Credit Agreement, dated January 4, 1973, the PCR, and the OPCR. Corre- spondence with the Borrower and internal Bank memoranda on project issues as contained in relevant Bank files also have been consulted, and Bank staff associated with the project have been interviewed. Several relevant Govern- ment and academic studies were also reviewed. An OED mission visited Papua New Guinea and Australia in September- October 1981. Discussions were held with officials of the Department of Primary Industry (DPI), and the Papua New Guinea Development Bank (PNGDB). Project sites were visited in the Markham Valley and the Eastern Highlands. Discussions were held with sub-borrowers, large ranchers, project staff, and staff of the Institute of Applied Economics and Social Research (IAESR). The author of the PCR was interviewed in Australia. The PCR is voluminous, providing detailed information on nearly every aspect of.the project. The audit finds the PCR both thorough and generally accurate with respect to the project's implementation. The audit memorandum summarizes the project and discusses several principal issues. OED gratefully acknowledges the valuable assistance provided by Government staff, as well as the other individuals interviewed. Their assis- tance greatly contributed to this report. A copy of the draft report was sent to the Borrower on April 27, 1982; no comments have been received.  PROJECT PERFORMANCE AUDIT REPORT PAPUA NEW GUINEA - SMALLHOLDER LIVESTOCK CREDIT PROJECT (CREDIT 348-PAP) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Actual as % of Estimate Estimated Actual Appraisal Estimate Project Costs (US$ million) 9.8 9.8 100 Credit Amount (US$ million) 5.0 5.0 100 Board Approval Date - 10/31/72 - Credit Agreement Date - 01/04/73 Effectiveness Date 05/16/73 11/15/73 - Date Physical Components Completed 12/31/76 06/30/80 184 Proportion Completed by Above Date (%) 50 100 - Closing Date 12/31/76 06/30/80 184 Economic Rate of Return (%) 14 2 - Financial Performance - Worse than expected - Institutional Performance - Worse than expected CUMULATIVE DISBURSEMENTS FY73 FY74 FY75 FY76 FY77 FY78 FY79 FY80 Appraisal (US$ million) 770 2040 3520 4760 5000 - - - Actual (US$ million) - 912 1872 2362 2900 3293 4175 5000 Actual as % of Appraisal 0 45 53 50 58 66 83 100 Date of Final Disbursement 02/28/81 MISSION DATA Date No. of Manweeks Specializations Performance Types of Mission (Month/Year) Persons in Field Represented/b RatinI Trend/-d Problems/e Appraisal 02/72 4 12 AE, AE, L, A - - - Supervision I 05/73 /a 4 L, OP 1 2 - Supervision II 03/74 1/a 2 L 2 3 M, 0 Supervision III 02/75 i/a 17-1/2 A, L, AE, AE, AE 2 2 M, O Supervision IV 08/75 2/1a 2 L/E, OP 2 2 M, T, F Supervision V 06/76 1 1-1/2 L 1 - - Supervision VI 07/77 2 2 L, AE 1/2/ 2 - Supervision VII 11/78 2 3 AE, A 2 2 M, 0* Supervision VIII 11/79 1 1 AE 2 2 M,T OTHER PROJECT DATA Borrower The Administration of Papua New Guinea Executing Agency Papua New Guinea Development Bank/ Department of Primary Industry Follow-on Project Name Agricultural Credit II Credit Number 1149-PAP Credit Amount (US$ million) 15 Date Board Approval 12/07/81 /a Mission responsible for supervi3ing other projects concurrently. /b Specialization: L- Livestock, OP - Oil Palm, A - Accounting, AE - Agricultural Economist. /c 1 = problem free or minor problms, 2 - moderate problems, 3 - major problems. 7- 1 = improving, 2 - stationary, 3 - deteriorating. /e F financial, M - managerial, T - technical, P - Political, 0 - Other (Country Economic), 0* - Other (Prices, Currency Overvaluation). /f Two Summary sheets were produced. The first, dated August 2, 1977, gave the project a I rating, the second, dated August 22, 1977, shows a 2 rating.  - iii - PROJECT PERFORMANCE AUDIT MEMORANDUM PAPUA NEW GUINEA SMALLHOLDER LIVESTOCK CREDIT PROJECT (CREDIT 348-PAP) HIGHLIGHTS Papua New Guinea has excellent potential for cattle production. Cattle raising was, however, an expatriate activity until the early 1960's. The project aimed to rapidly increase smallholder (national) development together with smaller investments for pigs and for poultry. Funds were also provided for livestock extension, training and research facilities. A major expansion of the Department of Primary Industry's (DPI) livestock extension staff was a condition of lending. The project largely failed to achieve its objectives, although it provided intangible benefits through experience which will be of future value to smallholders. Credit 348 financed 1,028 smallholder cattle units over nearly eight years. These units contributed to a smallholder cattle herd increase from 19,000 animals in 1972 to 49,500 by 1976. Severe problems then began to emerge. Inadequate extension and farmer training, poor small- holder management, land tenure conflicts, and deteriorating beef prices caused a decline in herd numbers to about 30,000 in 1981. Many smallholders lost control of their cattle and about 7,000 cattle went feral, necessitating an extermination campaign. Only about 33% of the units established are still operating. Most of the pig and poultry subprojects originally established failed so new designs were introduced. Most of the new pig projects also had failed by 1980, but about 130 of the new 300 poultry projects are doing well under a nuclear estate arrangement. The reestimated ERR is 2% compared with an appraisal estimate of 14%. Other points of interest are: - extension, training, and research facilities proposed were completed as planned, but substantial DPI staff turnover reduced the extension and research programs' effectiveness. (PCR paras. 6.06-6.08); - attempts to introduce new technologies proved optimistic, because under the traditional tenure system no individuals controlled sufficient Land for an economic cattle unit, most had to be estab- lished by groups, complicating management. Project lending was slower than expected because it proved difficult to establish cattle units of economic size. PNGDB financed other, smaller cattle units (which had great appeal) with non-IDA funds. (PPAR para. 24, OPCR para. 23); - deterioration of the smallholder sector after 1975 received in- sufficient attention. (PPAR para. 50); - perceived risks need to be highlighted in the appraisal reports to ensure that they are monitored carefully during supervision (PPAR paras. 48-49); - the need tc ensure that the extension demands of farmer investments financed from non-IDA sources do not detract from previously agreed project support (PPAR paras. 21 and 38).  - 1 - PROJECT PERFORMANCE AUDIT MEMORANDUM PAPUA NEW GUINEA SMALLHOLDER LIVESTOCK CREDIT PROJECT (CREDIT 348-PAP) I. PROJECT SUMMARY Formulation and Design 1. Cattle production was introduced to Papua New Guinea in the late 19th century, but remained a European activity until the early 1960's when several smallholder projects were initiated by Government. Shortly there- after, a Bank mission recommended expansion of the cattle sector because of PNG's production potential, its growing internal beef market, and the belief that villagers could easily acquire the requisite management skills, The Bank also recommended that Government and European-owned herds first be ex- panded and, subsequently, using the breeding animals so produced, the small- holder sector. 2. The Government of PNG adopted this strategy and the Bank provided two loans to assist implementation. The first stage was supported by the Second Agricultural Development project (Credit 175-PAP, US$5.0 million, signed on January 30, 1970), while the second stage was financed by the Smallholder Livestock Credit project. The principal financing component of this latter project included both smallholder investments for beef, pig and poultry production and livestock extension, training and research facili- ties. A condition of lending was the major expansion of DPI livestock exten- sion staff. DPI was responsible for project implementation, including the preparation of farm plans, in conjunction with the PNG Development Bank, which received IDA funds for on-lending to farmers. 3. The project was prepared by the Government in 1971 and appraised in January/February 1972. The credit was negotiated in September 1972, signed on January 1973, and became effective in May 1973 although retro- active financing of PNGDB cattle subloans was permitted from May 1, 1972. The loan was closed fully disbursed on June 30, 1980 (extended from the original closing date, December 31, 1976). 4. The appraisal report expected 870 smallholder cattle subprojects, which comprised 92% of anticipated project sub-lending, 270 two-sow sub- projects and 120 poultry subprojects (half layers and half broilers). Total project cost was estimated at US$9.8 million, of which the sub-lending com- ponent was estimated at US$6.7 million (68%). 5. The financial rate of return to smallholders was estimated at 11% for cattle, 25% for pigs, and 69% for broilers; all investments were considered high risk, particularly those for pig and poultry production. The economic rate of return was estimated at 16% for beef, 25% for pigs, and 33% for poultry, 14% overall. - 2 - 6. The PNGDB initiated smallholder cattle lending in 1967 using its own funds and by 1971 this lending had grown substantially. Smallholder cattle units had great appeal to villagers for both economic and non-economic reasons and were also promoted by DPI and PNGDB, partly for political reasons. Many of the units established were uneconomically small as it was difficult within the traditional land tenure system (where many land rights were com- munal) to obtain sufficient land for a commercial project. 7. Project design of the smallholder cattle subproject, which reflected previous DPI experience, was intended to make the most effective use of the land tenure system prevailing in PNG. Prior to lending, the borrower was required to obtain a Clan Land Usage Agreement (CLUA) indicating that he had been granted lifetime use of the area to be fenced and stocked by his clan, and he had to erect, under DPI extension officer supervision, a suitable fence and stockyard. Although the CLUA was not legally binding, the Bank felt it would provide some guarantee of tenure and hoped that future legislation might permit freehold title to all such land, thereby promoting more intensive land development. Loan criteria required that only subprojects with a minimum unit size of 15 breeders, 1 bull, and 9 steers would be financed; smaller units were considered uneconomic. 8. The pig and poultry subproject designs were also largely based on DPI tested systems. The native pig is the most important domesticated animal in PNG and has been a major part of the culture for centuries. The project hoped to increase productivity through improved husbandry and increased feed rations. The poultry subproject aimed at increasing egg and broiler produc- tion through small semi-intensive systems located near main towns where markets and extension were on hand. Implementation 9. The project was implemented more slowly than expected, and a sub- stantial change was introduced to the pig and poultry components. The PNGDB had initiated smallholder cattle lending in 1967 using its own funds and by 1971 this lending had grown substantially. The rate of lending began to drop just after the project was approved, however, and loans also averaged only about half (in real terms) the amount expected. The slower rate of lending resulted in extension of the project with 1,028 cattle loans to smallholders over eight years versus the expected 870 in three and one half years. 10. Most of the pig and poultry subprojects failed so new designs were introduced during the project. The new peri-urban pig subprojects were larger in size, receiving only weaner pigs and sufficient feed for production to slaughter; these had some initial success, but most had failed by 1980. The new 1,000 unit poultry subprojects were designed to operate in conjunction with an expatriate-owned processing complex providing all the major inputs, a high level of management service, and a guaranteed market. Of 307 such subloans, 132 were still operating in 1980 and most were doing well. - 3 - 11. The extension, training, and research facilities proposed were all completed as planned, but substantial DPI staff turnover associated with PNG self-government in 1973 and Independence in 1975 together with recent budget constraints have limited the effectiveness of the extension and research programs. The farmer training program failed largely because of inadequate design. Impact 12. The project has largely failed to achieve its objectives, although it may have provided intangible benefits through experience in livestock management and in- commercial enterprises which will be of future value to villagers. 13. As a result of project and other PNGDB lending, the smallholder cattle herd grew from 19,000 animals in 1972 to 49,500 by 1976. Poor small- holder management, inadequate extension, land tenure conflicts, and declining beef prices then caused an initial herd decrease to 48,000 in 1978, followed by an abrupt decline thereafter. DPI staff estimate that the smallholder herd was only about 30,000 in 1981. As a result of deteriorating infrastructure, many smallholders have lost control of their cattle, which, in some cases, have become feral. PNGDB is now making very few new loans for smallholder cattle and only weak efforts are being made to consolidate and assist the existing sublenders. For lack of accurate data, calculation of an actual FRR or ERR is difficult. Each must lie below appraisal estimates. The PCR estimates the ERR is 2% and the FRR 10%. Bank Performance 14. Most of the problems ultimately faced by the project were identified at appraisal, which was very perceptive. Nonetheless, project design was optimistic given the risks identified, and the important issues requiring continued attention were not highlighted in the report, being discussed only in the annexes. 15. Bank supervision was light (eight missions in eight years), the first four missions were carried out in conjunction with other projects, staff continuity was lacking, and an emphasis was placed on lending issues. The growing deterioration of the smallholder sector after 1975 received little attention. 16. The Bank decided not to lend further for smallholder cattle subproj- ects in the follow-up project (Second Agricultural Credit project, Credit 1149-PAP, US$15 million, signed December 7, 1981), believing that no further funds should be provided to this deteriorating subsector. Additional funds were provided for pig and poultry development. The audit mission believes that it would have been useful to provide small additional funds to improve extension and training for the existing cattle smallholders. II. MAIN ISSUES A. Smallholder Cattle Component Design 17. The appraisal mission sought to design a project which would con- tinue expansion of the smallholder sector on a sound basis. The approach adopted was heavily influenced by DPI's previous experience, on which the mission sought to improve. The two basic problems confronted were 1) the constraints imposed by the traditional land tenure structure, and 2) the villagers' lack of experience with cattle. The possible effect on DPI ex- tension staff of PNG's impending self government in 1973 and independence in 1975 was an additional concern. 18. Customary ownership and use of land in PNG was subject to agreement. Certain rights were communal, such as hunting, while other rights were often proprietary, such as control of land on which subsistence gardens had been planted. The traditional tenure system was complex, but flexible, and often depended on the relative strength of litigants. 19. The PNGDB was aware, when it initiated smallholder cattle lending, that no single individual within a village controlled sufficient land to establish a commercial cattle unit. It first experimented with a day herding- night paddock system in an attempt to work within the existing tenure system, but subsequently moved towards larger projects on a fixed piece of fenced land in order to achieve improved project management and avoid the worst effects from communal grazing. To achieve the larger projects, the PNGDB introduced the Clan Land Usage Agreement (CLUA) by which two leaders of the loan applicant's clan had to indicate that other members of the clan had granted the applicant the sole right for life to use the land enclosed for cattle raising. 20. Although not legally binding, the PNGDB hoped that this Agreement would permit effective implementation of the cattle subprojects and, eventual- ly, that enclosure of clan land would lead to the development of freehold title. The profitability of livestock was expected to encourage additional enclosure of common lands, permitting the evolution of private land rights which would promote individual initiative. Eventually, livestock raising would give way to crop production. Thus, the cattle lending program was seen as an instrument of both short and longer term economic and social development. - 5 - 21. The appraisal report followed this same approach!/ indicating that there was no evident alternative by which on-lending could be imple- mented, nor by which a transformation from traditional land rights could begin. The mission also decided that project financing would be permitted only for cattle units meeting a minimum size criterion. Many villagers wanting a cattle unit had previously formed groups lacking sufficient re- sources to implement a project of economic size and DPI agents had not always excluded such projects from PNGDB financing, particularly as there was great political support for the smallholder cattle projects. The mission estimated that the FRR of a subproject containing 15 breeders, 1 bull, and 9 steers on 60-80 hectares was 11%, and imposed this unit (or its equivalent) as the minimum size. Nonetheless, no constraints were placed on DPI and PNGDB to restrict their lending from other fund sources for smaller cattle subprojects. 22. Few villagers had any previous contact with cattle. The pig was traditionally the largest domesticated animal in PNG and cattle were seen by the villagers as huge and dangerous. Most of the cattle supplied to smallholders came from large ranches where they had been handled infrequently and then by men on horseback. The cattle were difficult to handle until they had been "gentled", and the inexperienced and often fearful smallholders, working on foot, needed encouragement and instruction to achieve this. Smallholders also knew nothing about animal and pasture management, but good project results depended on their acquiring and using this knowledge. 23. The appraisal report noted the role which farmer training and, particularly, DPI extension would have to play. The extension staff: farmer ratios proposed in the report were unusually high, but the appraisal report argued that 1) most of the cattle expansion would take place in a few dis- tricts where the existing staff were already spending up to 90% of their time on the existing cattle units, 2) upcoming PNG independence would likely bring significant staff changes, with departure of some expatriates and an increase in newly trained nationals, 3) subprojects would be widely scattered in relatively inaccessible areas. The appraisal report also provided an incisive list of the tasks which extension would have to fulfill (Appraisal Report, Annex 2, para. 16). The large and costly expansion of DPI extension staff was initially questioned in the Bank, but the appraisal mission's arguments proved persuasive. Implementation and Impact 24. Disbursement was slower than expected because it proved organiza- tionally difficult to establish cattle units meeting the minimum size criteria imposed. Rather than fully adjust to this criteria, PNGDB continued to lend for smaller units, using non-IDA funds to finance those units smaller than 1/ Appraisal Report No. PA-144a, Annex 6, paras. 12-16, dated October 16, 1972. - 6 - the minimum project size. The demand for new cattle units then declined after 1974. Data on the precise timing of project subloans are not avail- able, but analysis of PNGDB's total lending for smallholder cattle projects provides a useful perspective (Table 2). The rate of lending for smallholder cattle projects peaked in 1971/72 and declined after 1973/74, despite the additional funding available from Credit 348.1/ The evolution of PNG's smallholder cattle herd followed closely the pattern shown by the rate of lending. The smallholder sector grew at 40% per year from 1967 to 1973, and at 22% per year from 1973-1976, when growth stopped abruptly and decline began (Table 3). 25. Severe problems became evident within the smallholder sector in the mid-1970's. Low levels of management and deteriorating infrastructure resulted in an increasing number of smallholder cattle becoming uncontrolled. DPI's veterinary staff first became concerned in 1975 because of the poten- tial disease problems. Buffalo fly had entered PNG in 1974 and efforts to eradicate it were already being frustrated by the inability of smallholders in the Markham Valley, the principal cattle area, to fence their animals. 26. The problem grew more severe, with many escaped animals becoming feral. A costly campaign to eliminate feral cattle, involving DPI staff, private contractors, helicopters and trucks, was initiated in 1976. The carcasses of animals killed were taken to abattoirs for processing and sale, and proceeds were credited to the accounts of smallholders who could identify their animals; 20% of the value was deducted for campaign costs. (Actual costs were about 40%). By 1978, when it was halted by the shortage of funds and personnel, the campaign had killed over 3,000 animals. 27. The government originally planned to reconstitute the smallholder projects liquidated through the feral cattle campaign, but ultimately decided not to do so because of farmer disinterest, DPI staff shortages, and a deci- sion that smallholder cattle sublending ought to be deemphasized as a develop- ment priority. Since 1978, very few cattle subloans have been granted by the PNGDB. 28. DPI also undertook a smallholder cattle census in 1978. The census found 46,638 cattle on 3,325 units, averaging 14 animals on 27 fenced hect- ares. However, approximately 25% of these animals could not be fenced for a variety of reasons. The feral animal problem originated from this situation and is continuing. DPI staff estimate that about 500 to 1,000 feral cattle remained in the Markham Valley in 1981, and another 3,000 to 4,000 in Popen- detta (Northern Province). No official estimates of the smallholder herd 1/ The Bank's efforts to impose a minimum subproject size also had moderate impact. The PCR (Annex 1, para. 1.08) estimates that the average initial stocking of IDA-financed units was 1 bull, 9 heifers, and 8 steers on only 21 hectares. The IDA-financed units remained larger, on average, than the PNGDB smallholder cattle units financed from other sources. - 7 - have been made since 1978. The PCR (Annex 1, para. 1.45) estimates that 90% of the smallholder projects were still in operation in 1979. Considerable evidence indicates, however, that the size is declining rapidly. Smallholder slaughter (including that in informal markets) has amounted to an estimated 10,000 animals per year, well beyond the sustainable level. DPI estimates that smallholders are slaughtering twice as many females as males, liquidating their breeding herds. Senior DPI officials told the audit mission that only 33% of the projects initiated remain operational in 1981 and they expect that not more than 25% of the original total will survive. One senior experi- enced veterinarian estimated that the total PNG herd had declined to 95,000 animals in 1981, with the smallholder herd containing only about 20,000. This estimate seems low to the audit, but it is an indication of the sharp decline which has occurred. The evidence available indicates, therefore, that there are not substantially more smallholder cattle subprojects, or cattle, than there were when the project became effective. 29. The technical coefficients achieved on smallholder cattle subproj- ects (Table 4) are substantially below the (modest) appraisal expectations and the coefficents may be worsening with time. The on-farm technical problems which have led to these results are well described in the PCR (Annex 1, paras. 1.32-1.38). 30. The resulting financial situation is also less favorable than ex- pected at appraisal, but still somewhat unclear. The PCR estimates the FRR to be about 10%; this seems high to the audit. The PCR estimates the ERR to be 2%, reflecting the various subsidies which smallholders have obtained. Neither the estimated FRR nor the ERR include the intangible benefits which the project may have provided in terms of experience gained by the small- holders in cattle production. 31. The loan repayment situation is also still unclear. Most small- holders have attempted to pay their debts and to do so have often used income from other sources, e.g., coffee. The DPI also mustered and sold the stock of some units which were in arrears or whose animals had gone feral. None- theless, the PCR (para. 7.15) indicates that 70% of the cattle subprojects were in arrears on June 1980, with 80% of these arrears over three months and 50% over six months. The total amount in arrears was K649,884, or 22% of the total initial value of cattle subloans (K2,901,624); 41 projects, with bad debts of K47,140, had been written off. The PNGDB believes that many of the arrears were caused by an overoptimistic scheduling of loan repayments and hopes that these problems can be resolved by review and rescheduling. The audit was not able to determine to what extent this has been possible when smallholder projects are liquidated. -8- B. Village Land Tenure, Social Relations and the Cattle Subprojects 32. Land tenure considerations have played a major role in the estab- lishment, management, and closure of smallholder cattle units supported by the project, and the subprojects themselves have had great impact on the villages principally because of the tenure transformations wrought and the large number of individuals affected. Th se issues are well discussed in a number of DPI staff and academic studies.! 33. A cattle subproject was usually organized by a man of some esteem in the village who enlisted friends and relatives together with the land over which they have a claim. The leaders appealed to their followers by asserting that participation would enhance prestige, earn considerable income, and provide fresh beef. The exact distribution of the benefits was left unstated, but the followers expected to benefit substantially because they were giving up part of their land and were contributing labor and finance to the project. After enlisting followers, the leaders sought wider support for the proposed land takeover, requesting permission from the elders of their patrilineage cluster. Conflict became more intense and specific at this time because the boundary often enclosed land, including subsistence plots, to which non-group members had a claim.Z. 34. The leaders were not the sole owners of the cattle units, yet they were the dominant force, both in terms of operation and in terms of the prestige and benefits obtained. Most leaders were already men of consequence, but they became more important as a result, even being separated from their followers by nomenclature, i. e., the "cattle bosses". The bosses could not operate the units alone because of the work involved, nor did they wish to: part of the benefits obtained was being the boss. The followers assisted because of kinship ties, to maintain their group's prestige, and to obtain future material benefits. But they also wanted to limit their contributions in line with that they expected to receive in return. Table 5 gives an indication of cash benefits received both by the bosses, which were generally positive, and the followers, generally negative. For the three oldest sub- projects, the bosses had contributed 58% of the money but had kept 93% of the net revenues. In each case, the bosses also obtained additional finance from their followers to help with loan repayments. 1/ See, in particular, Lawrence S. Grossman, "Cash, Cattle and Coffee: The Cultural Ecology of Economic Development in the Highlands of Papua New Guinea," Ph.D. dissertation, The Australian National University, 1979. 2/ Subsequent studies have shown, however, that at least in some villages, nearly all villagers eventually participated in a cattle subproject. See Grossman, op. cit. - 9 - 35. These data do not fully reflect the benefits, because the cattle bosses had given their followers free meat or offal on various occasions. Nonetheless, the data suggest a strongly skewed distribution of costs and benefits amongst the participants. 36. The problem which the audit wishes to emphasize is not the uneven distribution of income and wealth which the smallholder cattle program may have helped to create, but the difficulties which the cattle groups (bosses and followers) soon encountered regarding the appropriate distribution of costs and benefits and which strongly affected success. In 1ddition to the problems of coordinating such a complex group undertaking,!' the followers reduced their contributions and even engaged in sabotage if the bosses did not adequately share project revenues with the followers or demanded too much cash or labor without reciprocating. A great deal depended on perceptions, and the high expectations created at the beginning of the subprojects subsequently became part of the problem. The long gestation-slow pay out of cattle breeding compounded this situation, as did the posture of the PNGDB which sought a more rapid payback of the loan than had been planned at appraisal (7 years vs. 12 years). When animals were sold to an abattoir, proceeds had to be transferred directly to the PNGDB which deducted the scheduled payments. Since the scheduled payoff period was short and the technical coefficients achieved were less than estimated, the PNGDB generally kept more than 80% of slaughter proceeds. 37. As the bosses and followers became disenchanted, the management of the cattle declined further. The herd was rotated infrequently and pastures became overgrazed; cattle were handled only occasionally, and became increasingly unmanageable; fences and yards deteriorated and the animals escaped; if not rounded up, the cattle foraged in subsistence plots, resulting in complaint and claims Eor compensation. Or worse, they escaped and were lost for good. It is not surprising that within this context many bosses and/or followers decided that it would be better to liquidate the cattle subproject rather than continue. Of course, liquidation, when it occurred, did not determine the ultimate disposition of the project land. It is clear that the bosses generally wanted to retain claim to this land for themselves; the followers obviously have resisted. How this situation will be resolved is still unclear. C. Extension and Farm Trzining 38. DPI and PNGDB staff strongly encouraged the smallholder cattle units from the initiation of lending in 1967 through the mid-1970's. The subprojects appeared to offer a means to increase profitably villagers' commercial involvement and to speed PNG's development. Furthermore, in the early 1970's after independence had been set, such on-lending was a politically attractive means of demonstrating the goodwill of the DPI and the 1/ It appears that although village residents were previously accustomed to mutual assistance on certain agricultural activities, there was little or no prior experience with command operation of any economic activity. - 10 - PNGDB toward nationals. The PNGDB was strongly criticized within PNG in the early 1970's for its alleged overly strict dealings with nationals, especially small farmers, in favor of expatriate borrowers. Smallholder cattle projects thus simultaneously became very popular with both villagers and with extension staff, leading to a sharp increase in lending in 1971-1973. 39. DPI extension officers encouraged villagers to apply for loans, often reassured them that the financial returns would be substantial, and provided considerable assistance in starting up the cattle unit. The effort to develop units rapidly meant that the extension staff were spread very thin and, partly as a result, DPI staff emphasis was on supervision and regulation, telling villagers what to do or doing it for them, rather than on education, which would have made the villagers more self-reliant. The lack of con- sultation with villagers on important decisions often offended them and reduced their motivation. 40. The quality of extension was mixed even before the project. Many extension officers, although enthusiastic, were not cattle experts which contributed to subsequent management problems. Others promoted the ventures because they believed their own prospects for advancement were partly depen- dent on the number of new cattle units established. With a massive increase in paperwork as the project continued, extension agents spent increasingly less time in the field and, when there, greater time on loan issues. Exten- sion staff found themselves required to act as collection agents, which made their visits less attractive to farmers, and less productive in terms of management improvement. More importantly, however, it proved impossible to achieve and maintain the projected increase in extension staffing, at least in terms of quality, as PNG moved towards independence. There was a large scale departure of expatriate staff and a rapid increase in local staff, who pos- sessed less training and experience. Local staff were frequently promoted rapidly, switching from one position to another before they could become effective. Control over DPI extension staff was then decentralized to the provincial level, where priorities were not always consistent with intensive assistance required for smallholder cattle units. 41. The farmer training courses also failed to achieve their goals. It was usually a condition of loan approval that at least one group member attend a course. However, it was not required that this person was the "cattle boss" and, in practice, was usually not. The bosses preferred to remain in the village to look after their interests and sent a younger helper instead. But the younger helpers (trainees) were usually without significant influence in the village and, on return from the course, were unable to have much impact, either by conveying knowledge to the boss or to the others who carried out much of the necessary work. As the bosses were usually unwilling to pay them for work, the trainees put in little time with the animals, had little opportunity to learn-by-doing, and often drifted off. 43. Farmer training declined after 1976 because of the increasing unwillingness of farmers to stay away from home, disillusionment with cattle production, and a shortage of provincial travel funds for farmers. Training staff were also drawn away for work in the feral animal campaign and in the smallholder cattle census. - 11 - D. Beef Prices, Imports, and Domestic Production Trends 44. The OPCR (para. 14) emphasizes that declining beef prices played a crucial role in slowing smallholder cattle growth after 1974, and attributed part of the price decline to Australian "dumping" in the PNG market during the international beef market downturn from 1974 to 1977. The audit agrees that declining beef prices have been a factor in the smallholder sector's decline, but believes the OPCR exaggerates its importance relative to other factors. The audit also found no evidence of beef "dumping". Table 6 shows the price movements for prime steers slaughtered in the Lae abattoir, the largest slaughter outlet in PNG, between 1971 and 1980. Nominal beef prices rose throughout the period, but real prices declined sharply in 1974 and have fallen further with an overall decline since 1971 of about 30%. Nonetheless, representatives of the Graziers' Association, including the major private producers in the Markham Valley, told the audit mission that beef prices had been maintained throughout most of the 1970's at satisfactory levels and that no dumping occurred during 1974-1977. They stated, and DPI confirmed, that an informal agreement has existed in PNG between domestic producers and retailers (with active government support over most of the 1970's) to maintain domestic beef prices and encourage the domestic industry. Internal prices declined after 1974, but still remained above international prices until 1977. Sub- sequently, international prices rose in real terms while internal prices continued to decline. Continuing appreciation of the Kina since 1977, at least partially to restrain the cost of imported goods, including beef, was an important factor. 45. The Graziers' Association indicated that imports of cheap meat cuts, such as brisket and mutton, have become increasingly competitive during the last six years, but this competition was still controlled until 1979-80 when large amounts entered and were sold principally through newly established (and often temporary) retail outlets, thus escaping the marketing system which previously helped control imports. These imports also entered just as the domestic beef mlrket was declining after a sharp decline in coffee earnings in the Highlands.- 46. This analysis seems consistent with the data on consumption, domestic production, and imports, shown in Table 7. Total domestic consump- tion of fresh and frozen beef rose by 11%/yr. between 1971/72 and 1976/77, but domestic output rose by 20%/yr. while imports grew by only 5% per year. 1/ The Graziers' Association indicated that the Government is now trying to restrict the import of cheap meats. The audit mission received mixed reports regarding the Government's intent from other sources. Because the local beef industry cannot produce whole carcasses at such a low unit price as the cheap imported cuts, which are heavily discounted in the Australian and New Zealand markets, trade restriction will reduce con- sumer welfare in PNG, particularly for low income earners. This must be balanced against the production effects. - 12 - These data probably underestimate both total consumption and domestic beef production due to exclusion of informal (village) slaughter which became increasingly important in the late 1970's. The domestic beef production data shown in Table 7 are derived from official slaughter data, Table 8, which does not include animals slaughtered for traditional ceremonies. Historically, the pig was the animal slaughtered on such occasions, but cattle (known as "big pigs") have assumed some of their role in the 1970's. The informal rura market was estimated to account for approximately 4,000 animals in 1976/77.1 The slaughter in informal markets is also important because the average price paid exceeds the abattoir price by about 50%, on average. Informal market purchases are for celebrations and exchanges where the prestige and satisfac- tion is directly dependent on the price paid. The income so earned also passes directly into the seller's hands, while sales to the abattoir are routed first to the PNGDB which determines how much will be kept for loan repayment. 47. Because of the high and rising informal market price during the mid-1970's, it is difficult to determine whether smallholder producers, particularly in the highlands, received relatively lower average prices for their output during this period. A higher proportion of smallholder subproj- ects in the highlands appear to have failed than those elsewhere.2' This probably occurred because conflicts over project land have been most severe here, and because project land is also suitable for coffee, whose profita- bility rose dramatically. E. Bank Performance Appraisal 48. The appraisal mission was very perceptive in its analysis of the existing situation in 1971 and it was innovative in designing institutional mechanisms to overcome the problems which the proposed project was likely to face. The audit believes, however, that the appraisal design was over- optimistic about the capacity of PNG to implement the system even as modestly designed. And insufficient emphasis was placed on project risks in the main appraisal report, as their discussion was confined to the annexes. 49. The project attempted to introduce new and only partially tested (within the PNG environment) technologies and commercial enterprises to largely uneducated farmers having little knowledge of the animals concerned, little concept of management systems, and little experience in business. The livestock units were heavily dependent on extension assistance at a 1/ G. Malynicz, P. Herington, and J. Eccles, "The Livestock Industry of Papua New Guinea", Department of Primary Industry, January 1, 1977, draft. 2/ Roger Nelson, "Current Situation in in the New Guinea Highlands," Veteri- nary Officer's Conference, June 1981. - 13 - time when PNG was moving rapidly toward independence, with attendant national euphoria and administrative change. DPI's senior staff was largely expatriate and it was already perceived that many of these staff would leave with in- dependence, and that it would be difficult to replace them with other staff of equal training. And the cattle units were expected to operate within a complex land tenure structure which usually necessitated multi-person owner- ship and management, thereby complicating substantially the transfer of technology and day-to-day management. Within this setting, the development of livestock units was likely to be very difficult. Regardless, given the perceived risks, the report ought to have emphasized the need to monitor key indicators, like the farmer/extension agent ratio, village tenure development and on-farm technical coefficients. Supervision 50. Because of thE substantial risks inherent in the project, Bank supervision should have been intense. Instead, supervision was light and from their reports most missions were insufficiently concerned with on-farm prob- lems. There were only eight supervision missions in eight years and the missions were usually brief; the first four missions also supervised other projects. There was little staff continuity; on only two missions was a member from a previous mission present. More emphasis was placed on lending issues, particularly the slower than expected disbursement, than on technical and managerial problems. On two occasions, Vn June 1976 and July 1977, the project was classified "problem-free (1)"' although on these occasions, smallholder management was bad and unit infrastructure was deteriorating, cattle were escaping, the feral animal campaign had begun, the farmer training program had nearly halted, and the extension service was declining in qual- ity.2/ In each case, the draft paragraph for the President's report included the assessment "Despite major changes within Papua New Guinea following self government in 1973 and independence in September 1975 the project has pro- ceeded quite well." 51. There was no supervision mission from August 1977 until November 1978 (16 months). This mission placed substantially more emphasis on project impact at farm level, and was much less sanguine. It stated that the genuine transfer of production technology and managerial skills to small farmers had been minimal, that farmers were almost totally dependent on DPI staff for technical, financial, and managerial decisions, and that there was little 1/ The SPN mission 1977 produced two Summary sheets, one dated August 2, 1977 gives a 1 rating, the second dated August 22, 1977 give a 2 (moderate problems) rating. 2/ The project was given a 1 rating, because it was not deemed necessary to go into these issues in detail since the Bank had already decided not to pursue support of this industry further and because senior DPI staff were aware of the problems and were as capable as Bank staff. The audit mission believes that these views ought to have been stated in super- vision reports. - 14 - prospect for farmers to become less dependent on DPI assistance in the near future. It suggested that a particularly thorough evaluation was needed if a follow-up project was contemplated. Even this mission however, assigned the project only a "moderate problem (2)" rating and the draft paragraph for the President's Report stated "Considering the major changes within PNG...., the Project has generally proceeded relatively well." 52. The major recommendation of the 1976, 1977, and 1978 supervision missions was that DPI and PNGDB should establish a system for continuous monitoring and evaluation of the project. In fact, there was substantial understanding of the problems faced by smallholders within DPI, and a number of useful studies were carried out with DPI staff cooperation which would have been useful for policy formulation. Thus, an effort to use the existing knowledge more effectively might have been more productive than an effort to establish a formal monitoring and evaluation system. The latter was needed, but the emphasis placed on it seems to have diverted attention away from the obvious problems the industry was already having, and postponed their reso- lution to the future. As it is, no clear policies resulted within the Bank or within PNG other than a deemphasis of smallholder cattle lending. F. Conclusions 53. The project was not successful in establishing as many profitable cattle subprojects as was estimated at appraisal, and project costs have been large relative to the direct benefits achieved. The project did allow nationals to gain experience in livestock management and in commercial enter- prise, and provided certain consumption benefits (including the prestige of owning and operating a cattle unit) which must be given some positive weight. It also hastened the transformation of traditional land rights, which must eventually occur, although it is unclear whether the project has done so in the most beneficial social manner. Although declining beef prices depressed cattle profitability, with detrimental effect on smallholder management, this is only one of the factors which has resulted in the poor performance of this subcomponent. The audit believes that other factors, principally the proliferation of uneconomically small subprojects, land tenure problems, the failure of the farmer training program, inadequate extension input, and an overestimation of farmers' abilities to absorb rapidly the relatively sophisticated project approach, were more important determinants of the project's outcome. 54. Beef cattle development probably has a considerable role to play in PNG, as was noted in the Bank's original analysis of sectoral priorities. The demand for beef substantially exceeds domestic production capabilities and is rising rapidly, with serious balance of payments implications. And, as the PCR states (para. 5.08), the smallholder cattle sector has had, and is likely to continue to have, far greater impact on rural development than any other livestock subsector. The PNGDB has deemphasized smallholder cattle lending since 1978 and the Bank decided at the project's close not to lend further for beef cattle development. The audit mission understands the hesitancy of the PNGDB and the Bank to encourage additional smallholder cattle deve- lopment, but believes assistance is urgently needed for the consolidation of the subprojects already started. - 15 - Table 1 PAPUA NEW GUINEA - CATTLE NUMBERS 1956 - 1978 Smallholder Ranch Govt./Mission TOTAL 1956 10,435 1959 15,123 1960 16,320 1961 20,247 1962 23,508 1963 691 ----26,429---- 27,120 1964 1,488 ----29,732---- 31,220 1965 1,919 ----33,263---- 35,182 1966 2,892 ----36,722---- 39,614 1967 3,635 24,961 16,026 44,622 1968 4,425 32,124 16,071 52,620 1969 5,870 37,108 17,750 60,728 1970 8,542 48,477 19,075 76,094 1971 11,817 51,662 21,263 84,742 1972 19,176 56,424 20,528 96,128 1973 27,346 58,542 21,057 106,945 1974 35,439 63,038 21,490 119,967 1975 42,451 e 65,340 19,145 126,936 e 1976 49,463 63,129 19,004 131,596 1977 n.a. n.a. n.a. n.a. 1978 p 48,004 ----81,798---- 129,802 1981 * 30,000 ----75,000---- 105,000 Source: Huntings Technical Services, Ltd., "Beef Cattle Ranching Development Study", Inception Report, April 1979, Table 2.1. Original data from DPI and PNG Bureau of Statistics, except as noted: e = estimated by Huntings, p = preliminary results of Smallholder cattle census, n.a. = not available, * = estimated by audit mission. - 16 - Table 2 PAPUA NEW GUINEA DEVELOP1ENT BANK SMALLHOLDER CATTLE LOANS 1967-69 TO 1977-78 Cumulative No. of Cumulative Amount Amount Loans # of Loans (K'000) (K'000) 1967 - 69 232 232 581 581 1970 - 71 342 574 509 1,090 1971 - 72 542 1,116 1,158 2,248 1972 - 73 463 1,579 1,093 3,341 1973 - 74 503 2,082 1,085 4,426 1974 - 75 434 2,516 956 5,382 1975 - 76 312 2,828 876 6,258 1976 - 77 263 3,091 570 6,828 1977 - 78 220 3,311 463 7,291 Source: P. N. G. Development Bank - 17 - Table 3 GROWTH OF PNG CATTLE STOCKS BY PERIOD (% of Annual Growth) Total Ranch Govt./Mission Smallholder 1964 - 67 13 11 ----- 34 1967 - 73 16 15 5 40 1973 - 76 7 3 (3) 22 1976 - 78 (1) ----- 0 (2) 1978 - 81 (6) ---- (3) (11) Source: From Table 1. - 18 - Table 4 ON-FARM TECHNICAL ACHIEVEMENTS Unit Appraisal Actual 1978 1981 Calving Rate % 60 52 45 Extraction Rate % 17 15 14 Cow Mortality Rate % 5 11 11 Carcass Weight, Steers kg 250 200 188 Slaughter Weight, Steers yrs 2.5 - 4 - 4.5 Source: Appraisal Report PA - 144a, Annex 6, Table 5, dated October 16, 1972, OPCR (para. 15) and DPI, Papua New Guinea. FINANCIAL DATA CONCERNING THE CATTLE PROJECTSa (in Kina) Initial Contributions Distribution of Proceeds (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Project Years Number Stocked Loan Loan Followers' PNGDB Total Repayments Repayments Revenue Revenue Net Income of Net Income of Bosses' Followers' Households Loan Gross Plus by Given to Kept bg Bossese Followersf ConLribULILLr 17C,c ini=2 F,ca"e V.11-wers Bosses Expenses 1 8 240 30 1 2427 2383 2323 230 0 290 50 -260 2 4 330 40 3 3690 2525 1850 68 20 723 393 -88 3 4 404 146 7 6010 4289 3180 195 15Sc 1149 745 -186 1 4 2 200 450 28 4449 1335 1510 90 16 -101 -301 -524 5 2 180 235 7 4320 1354 1280 0 14 60 -120 -221 6 2 284 436 8 3450 1000 930 0 90 -20 -304 -346 7 2 150 200 17 2798 464 280 100 0 284 134 -300 a As of November, 1977. b 'Incidental expenses' include costs for various services and supplies not covered by the loan. For the first three projects, K200 has been added for such expenses and K50 for the last four projects. These figures are approximations. c Not included in this total is the value of the two calves given to two followers. d Column 10 = (Columns 6 + 8) - (Columns 7 + 9). e Column 11 = Column 10 - Column 2. f Column 12 = Column 9 - (Columns 3 + 8). Source: Grossman, op. cit., Table 3.1. - 20 - Table 6 REAL CATTLE PRICES IN PNG Nominal 1/ Real 2/ Year Price Price (per kg) (100 = 1971) 1971 64 100 1972 (65) (94) 1973 (66) (91) 1974 68 73 1975 75 75 1976 76 70 1977 80 71 1978 82 70 1979 90 69 1980 95 69 1/ Price per kg cold carcass weight, prime steer, 1971 and 1974 are indicator prices. 2/ Nominal price deflated by urban cost of living index, with 1971 set equal to 100. Source: Huntings Technical Services, Ltd., op. cit, Table 3.4; National Statistics Office, and audit mission calculations. - 21 - Table 7 BEEF CONSUMPTION 1971 TO 1978 Domestic Production Imports TOTAL % of Total Imported Tonnes Tonnes Tonnes 1971 - 72 1,086 2,116 3,202 66% 1972 - 73 1,382 1,900 3,282 58% 1973 - 74 1,950 1,304 3,251, 4C 1974 - 75 2,145 1,804 3,949 46% 1975 - 76 2,600 2,430 5,030 48% 1976 - 77 2,740 2,700 (1) 5,400 50% 1978 (2) 2,460 6,064 8,524 71% (1) Estimate based on Australian and New Zealand export statistics. (2) 1978 imported based on first six months imports. Source: Huntings Technical Services, Ltd., oR. cit, Table 2.4. - 22 - Table 8 CATTLE SLAUGHTERINGS AT LICENSED PREMISES (1) 1971-72 TO 1977-78 Number of Cattle Year Smallholder Extensive Total % Extensive 1971 - 72 649 7,127 7,776 92% 1972 - 73 2,161 7,880 10,041 78% 1973 - 74 2,432 11,319 13,751 82% 1974 - 75 2,830 11,853 14,683 81% 1975 - 76 4,123' 13,096 17,219 76% 1976 - 77 7,134 (2) 11,133 18,267 61% 1977 - 78 n.a. n.a. 16,400 n.a. (1) Licensed premises include Abattoirs and Licensed Slaughter houses. Cattle killed for own consumption and 'sing sing' killings are not included. (2) Includes approximately 2,000 head of feral cattle. Source: Huntings Technical Service, Ltd., op. cit., Table 3.3. - 23 - IDA OVERVIEW OF PROJECT COMPLETION REPORT PAPUA NEW GUINEA SMALLHOLDER LIVESTOCK CREDIT PROJECT CREDIT 348-PAP May 6, 1981 East Asia and Pacific Projects Department Agricultural Division II h - 25 - PAPUA NEW GUINEA OVERVIEW OF PROJECT COMPLETION REPORT SMALLHOLDER LIVESTOCK CREDIT PROJECT (CREDIT 348-PAP) Introduction 1. The Borrower presented a final draft of its PCR to the Bank on March 24, 1981. It was a voluminous work touching on many aspects of the project, but was disorganized and repetitive. The quality of the coverage of the report was limited by the lack of available historical data on project progress which was due, in turn, to a lack of monitoring. Taking into account this limi)ation, we find the PCR satisfactory as an evaluation of the project. 2. On January 30, 1970, the Bank approved the Second Agriculture Development Project (Credit 175-PNG) for Papua New Guinea as part of that country's development effort to diversify exports (oil palm), expand pro- duction (coconut and rubber), and substitute imports (cattle). That project also attempted to involve more nationals as opposed to expatriates in the growing of tree crops. Regarding the cattle component, the project financed large expatriate and government ranchers with the objective of increasing the national herd and supplying heifers and bulls to smallholders under the following project - Smallholders Livestock Project. The cattle component was considered successful in design despite the fact that private ranch and herd development reached only 60% of appraisal estimates. Furthermore, Credit 175 established three training schools for farmers who were potential participants in the smallholder project. Government was responsible for project preparation. 3. The objective of the Smallholder Livestock Project (Cr. 348) was to help finance medium- and long-term investments to increase smallholder beef, pig and poultry production and to improve livestock extension education and research facilities in the Department of Primary Industries (DPI). DPI was responsible for project implementation in conjunction with the PNG Development Bank (PNGDB), which received IDA funds from Government for on-lending to farmers. Subloans to farmers were made on the basis of farm plans and budgets prepared by DPI extension staff, who also supervised their implementation. Project Implementation 4. Total project cost was US$9.8 million equivalent of which the cost of the project's sublending component was estimated to be US$5.2 million. Compared to the appraisal report, actual total value in current prices of subloans fell short by only 25% for beef cattle development and 6% for pig breeding and fattening. Actual total value of subloans for broiler production was seven times as much as the appraisal estimates. When - 26 - calculated in US dollars in 1972 prices, the year of the appraisal, actual amounts spent for subloans were only US$3.8 million as compared to US$5.2 million. See Table 1 for details. 5. The project directly benefited 1,572 farmers or farm groups compared to 1,260 farmers estimated at appraisal. There were considerably more subloans for broilers and beef, but fewer subloans for pig production than was anticipated during appraisal. In current prices, average subloans for beef production were about US$ 4,000 compared to US$5,500 estimated at appraisal; subloans for pig development were about US$1,200, as expected; and, subloans for poultry production were US$2,700 compared to US$2,000 at appraisal. In US dollars of 1972 prices, average subloans were US$3,000 for beef production, US$900 for pigs and US$2,000 for poultry. 6. The technical services of extension, research and training were estimated to cost US$3.2 million at appraisal with the IBRD credit proceeds financing the capital cost of US$400,000. Actual drawdown of the Bank Credit for financing technical services was US$456,000 of which US$29,000 was used by Government for preparation of a proposed second agricultural credit project, which provided a modest support for financing of pigs and poultry on large scale agroindustrial operations. Appraisal estimates show that Government was expected to contribute US$2.8 million equivalent to help finance the technical services component of the smallholder livestock project. It is very likely that these expenditures substantially exceeded the amount estimated at appraisal as the project required eight years rather than three and a half years to disburse, and technical services largely comprised current expenditures, which were financed by local funds. Beef Cattle 7. It is significant to note that the project financed nearly 20,000 head of breeding stock. Other investment items were fencing, pasture improve- ment, and animal health. The table below shows the percentage distribution of inputs in beef cattle production: Appraisal estimate Actual - percentage distribution of investment - Purchase of cattle 60 81 Fencing 30 15 Pasture improvement 3 1 Animal health - 1 Other 7 2 Total 100 100 - 27 - Pigs 8. In general, the pig production component of the project was unsuccessful even though over 200 subloans were made to small farmers (270 subloans were expected at appraisal), and 45 additional subloans were made to firms initiating medium-scale breeding and fattening operations adjacent to the urban centers. Under the "peri-urban" pig subloans, smallholders were financed for an enterprise with capacity to fatten 3-4 batches of about 20 weaner pigs each year. This activity was limited by the availability of weaner pigs and feed, and the size of the market. Poultry 9. The Broiler and Layer subcomponent of the project for smallholders as designed in the app:raisal report appeared to become obsolete soon after project effectiveness. The DPI decided that smallholders could not operate on their own due to the lack of availability of low-cost feed and day-old chicks and management expertise. However, as in other parts of the world, large integrated broiler firms which control feed supplies, breeder and hatchery operation, chick growing buildings and slaughter facilities have proven to be successfuL in the country. The integrated broiler operation of the Niugini Table birds has become highly successful, with a high level management of the financial and technical aspects of broiler production. The number of subloans made under the project was 2.5 times those compared to appraisal estimates for all poultry and the total value of actual subloans exceeded US$600,000, compared to US$48,000, in terms of 1972 prices, estimated at appraisal. Training 10. In general, the project investments in training of extension staff and farmers, mostly in cattle management practices, were completed and considered successful. There was, however, considerable delay in the con- structing and equipping of a livestock farmers' training center at Moitaka near Port Moresby due to lack of appropriate road access and administrative problems. There was strong farmer response to the training activity in the first half of the project but, subsequently, farmers viewed adversely the marginal benefit of additional training against the need to attend to their farms. Also, in 1976 there was a shift of emphasis from expansion to consolidation in the snallholder cattle industry, with fewer new farms being established. Research 11. Research under the project was of an applied nature, testing production systems for beef in different ecological zones. Weaknesses in the research component for beef not mentioned in the PCR were insufficient research effort devoted to improved pastures, minimal daily salt intake and supply thereof, and animal nutrition. The emphasis in pig and poultry work - 28 - was placed on nutrition and management aspects. Buildings were constructed as planned, and equipment provided for the beef cattle research unit at Erap and the poultry research center at Labu. Five vehicles were provided and three houses were constructed at Goroka, Urimo, and Moitaka. Virtually all of the research staff positions were filled. Research results were excellent for pigs, fair for cattle, but rather poor for poultry. Extension 12. A major investment was made in improving the physical infrastruc- ture, equipment, vehicles and quality of staff in the extension service under the project. The DPI was successful in meeting appraisal mission recommenda- tions as all 78 houses and 11 office stores for extension staff were construc- ted and all 35 four-wheel drive vehicles and 11 motorbikes were purchased. All staff, apart from two field supervision positions, were hired according to the design of the extension service at appraisal. Despite interruptions the PCR points out appraisal estimates were over optimistic as to what the extension department could accomplish under the project. However, the extension staff of DPI did take an active role in project implementation and gained a valuable experience while doing so. The agents performed reasonably well under the project given the fact that: (a) during the move to national independence, expatriate extension people were replaced to a great extent by locals, who were less qualified and less accepted by project farmers; (b) there occurred a shift of decision making from the Agricultural Provincial Officer to the central DPI offices in Port Moresby, which had the unfortunate effect of distorting the lines of communication between field agents and their supervisors; and (c) with the collapse of the world beef market in 1975-77 few farmers entered the project, thus, decreasing demand for extension agents services to new ranching operations. Agricultural Impact Beef Cattle 13. The project beneficiaries increased their beef cattle herds by 30,000 animals which was roughly one-half of the expectation set out in the appraisal report. They accomplished this in eight years, almost twice the time considered in the appraisal report. Apart from those explanations given in the Government's PCR (para. 5.03), Bank supervision missions also noted that pasture degeneration due to overgrazing, infertile bulls and marketing constraints were major problems in beef raising. The principal marketing difficulties were lack of transportation, high marketing margins to agents and poor geographic distribution of abbatoirs. - 29 - 14. The PCR does not emphasize that perhaps the most crucial blow to the smallholder beef cattLe industry was beyond its control. After a cattle price boom in 1972-74, the cattle industry suffered tremendous oversupplies as the world market virtually closed for frozen and chilled beef. The steep price decline staggered the beef producing sectors in all countries. Australia began a beef dumping operation in PNG which crippled the local industry, driving the market price still lower in real terms. In 1975, the DPI began a beef consolidation effort, trying to ensure that beef cattle clients under the project would maintain their herds on the farm until the market improved; they also discouraged smallholders to enter into beef production during the 1976-78 period. In 1978, beef market prices began to improve slowly and new subloans were resumed. Another development not emphasized in the PCR and which had unfavorable results for the beef component of the project located in the uplands was the explosive rise in coffee prices which began in 1974. Many project beneficiaries who initiated livestock activity under the project also had coffee trees, and when the coffee price rose, they tended to give less time to livestock raising and more time to coffee replanting, cultivation and harvesting. 15. Productivity coefficients for beef cattle are shown in the following table: Unit Appraisal Actual Calving rate % 60 52 Extraction rate % 17 15 Adult cow mortality rate % 5 11 Bull ratio - 1.15 1.76 Carcass weight /a kg 220 200 /a Dressed cold carcass weight. While the actual calving and extraction rates are acceptable under small- holder production, the adult cow mortality was the single most damaging indicator of inadequate animal management which contributed to relatively low rates of return on farmer investment. Bull performance was not adequate in the beginning of the project; of 1,700 smallholder cattle farms surveyed, 800 or close to one-half had no bulls in their herds. In 1979, DPI began raising performance-tested bulls for sale to smallholders at cost. Many smallholder livestock farmers could not benefit from the project's credit program because they were required to have at least 15 adult head of beef cattle. The overall average herd size financed by the PNGDB, including the project, is 14 with an average farm area of 26 ha. In 1977, the Borrower requested that farms with herds fewer than 15 adult animals be eligible for beef cattle financing as about 40% ol subloans made by PNGDB for beef cattle were - 30 - otherwise excluded from project financing. The supervision mission of August 1977 supported this request, but there was no follow-up by the Bank. Pigs 16. Pig production under the project has not been successful either at the smallholder level or at the commercial farm level. The productivity of pigs in the village situation is low, especially in highland areas. The principal reasons given in the PCR for poor project performance on pig development were: (a) inadequate extension service to farmers producing pigs on a commercial scale; (b) unavailability of protein concentrates; (c) inadequate feed mixing of sweet potatoes with concentrates; (d) high mortality rate due to disease; and (e) relatively high cost of animal feeds. While findings in the Bank supervision reports support most of these explana- tions for the pig component failure, they did point out that, with good management, commercial pig production could have been a relatively high cost of animal feeds. While findings in the Bank supervision reports support most of these explanations for the pig component failure, they did point out that, with good management, commercial pig production could have been a relatively profitable enterprise. Also, the Bank reports stress the disorderly marketing system which confronted smallholder pig producers, and marketing arrangements for pigs from the traditional rural sector was unstable and somewhat disorderly. Future efforts at commercial pig development will have to devote special attention to adequate feed supplies, disease problems and marketing. Poultry 17. The project impact on national broiler production was significant. As of March 1981, there were 132 poultry units of 1,000 birds each; these units produced 1,016 metric tons of poultry meat annually, or about 13% of the total market. The major factor in the relative success of the broiler production is due to the large scale integrated approach which is being adopted worldwide. For example. the largest intpgratred onerition. i.e. Nincrini Table Birds. depends on the smallholder broiler farmers for 9rowine the birds while the comDanv orovides all the malor innilts including finance. day-old chicks. feed, transport, a high level management supervision. and a auaranteed market with price based on costs of nroduction. - 31 - 18. Until 1979 Niugini Table Birds relied 100% on the smallholder for their supply of poultry for processing, but is now beginning to shift birds to their own growing sheds on a larger scale. One of the reasons for growing their own birds is the relatively poor productivity performance of the smallholder as shown in the following table: Average for DPI smallholders Lubu South Tegels Ingham Unit PNG PNG Africa Singapore Sydney Brisban Age in days days 63 63 56 63 56 52 at slaughter Average body kg. 1.69 2.09 2.24 2.10 2.16 1.75 weight Feed conversion ratio 2.64 2.36 2.16 2.30 2.10 2.10 Mortality % 12.40 7.78 5.41 3.40 5.00 4.50 These figures indicate that the performance of smallholders in PNG compared to smallholders in other countries is deficient. Specific factors contributing to the smallholders' problems were considered to be: (a) wrong or inadequate feeding; (b) poor feed storage leading to quality deterioration; (c) inadequate drinking water; (d) inadequate housing, causing disease problems through wet litter; and (e) losses through stealing and predators. Rates of Return 19. Although an overall economic rate of return was calculated in the PCR, prepared for the Government by a consultant, insufficient data was found in the report to check it. The financial data on the beef cattle model was complete and was calculated according to Bank standards. Communications were sent to Government requesting further information and worksheets for the rate of return calculations, but these have not been adequate to recalculate the ERR. In any case, since there was no monitoring of the project despite repeated suggestions by Bank supervision missions and a Loan Covenant requiring it, serious gaps exist in available information; the bank should accept the results in the country PCR and focus in the future on improving - 32 - project monitoring. The rate of return for the project and its components are shown below: Appraisal Report Estimate Actual FRR ERR FRR ERR Overall ERR - 14 - 2 Beef cattle 11 16 10 2 Pigs 25 25 - - Broilers 69 33 44 - Discussions with the consultant who prepared the PCR confirm that inadequate data was available for calculating a reliable ERR. Furthermore, economic factors such as use of border prices for pricing project outputs and inputs and shadow pricing of labor and foreign exchange was not used in estimating the ERR, which would lead to the conclusion that the ERR of the project was probably substantially higher than the 2% ERR derived by the Project Unit, but the considerable gaps in information preclude the presentation of an alternative rate of return. Under the circumstances, the ERR of 2% pre- sented in the PCR can only be viewed with considerable reservation. Institutional Performance and Development 20. PNGDB's overall performance was satisfactory during project disbursement. In December 1979, the staff numbered 381 of which 38% were stationed at the head office. The PNGDB has a healthy financial condition, especially when due account is taken of its special development role, which entails higher than average lending risks and administration costs (11% of average loan portfolio in 1979), both connected with rapid increase in staff and training costs, and lending to traditional, agricultural producers scattered throughout the country. In July 1979, the project financed a consultant to undertake a review of PNGDP's portfolio and procedures, to determine its strengths and weaknesses, and to recommend improvements in the institution for lending for agricultural development. Under the project, there were only 133 subloans written off as bad debts which amounted to US$130,000 equivalent in current prices or 2.2% of total amount sublent under the project. Arrears greater than six months amounted to US$530,000 equivalent or 14% of total subloans. One of the curious features of these relatively low levels of default and arrears was the unique action on the part of PNGDB and DPI to round up cattle owned by defaulting borrowers, and to sell than on the market. These receipts were used to liquidate the outstanding debts of smallholders and any surplus was returned to them. - 33 - Bank Performance 21. The PCR does not include any direct criticism of Bank performance in the project. However, there are a number of points related to Bank staff's judgement at appraisal and supervisions which deserve comment. (a) World Bank projections of beef prices on the world market were much too optimistic in 1972 when project appraisal. was undertaken. 'Farmers found beef and pig raising profitable until the decline of the market in 1975-77. Interest on the part of farmers to monitor their cattle herds declined and few new farmers began cattle operations during those years. (b) Shortly after project effectiveness, Government decided that broiler and egg production at the smallholder level would not work. The appraisal design for poultry develop- ment was not appropriate under the circumstances of unsure supplies of day-old chicks and feed. It is to the credit of the Bank that flexibility was shown and an understanding was reached whereby layer production was not to be pursued and broiler production would develop only on the basis of an integrated operation, whereby the integrator supplied chicks, feed and management to smallholders for growing birds for the firm. This approach proved successful under the project. (c) Evidently, Government was never really convinced on the need for project nonitoring and regular audits despite Loan Covenants requiring these measures and repeated insistence by Bank staff to begin work on these basic management tools. Nevertheless, Bank pressure on Government did result in the preparation of a study, "Procedures Manual for Smallholder Loans", and the National Smallholder Cattle Census which was carried out once during project disbursement. At no stage did the Bank suggest that noncompliance with covenants would result in the suspension of disbursements. (d) Advice on applied research for livestock given by Bank super- vision missions in the first three years of project execution was not sufficiently followed up by subsequent missions. More attention should probably have been devoted to the nutritional aspects of cattle development. Poultry research was not well done. (e) It would appear that this project could have profited from more Bank supervision. In eight years there were only eight supervision missions and all of the missions devoted only part of their time in PNG to the project, which was undertaken in conjunction with the supervision of other ongoing projects. - 34 - Conclusions 22. Measured by the expectations of the appraisal report, project performance is disappointing. It is disappointing not just because the production and productivity indicators fell short of expectations but also because the project design in the case of pig and poultry farming was not appropriate and the period in which the project was to be implemented proved to be over optimistic. Hope remains for further beef cattle raising by smallholders, but the success factors here are not at all clear and, certainly, inadequate at this time to permit the design of a follow-up project solely for beef production. Why such disappointing results of this project? 23. The answers are complex and difficult to assess. There is in PNG a traditional system by which smallholders raise pigs and chickens; few small farmers have beef cattle, except those adjacent to the large expatriate ranches. Management of stock is rudimentary and few commercial inputs are used, although a relatively but steady level of output was achieved. A complex of social values and attitudes influence the economic behavior of the smallholder. On the other hand, the project intended to replace the traditional system of animal husbandry by a commercially oriented system, which was based on economical inputs, efficient feeding practices and feed mixes. Seemingly, those elements which maintained production, although probably not very efficiently, in the traditional system were not transferred to the project system, nor were there significant technology transfers by extension agents to small farmers. 24. Project farmers were not experienced in making their own financial and technical decisions and, consequently the extension workers undertook the role of the decision-maker at the farm level, reducing the farmer to little more than workers. There was considerable success in the beginning of the project when the Provincial Agricultural Officers were the key project personnel and a larger number of expatriates staffed the extension service; farmer expectations were high. By 1975, project performance began to deteriorate. 25. While it is clear from the PCR that the smallholder pig, layer and broiler components of the project were not successful in terms of goals set by the appraisal mission, the report points out considerable success of a large-scale, integrated broiler enterprise. However, it would appear that smallholder pig and chicken enterprises are not effective vehicles for rural development in PNG. The beef cattle component has not been a failure. The project helped establish 15% of the current national cattle herd; the 1,028 smallholder cattle farms developed under the project have had a favorable impact on rural development and the project appears to have provided a reference point for livestock development and rural credit in Papua New Guinea. However, it remains to be seen whether the smallholder can acquire management skills needed to achieve productivity gains in order to - 35 - make beef raising a more attractive activity. Marketing disorders and instability require a major effort by Government to provide credit for improved transportation, abbatoirs and orderly market outlets for beef. 26. Bank staff disagree with some positions taken in the PCR as follows: (a) Insufficient emphasis is given to the serious decline in the world beef market during the 1975-77. This factor not only caused lack of markets and low prices for beef, but also adversely affected the demand and price of pork as well. The reason why beef ranching became less attractive in PNG was the overall effect of world beef prices and the importation (dumping) of cheap beef to the country. (b) The PCR shows a lack of appreciation of market availability and marketing constraints. While the problem is mentioned in the PCR, it is not adequately discussed. The lowest cost area for cattle raising is in the mountain areas, but the major markets are on the coast. The problem of transportation of beef either on foot, truck or by plane is critical in the country, and has yet to be resolved. Further, the poor distribution of abattoirs in the cattle raising areas prohibits efficient slaughter flows to the urban centers. (c) The PCR overEmphasizes the importance of expatriates in the project with particular reference to the phasing out of expatriate extension stff during the move to independence. No doubt the extension to smallholders was important to the successful operation of livestock enterprises, but the extension system did not collapse with the withdrawal of expatriates. Lessons Learned 27. One of the lessons learned under this project is the need for expert technical and financial management for pig and poultry production, especially if production is to be expanded on a national scale. In the short run, it is probably wise to concentrate the production of pigs and poultry in integrated firms which can achieve economies of size. However, in the long run, it is necessary for Government to assist the smallholder in transforming pig production to a conmercially-oriented system. It is not clear how this should be accomplished,, but serious effort should be made by Government and the Bank in developing a meaningful dialogue on this issue. 2R. A continuing issue which has, evidently, not been resolved with this project is the role of DPI and PNGDB at the field level. Clarification of the roles is required. The extension agents should not be responsible for subloan collections as it seriously erodes their credibility and accept- ability with the farmers with whom they work. Perhaps they should not do - 36 - subloans appraisals nor supervisions for the PNGDB either as this dilutes their major purpose which is to extent agricultural and livestock improved practices to the farmer. 29. This project introduced modern cattle raising practices to over 1,000 smallholders. Despite the adaptation difficulties which many small farmers experienced in raising cattle, almost 90% of project beneficiaries who received subloans for cattle raising remained in the activity at the Closing Date. Increased efforts are required by Government to consolidate the gains reached under the project, through steps aimed at improving the design of future projects in cattle development. The minimum range of herd size for smallholders needs further study before a second beef cattle loan. There may be scope for reduced herd size if small cattle ranches also have crop output, which can be fed directly or its byproducts fed indirectly to cattle. PAPUA NEW GUINEA SMALLHOLDER LIVESTOCK CREDIT PROJECT - CREDIT 348-PAP Project Completion Report Project Implementation: Number of Farmer Beneficiaries, Total Value of Subloans and Technical Services and Average Value of Subloans, Appraisal Estimates Compared to Actual Results Number of farmer Total value of subloans henpficiAries and technical services Average Value of subloans Appraisal Actual Appraisal Actual /a Actual Appraisal Actual /a Actual (In US$ 1972 prices) (US$ in (In US$ 1972 prices) (US$ in current prices) current prices) Beef cattle 870 1,028 4,804,000 3,060,000 4,070,000 5,522 2,977 3,959 Pig breeding/fattening 270 247 319,000 226,000 300,000 1,181 915 1,214 Broiler production 120 307 118,000 620,000 825,000 983 2,020 2,688 - Technical services /c - - 400,000 343,000 456,000/b - - - Total 1,260 1,572 5,641,000 4,249,000 5,641,000 4,477 2,703 3,588 /a Australian dollars and PNG Kina were changed into US dollars at the average exchange rate (KI = US$0.72) for the period of project disbursement. In order to arrive at US dollars in 1972 prices, current dollars were deflated by the US Wholesale Price Index. /b Of the US$456,000 used for technical services, US$29,000 was used by Government for preparation of a second agricultural credit project, which also contains a minor portion of the proposed IBRD Credit for financing poultry and pig producers on large-scale farms. /c These are the proceeds of the IBRD Credit only. Appraisal estimates show :hat Government was to contribute M US$2.8 million equivalent to help finance the Technical Services component of the project. Because there was little record keeping during project disbursement, it is impossible to determine Government's other outlays for technical services expenditures. Presumably these expenditures substantially exceeded the amount estimated at appraisal because the project required eight years rather than three and a half years to disburse. Technical Services were largely comprised of current expenditures, financed by Government.  - 39 - PROJECT COMPLETION REPORT PAPUA NEW GUINEA SHALLIZOLDER LIVESTOCK CREDIT PROJECT CREDIT 348-PAP This completion report was prepared by the Borrower with substantial consultant assistance. Revised Edition March 1981  - 41 - PAPUA NEW GUINEA SMALLHOLDER LIVFSTOCK CREDIT PROJECT COMPLLTION REPORT SUMMARY AND CONCLUSIONS 0.01 This study is the Completion Report for credit 348 for the development of smallholder beef', pig, and poultry enterprises in Papua New Guinea. 0.02 The Agreement was signed in January 1973 for IDA credit of US$5 million for 4 years. This was considered to be the foreign exchange component of a total project cost of USS9.8 million. The credit was interest-free and carried a service charge of 3/4%. 0.03 US$4.6 million was to provide 90% of finance for 870 smallhok-'3r beef farms, 270 smallholder pig farms and 120 smallholder poultry farms. Average farm sizes of 15 breeders for cattle, 2 sows for pigs, anc for poultry, 600 broilers per annum and 100 layers -were envisL, =- o appraisal. 0.04 US$4.6 million for smallholder finance was made available to the PNGDB under a Subsidiary Loan Agreement with the Government. US$U.0 million was a loan at 1 1/% interest per annum, and US$600,000 a capital grant. Smallholder farms were financed through the PNGDB Supervised Rural Credit System. The remaining US$400,000 was for technical services, buildings, vehicles and equipment to be funded directly by IDA. 0.05 At appraisal, the project's economic rate of return, including technical services, was calculated to be 14%. 0.06 At completion in March 1980, 3 1/4 years after the agreed completion date of 31st December, 1976, total disbursement was US$4,14LL,000. 0.07 1,028 smallholder cattle farms averaging nine breeders were established under the project. 900 of these farms are still functioning and the project herd impact of 31,000 beasts (including slaughterings) over the project period of' 8 years was KX.7 million. This is 48% of the appraisal projection of 64,000 beasts (including slaughterings) over 1 years valued at K10.1 million. 0.08 247 smallholder pig farms were established, 202 were for breeding/fattening units similar to those envisaged at appraisal, and 35 were peri-urban 20 pig pork fattening units. Three peri-urban fattening units are still operating. It is doubtful if any of the 202 village breeding/fattening units still operate on a commercial basis. 0.09 Three hundred and seven 1,000 bird broiler units were developed under the project. The smaller broiler and layer units envisaged at appraisal were found impractical at project commencement, and none were financed. Of the 307 units developed, 132 are still operating, with production in 1979 of 1,016 tonnes of poultry meat valued at Kl.01 million. These projects are managed by a nucleus estate company. - 42 - 0.10 Animal Production Research Services and Farmer Training expansion requirements were met, and facilities fulfilled the functions outlined at appraisal. 0.11 The project's economic rate of return including extension costs and technical services for the 8 years of project development, was 2% over 20 years. 0.12 Credit 348 failed to achieve appraisal projections in terms of disbursement period, cattle and pig development and economic performance. The major contraints to project performance were problems associated with extension, management and poor financial incentives for farmers. 0.13 The project established 15% of the current national cattle herd, and a viable smallholder broiler industry in Lae as a model for future urban smallholder livestock development. The 1,028 smallholder cattle farms developed have had a significant impact on rural development and the project has provided a reference point for livestock development and rural credit in Papua New Guinea. - 43 - PAPUA NEW GUINEA SMALLHOLDER LIVESTOCK CREDIT PROJECT COMPLETION REPORT INTRODUCTION 1.01 Papua New Guinea has a population of three million people occupying a land area of 475,365 sq km. The country is mountainous and there are a large number of islands to the east and north in the Solomon and Bismark Seas. Rainfall is abundant except in the Central Province. 1.any of the soils are fertile and the country is endowed with good fish and forest resources. 1.02 The population is Melanesian, rural, and growing rapidly. There are some 700 languages. Two lingua francas, Melanesian Pidgin and Motu, are widely spoken, and English is used in schools and Government. 1.03 The country became independent from Australia in 1975 in a smooth transfer of power. The Government is a Western-style democracy and the current Prime Minister, the second since Independence, is Sir Julius Chan. 1.04 At appraisal Papua New Guinea's combined gross domestic product was calculated to be AS5117 million and had been growing at the rate of 10% p.a. for 5 years. At completion in 1979, PNG's gross domestic product was estimated to be Kl,h60 million at current prices. This represented an average increase of 8% in current price terms, but a decrease of .01% at constant prices since 1975. 1.05 The major elements of production in the economy are agricultural: tree crops for export (mainly coffee, cocoa and copra), and copper produced from one open cut mine. Of the gross domestic product, exports generally account for bO%, manufacturing services and commercial activities 20%, subsistence activities 10%, and Government activities 30%. 3,4% of total Government receipts in 1978 was provided as Australian aid. The per capita domestic product was estimated to be K350 in 1976. 1.06 The politically stable, and economically important Government, exerts a strong influence on development. The Government's basic strategy is to develop the agricultural base of the economy, emphasising rural development, through self-employment and small-scale units of production. BACKGROUND 2.01 Since 1963, World Bank Economic Missions have encouraged beef development, and in 1970, IDA approved the allocation of US$1.7 million, as part of *a USS5.0 million credit (PNG 175) to help finance the development and stocking of 150,000 acres of large cattle ranches and the building of educational facilities to train livestock producers. -44- 2.02 The remainder of the credit went to further expansion of smallholder oil palm settlements, and for planning and replanting existing coconut estates. This was the second IDA credit to Papua New Guinea for Agriculture, the first being a US$1.5 million credit for an oil palm settlement scheme in January 1969. 2.03 In 1971 the Government of *Papua New Guinea, through the Australian Government, requested IDA credit to help finance the establishment of smallholder beef farms, and. pig and poultry enterprises, as well as assisting in providing the extension, training and research services required for this development. 2.01 In 1971 meat and egg imports cost the PNG Government A$16.2 million. PNG was 2% self-sufficient in beef, 20% self-sufficient in pig meat, 38% self-sufficient in poultry meat, and 85% self-sufficient in eggs. The major objective of the smallholder livestock project was to provide the basis for rural development, and meat and egg import replacement through the smallholder beef, poultry, and pig industries. PROJECT FORMULATION 3.01 The smallholder livestock credit project was proposed by the Government in 1971 as the next stage of livestock development in Papua New Guinea: that is, the establishment of Papua New Guinea-owned smallholder beef farms stocked with the turnoff from the large-scale ranch sector, expanded under credit 175. Minor project components were to be smallholder pig and poultry development, and the provision of expanded extension training and research services. 3.02 Table 1 gives the farm numbers and loan funds required under the Government's proposed project. Table 1 Proposed Project Farms and Loan Funds Enterprise No. of farms Average loan (AS) Total lending (A$) Beef cattle 2000 3500 7,000,000 Pigs 500 77h 387,000 Poultry - eggs 98 1724 169,000 - broilers 152 981 149,000 TOTAL 2750 7,705,000 Average beef farms were to carry 20 breeders, pig farms 2 sows, egr farms 100 layers, and broiler farms 600 birds per year. 1.03 A detailed project appraisal was conducted by the World Bank in February 1972 and a credit project agreement based on the appraisal recommeodations was signed in January 1973. 45 - 3.0h Table 2 gives the farm numbers and loan funds required under the credit 348 project agreement. Table 2 Credii: 3L8 Proiect Farms and Loan Funds Enterprise Nc. of farms- Average loan (A$) Total lending (A$) Beef cattle 870 200 3,65q,000 Pigs 270 890 240,300 Poultry - eggs 60 890 53,1100 - broilers 60 610 36,600 TOTAL 1260 A$3,9',300 Average farm sizes were to be: cattle 15 breeders, pigs 2 sows, egg 100 layers, and broilers 600 birds per year. BASIC PROJECT DESCRIPTION 4.01 IDA was to provide a loan of SUS$5 million ($A4.? million) to the Papua New Guinea Government. This was to be interest-free and carry a service fee of 3/% per annum. Repayments were to commence in 1983 and be completed by 2022. 4.02 Of this SUSS5 million, $4,200,C00 was to be re-lent to the PNGDB at a rate of 1 1/4; per annum. Another USS600,000 was to be given to the PNGDB as a capital grant. US$4,218,000 was to be for sub-loans for smallholder beef farms, US$297,000 for sub-loans for smallholder pig farms, and US$103,000 for smallholder poultry farms. The remaining US$OO,000 was to be for technical services: building, vehicles and equipment, to be funded directly by IDA. 11.03 The project was to be developed over 4 years with the 31st of December 1976 as the closing date. DETAILED PROJECT PERFORMANCE Beef (Annex 1) 5.01 The aim of category 1 of the project was to develop over 4 years 870 smallholder beef farms each stocked with 1 bull, 15 heifers and 9 yearling steers on 50 hectares of open grassland. The total cost of developing these farms was to be K)A,003,000 with a loan component of K3,124,400 or K3,591 per farm. At appraisal the economic rate of return for the beef cattle breeding/fattening enterprise was projected to be 161. - 46 - 5.02 At completion in March 1980, 1,028 smallholder farms with average stock numbers of 1 bull, 9 heifers and 8 steers on 21 hectares had been established over 8 years. 900 of these farms were still operational. The average sub-loan per beef farmer was K2,822 Tvr a total loan component of K2,901,244. The economic rate of aturn of beef development over 20 years, including extension eests at 90% of total prmjeat extension costs for 8 years, was calculated at 2%. 5.03 The major factors in the poor performance of the smallholder beef sector were associated with traditional attitudes towards land ownership, the poor financial returns relative to expectations, a heavy dependence on Government extension input, and an overestimation of the subsistence farmer's management ability. These problems are manifested in the poor technical performance of many smallholder farms, the slowing in the growth of the smallholder sector, and the lack of farm maintenance and stocK control. 5.04 Actual calving percentages achieved on smallholder beef farms was 52% commpared with appraisal projections of 60%. Actual breeder mortalities were 10% - 12%, compared with 5% projected at appraisal. 5.05 The National Smallholder Cattle Census in 1978 indicated that 25% of the smallholder herd of 46,000 beasts was not effectively controlled. 5.06 At completion in 1980, the project herd impact of 31,000 beasts (including slaughterings over the project period) was K4.7 million, that is 48% of the appraisal projection herd impact of of 64,000 beasts (including slaughter ings), valued at K10.1 million. 5.07 However there were 1,028 smallholder cattle farms established under the project of which 900 are estimated to be still functioning. The 1980 stock numbers on these projects of 19,200 beasts represents 41% of the total smallholder, and 15% of the total, herd. Studies in the Markham Valley indicate that in favoured areas, well-managed smallholder cattle projects can achieve financial rates of return in excess of 20%. 5.08 The smallholder beef industry performance was far short of the optimistic appraisal projections, and will not lead to national self-sufficiency. The industry is stagnating and has many problems, but has has a far greater impact on rural development than any other sector of the livestock industry is ever likely to have. Pigs (Annex 2) 5.09 Under category 2 of the project, 270 pig farms were to be developed over 3 years. Finance was to be provided for the purchase of 2 to 3 crossbred sows, two fattening pigs and the construction of cheap pig shelters made of local materials which would last 8 years. The average sub-loan would be about K733. Total loan funds for pig development would be K198,000. The economic rate of return for the projected pig development was 25%. 5.10 At completion in 1980, 247 pig suu-loan-s averaging K874 for a total of K215,898, had been financed. 202 of these sub-loans were for breeding/fattening units similar to those envisaged at appraisal, and 35 were for peri-urban pork fattening projects. These were based on a - 47 - system of smallholders fattening 3.3 batches of 20 porkers per annum for sale to the urban markets. Loan funds averaging Kl,100 per farm were for housing, initial batch of pigs, and feed. Between 1976 and 1978, twenty-five of these projects were established near Madang and 10 at Situm near Lae. Only 3 peri-urban fattening units (2 at Madang and 1 at Lae) are still operating. It is doubtful if any of the 202 village breeding units are still operatzng on a commercial basis. 5.11 The major factors contributing to the failure of the village breeding projects were inadequate feeding, the high cost of protein concentrate and high mortalities. These factors are inter-related. Protein costs doubled in the first eighteen months of the project and the organisation and work involved in feeding .2L'kg of protein concentrate and 8kgs of sweet potato daily for the return achieved became unacceptable to the villager. The DPI extension effort required to maintain management and feeding standards was not forthcoming, as resources were scarce and returns considered doubtful. The financial rate of return for the appraisal model witi 197L costs and return was 21% over 8 years. 5.12 The peri-urban fattening units collapsed because of the low profit margins, high prI.ce and performance sensitivities, and lack of management skills required for this type of intensive production. Return to investment of 16% in Madang and less than 2% in Lae were achieved. 16% may appear to be reasonable, but it includes no investment for labour, management or housing. 5.13 Pig development under credit 3L8 failed. The consultant considers that, because of the labour and management requirements for commercial pig production, smallholder piggeries are likely to succeed -only where nucleus estates provide all inputs including management, financial supervision and marketing. Poultry (Annex 3) 5.1,u Category 3 of the project was for the establishment of sixty 100 bird layer projects and sixty 50 bird/batch broiler projects over 3 years. Average sub-loan for layers would be about K890, and for broilers K610. This gives a total expenditure under category 3 of K90,000. 5.15 At completiun in March 1980, 307 poultry loans had been granted for a total of K594,344 with an average sub-loan of Kl,936. Of the 307 units developed, 132 1,000 bird broiler units were still operating, producing 1,016 tonnes of poultry meat of which the annual value was K1,005,840. These 1,000 bird units were all at Lae and managed on a nucleus estate principal with the parent company, Niugini Tablebirds, providing all inputs including management, supervision and marketing. 5.16 The original 100 bird layer and 50 bird/batch broiler projects were found to be impossible to develop and supervise in the PNG village environment. The relative success of the 1,000 bird peri-urban broiler projects at Lae is the result of energetic promotion by the DPI Poultry Extension Advisor in the early stages of the project, followed by close commercially-oriented supervision by Niugini Tablebirds, and the establishment of a poultry import quota system in 1975. - 48- 5.17 It is considered "nat smallholder poiltry development is only likely to succeed if it is developed on a fully commercial nucleus estate system. Training (Annex 4) 5.18 In 1972, Papua New Guinea had two DASF Agricultural Training Institutes: Vudal Agricultural College with a three year diploma course, and Popondetta Agricultural Training Institute with a two year certificate course. These schools had an annual output of about 40 and 55 students respectively. DASF also had 22 Farmers Training Centres, 11 specialist Livestock Farmers Training Institutes, and one small Animal Production In-service Training Centre at Lae. 5.19 Under the project, the Government agreed .to: - construct, equip, and staff a Livestock Farmers Training Centre at Moitaka (Luanakalana) near Port Moresby, with a capacity of 48 places to train farmers and junior DPI (DASF) officers - continue the livestock courses currently given at Baiyer River (72 places), Erap (L8 places) and Urimo (LI8 places). - contruct, equip and staff an In-service Training Centre of 30 places at Lae to train Diploma and Certificate holders of the staff of DPI - employ one additional staff member (of assistant Livestock Officer level) at each -of the four Regional In-service Training Centres at Kapogere, Mutzing, Korefeigu and Sohar.a. 5.20 The building at the In-service Training Centre at Lae was completed on time, but the Luanakalana Farmers Training Centre was not completed until 1978. No training courses were ever held at the Luanakalana Farmers Training School. The In-service Training Centre at Lae was fully utilized throughout the project period, with 45 courses of an average duration of 3-4 weeks for a total of 612 attendants. At the other three Livestock Farmers. Training Institutes and Bisianumu which was used whilst Luanakalana was being built, 68 courses for 1,737 attendants were held. The costs of Livestock Farmers Training excluding air fares and travel was estimated at K85 per student in 1976, whilst cost per student at the In-service Training Centre at Lae was estimated to be K586 per student in 1979. 5.21 Farmer training declined from 1977 because of the increasing unwillingness of farmers to stay away from home, disillusionment with cattle farming, and shortages of provincial travel funds for farmers. In an effort to revive interest in Livestock Farmers Training, training staff have conducted 16 District Farmers Training courses in seven provinces over the last two years. These courses are for established farmers with an emphasis on farm demonstrations and are conducted at Provincial request. 5.22 The waning interest in farmers training, and relatively poor performance of the livestock industry indicate that the training programme has not been very successful. The effective transfer of knowledge and skills - 49 - has been much less than was anticipated and has been adversely affected by rapid localization and reduced interest and resources. 5.23 At completion there were four Agricultural Colleges, a national Fisheries College and an Agricultural Faculty at the University of Papua New Guinea. During the life of the credit project, 963 students graduated at certificate level, .309 at diploma level and 69 at degree level from these i.nstitutions. Of these graduates, 65 at certificate level and 2 at degree level were -women. Animal Production Research Service (Annex 5) 5.24 At appraisal in 1972, Animal Production Research on cattle, pigs and poultry was being undertaken by five-graduates with facilities on eight Livestock Stations,, totalling over 18,000 hectares. Chemical analysis was done in the Central Veterinary Laboratories in Port Moresby. The Research and Surveys Division of the Agricultural Branch also had two graduates doing pasture evaluation work in the field. Field Research staff were administ%ratively responsible to the Regional Controller, and technically responsible to the Chief Animal Production Officer of the Livestock Branch who was in charge of the Animal Production Research Programme. 5.25 The Veterinary Research Programme was adequate for Animal Health needs. However, more production research was required in the areas of smallholder management and husbandry, testing systems being advocated, and developing improved systems for the future. 5.26 Under the project the Government agreed to increase Animal Production and Research professional and management staff to 22 (15 for the five Beef Research Stations, b for the Pig Research Station, and 3 for the Poultry Research Station), construct new poultry research laboratories and provide housing and equipment for the expanded staff at all locations. Research was to be of a strictly applied nature, emphasising testing production systems for smallholders in different ecological zones for beef, and situational and management aspects for improved smallholder production with pigs-and poultry. 5.27 By the 31st of December 1976, the project requirements for the expansion of research facilities had been fulfilled at a cost of K170,000. This included a Beef Research Unit at Erap, a new Poultry Research Centre at Labu, 5 vehicles and 3 houses. All but three of the research staff positions were filled. These were Senior Veterinary Officer (cattle) at Erap, Animal Production Officer (cattle) at Moitaka, and Senior Veterinary Officer (pigs) at Goroka. 5.28 Major areas of cattle research during the project were in indentifying mineral deficiencies, quality performance, and adaption trials and monitoring, pasture improvement research and legume seed production. 5.29 Pig research in the early years of the project concentrated on breed and feed performance comparision under smallholder conditions. The emphasis from 1977 has been on distribution of crossbred pigs and mixed feed from the Tropical Pig Breeding and Research Centre at Goroka and a study of disease, production and husbandry techniques in the Southern Highlands. - 50 - 5.30 Poultry research on feed requirements and quality, bird management techniques and performance in the smallholder environment has been conducted at Labu Poultry Research Centre, Lae. The Centre has continuing programmes of feed quality monitoring, breeding and distribution of poultry, and pig an'd poultry feed analysis. A National Poultry Hatchery was established at Erap in 1978 to supply Papua New Guinea's requirements for broiler and layer breeding stock. 5.31 The Animal Production Section of DPI Livestock has applied research programmes for buffaloes, bees, sheep, deer and goats. Research is hampered by difficulties in obtaining and keeping qualified staff, and limited transport and communications. Economic Rate of Return 5.32 The economic rate of return for smallholder beef development under credit 348 was 2% with extension cost! incluced, compared with lo projected at appraisal. As no pig industry was sustained, there was no return. 25% had been projected at appraisal for pigs. 'ITe economic raUe of return for poultry over 8 years, was 7M with extension costs included, compared with 33$ projectcf at app-aisal witn no extension costs included. 5.33 The project economic rate of return over 20 years including technical services and extension costs for the 8 years, was 2% compared with 14% projected at appraisal with no extension costs included. INSTITUTIONAL PERFORMANCE AND DEVELOPMENT Agricultural Extension and Support Services (Annex 6) 6.01 At appraisal, DASF (DPI) had 1,145 extension generalists (Rural Development Officers) and 229 Animal Industry field staff. It was considered that Rural Development Officers spent 20% of their time on livestock extension while animal industry staff spent 100% of their time on animal husbandry and disease control. DASF extension officers acted as agents for the Papua New Guinea Development Bank, and Agricultural Credit was a major tool in extension work. Total DASF staff in 1971 numbered 3,586. Of the 575 first division staff (Assistant Directors down to grade 1 diploma-level officers) only 92 were nationals, the rest being expatriates, mainly Australians. 6.02 The smallholder credit project was to be implemented by the District Rural Development staff, supported by the Animal Industry Division field staff. District Rural Development field staff were responsible to 4 Regional Controllers, one each for the Papuan, North Coast, Highlands and Islands regions. Animal Industry Field staff were administratively responsible to their Regional C - 51 - region and one field supervisor to the Headquarters of each of the six districts of major development, namely Port Moresby, Lae, Goroka, Mt. Hagen, Mendi and Wewak. The Project Co-ordinator was to be assisted at Headquarters by a Livestock Officer clerk and a typist. 6.04 It was envisaged that the number of farms per trained field staff would be 20-40 during consolidation (years 3-6), and 200 during maintenance (years 7-12). The cost of project supervision, livestock services and extension over the four years disbursement period was projected to be K1.4 million. This included 78 houses, 11 office stores, 35 4-wheel drive vehicles and 11 motorbikes. All staff, (apart from 2 field supervisor positions) buildings, equipment, and vehicular requirements for the project had been met by 1976. 6.05 At project completio1 in 1980, the number of farms per trained field staff unit was .11.5 and there were 2 Livestock Branch staff units involved. The total cost of the project supervision, livestock services and extens ion over the eight year project period was estimated to be K3.9 million. 6.06 The key extension staff in the development of this project were DPI field officers at senior Provincial level. At appraisal these officers were almost without exception experienced, skilled expatriates. Tney received direct support from Headquarters, were not subject to local pressures, and tended to emphasise supervision rather than transfer of knowledge. In the first years of the project, farmers were enthusiastic and had high expectations. 6.07 By 1977 almost all of these experienced expatriates had been replaced by young, relatively inexperienced national officers, subject to local and political pressures and involved in the concentrated training and transfer programmes required for promotion. 1. Trained field staff units are considered to be 15% of total trained field staff as it is considered that field staff spend 15% of their worktime on Smallholder Livestock Projects. 2. 1 Chief Livestock Development Officer 25% 1 Typist 25% 1 Livestock Devel.opment Officer (cattle) 50% 1- Livestock Development Officer (pigs) 50. 1 Livestock Development Officer (poultry) 50% 3. No actual figure could be obtained. This estimation is based on a comparison of actuai project duration ani staff members, appraisal projections for project duration, staff membe,rS and cost of extension services. - 52 - 6.08 Farmers, who had n,t. developed management skills as quickly as had been anticipated, were becoming disillusioned with low returns, pasture degeneration, feral cattle problems, and the reduced Government extension/supervision input. With the advent of Provincial Government in 1978, Provincial field staff were no longer responsible to Headquarters and received little real ;upport from DPI Headquarters. Provincial resourses became scarce, priorities changed and effective extension input into livestock was greatly reduced. However, even in the early years of the project, field officers considered the project extension requirements projected at appraisal to be unrealistically optimistic. Paoua New Guinea Development Bank (Annex 7) 6.09 The Papua New Guinea Development Bank (PNGDB), established in 1967, is a statutory authority of the Papua New Guinea (PNG) Government, with a wholly Government subscribed paid-up capital of K26.2 million. 6.10 The Bank provides finance fQr primary production, and the establishment, development, or acquisition of industrial or commercial undertakings. The Development Bank Ordinance states that the PNODB should consider prospects of success rather than security available, and the development aim3 of the Government, particularly with regard to the financial needs of small projects and the development of the citizens of ?NG. 6.11 Under the 348 credit project agreement, US$.6 million of the US$5 million credit was made available to the PNGDB under a subsidiary loan agreement between ,the PNG Government and the PNGDB. This loan was to finance over L years, 870 smallholder beef farmers of 15 breeder potential, 270 two sow pig breeding/fattening units, and sixty 100 layer and sixty 600 broiler per year smallholder poultry units. 90% of PNGDB financed subloan development expenditure was eligible for IDA re-imbursement under the project; amounting to US^b4,218,000 for cattle, US$279,000 for pigs, and US$103,000 for poultry. This subsidiary loan to the PNGDB was for USS4 million repayable at 1 1/11S interest over 15 years with 5 years grace. The remaining US$0.6 million was in the form of a capital grant to the PNGDB. 6.12 At completion in 1980, 1,028 cattle sub-loans valued at K2,901,624, 247 pig sub-loan for K215,898, and 307 poultry sub-loans for K59L1,334 had been advanced. This gave a total of 1,577 loans valued at K3,71i,856 over.8 years, compared with appraisal projections of 1,260 loans valued at K3,766,800 for L years. Total bad debts for the project were K82,669 for 133 sub-loans and arrears amounted to K678,291 for 897 accounts. 50% of these accounts are in arrears because loan repayments need re-scheduling. 6.13 Administrative changes within the Auditor-General's Department have delayed presentation of PNGDB audited accounts to the IDA as required under nection 4.02 of the Project Credit Agreement. The IDA has also expressed concern at the lack of specific reference to the project in the audit opinion. The 1978 audited FROM accounts are now at hand for presentatioh to IDA, and 1979 accounts are with the Auditor-General's office. PNGDB's audit reporting is now in accordance with World Bank guidelines on the audit of Development Finance Companies. - 53 - 6.14 A consultancy review of the PNGDB as part of a World Bank Agricu1tu,,j Credit Project Preparation in 1980 concluded that PNGDB performance compared favourably with other development banks in similar smal, rapidly developing, countries. The consultant considered that financially PNGDB was in a satisfactory state and in 1980 should eliminate its cummulative deficit which reached almost K1 million in 1974. 6.15 At appraisal in June 1972, total approvals were K27.8 million, anI during 1971-72 K8.2 million in loans was approved. By the end of 1979, total approvals had reached K105 million and during the year, K16.2 million in loans was approved: 33% for agricultural, 37% for the service industries, 18% for manufacturing and mining, and 12% for commercial !cans. Monitoring and Evaluation (Annex 8) 6.16 The credit agreement required the Government systematically to monitor and evaluate the financial and economic performance of the project. 6.17 The PNGDB has a standard system of evaluating all smallholder loans every six months as a basis for loan or repayment re-scheduling. The data is collected by DPI Rural Development Staff as PNGDB agents using forms designed jointly by DPI and PNGDB. At appraisal the credit 3148 Project Co-ordinator proposed to use the same system and modified forms for project monitoring. 6.18 These forms were received by the Co-ordinator on a regular basis, but no satisfactory monitor4ng or evaluation was carried out as the f6rmat was not suitable for statistical analysis. Following pressure from IDA, and recommendations of an IDA-funded DPI/PNCDB Work Simplification Study and the DPI Rural Statistician, a national Smallholder Cattle Census was conducted in 1978. 6.19 The Census was an excellent extension exercise providing detailed data which was easily to analyse on cattle numbers, control and performance, farm size and location. However, apart from indicating whether farms were PNGDB-financed and whether loans were on schedule, there was no financial or economic evaluation of IDA-financed cattle farms. The Census, designed as an annual activity, has not been repeated because of staff shortages in the DPI Livestock Branch and the Statistical Secti.on. 6.20 Bi-anhual PNGDB loan evaluation is also carried out for smallholder pig and poultry loans. But there is no evidence of any systematic economic or financial monitoring of IDA pig and poultry smallholders other than Niugini Tablebird's Annual Report which details financial performance for the 1,000 bird broiler units at Lae. These units were financed under credit 348. 6.21 The IDA required quarterly progress reports on the progress of project implementation. These were provided by the Project Co-ordinator on a regular basis until June 1976. Only 4 prc__ress repzrts were submitted from that date until project closure in March 1980. The reduced reporting %as the result of staff shortages and changes, the slowing down of disbursement, and a shift in Government policies and priorities. - 54 - Disbursements (Annex 9) 6.22 PNGDB disbursements to sub-borrowers were related to investment phasing and were generally up to two years for beef and one year for pigs and poultry. 6.23 There were delays between approval and disbursement to farmers of up to 10 months recause of cattle distribution, reconciliation problems, and slow farm development. The DPI and PNGDB did not consider this delay caused hardships to sub-borrowers because funds were allocated only. for purchase of capital items, with payments generally being made directly to the supplier. No allowances were given for living costs. 6.24 Loan disbursement, projected at appraisal to be over 4 years, was extended to 8 years from the 1st of September 1972 to the 30th of March 1980 due to slow rates of smallholder cattle development. By the completion date of the 31st of March 1980 disbursement was 89% of the US$5.0 million projected at appraisal. Three closing date extensions were granted from the 31st of December 1976 to the 31st of December 1978,.to the 30th of June 1979, to the 31st of March 1980. 6.25 Re-allocations were made in 1975 and 1977 to allow for pig and poultry expansion, additional poultry laboratory equipment, pasture seed production equipment, and consultancy services for work systems development, project preparation and the 348 Completion Report. Compliance with Covenants 6.26 The borrower and the PNGDB complied with all covenants except in the following respects. 6.27 Section 3.05, Credit Agreement: the establishment of "a system to monitor and evaluate the financial and economic benefits accruing from the Project." A cattle census conducted in 1978 was a useful extension exercise and provided detailed statistics on the technical performance of the smallholder cattle herd. However it did not provide specific monitoring or evaluation of the financial and economic benefits accruing from the project and has not been repeated. 6.28 There has been no system established to satisfy section 3.05 of the Credit Agreement with regard to smallholder pig projects. But Niugini Tablebird's Annual Report provides the necessary financial and economic evaluation of the project poultry development. 6.29 Section L.02, Project Agreement. PNGDB has not been complying with this covenant which requires it to submit to IDA audited financial statements within 4 months of the end of its fiscal year. PNGDB has not submitted 1978 or 1979 audited financial statements to the IDA, and the pre-1978 audited statements submitted have contained no separate audit opinion following the requirements of Credit 3L18. Administrative problems within the Auditor-General's office have caused auditing delays. These problems are being solved, and the 1978 PNGDB financial statement has been audited and is on-hand for delivery to the I3A. The 1979 financial statement is still with the Auditor-General's office. By 1980, PNGDB's audit reporting was in accordance with World Bank Guidelines on the audit of development finance companies. - 55 - WORLD BANK PERFO?MANCE 7.01 Because of staff changes within DPI and the PNGDB over the life of the project, first-hand information on details of pre-appraisal negotiations between the World Bank and the Government are unavailable. Performance at Appraisal 7.02 The World Bank Appraisal Mission in January/February of 1972 followed a project proposal from the PNC Government in 1971. The project development proposed in the appraisal report presented to the PNG Government in October 1972 was well prepared and detailed and differed very little from the Government proposal. The major differences were in the number of farms to be developed, the size of the beef farms and consequently, the size of the loan. 7.03 The World Bank development projections, although only approximately 70% of PNG Government projections, proved to be over-optimistic. However for the first 18 months, apart from a different poultry development strategy and the farm/extension staff ratio, the project was on schedule. It was evident to DPI field staff in the early stages that extension staff projections were inadequate for the project development. 7.04 The slowing down of the expansion of the smallholder livestock industries began in 1975 with rapid changes and a shift of emphasis from expansion to training within DPI, waning farmer enthusiasm and competition from other industries like coffee, transport and trading enterprises. Implementation 7.05 World Bank supervision missions were carried out in 1973, 197b, 1976, 1977, 1978 and 1979. The first four missions combined supervision of credit 3LI8 with other World Bank credits. Detailed supervision reports were prepared on each mission with implementation recommendations forwarded to the borrower. The time-lag between supervision missions and implementation recommendations to the borrower was usually about three months. 7.06 The rate of project development was specified in detail in. both the Appraisal Document and the Credit Agreement. However, the World Bank's approach was flexible, and symathetic to the needs of the rapidly changing situation in Papua New Guinea with Self-Government, Independence and rapid localization of Government services. This flexibility is manifested in three re-allocations and three closing date extensions. Recommendations following supervision missions were generally constructive. 7.07 The separate project monitoring and auditing insisted on by the World Bank may have been considered unecessary by the borrower, but it was stated or imlied in the Covenants of thc agreement. The standard of project monitoring and auditing desired by Lhe *:orld Bank was never achieved, but auditing and smallholder performance monitoring both in DPI and PNGD3 improved significantly during the project period because of World Bank pressure, advice and funding of consulting assistance in - 56 - the preparation of a "Procedures Manual for Smallholder Loans". The National Smallholder Cattle Census was initiated partly. through World Bank advice on the monitoring requirements of credit 348. 7.08 The major lesson to be learnt from the formulation and implementation of credit 318, is that detailed development strategies and schedules are required in project appraisal and agreement documents but should only be considered as a guide to project implementation. INTERNATIONAL LOANS TO PNG (Annex 10) 8.01 The three major sources of international loans for Development in PNG are the World Bank, the Asian Development Bank and the Economic Development Fund of the European Economic Community. 8.02 The World Bank has lent PNG USS71 million since 1968 through the International Bank of Reconstruction and Development, and US$25.2 million through the International Development Association. 8.03 The Asian Development Bank has lent US$65.6 million for development and granted US31.1 million as technical assistance to PNO since 1971. 8.0 The Economic Development fund of the EEC has an indicative aid prograTne under Loan 4 of 10 million European Units of Accounts (EAU) (US31) million). At June 1980, 20% of this indicative aid programme had been committed to Development Projects. RECOMMENDATIONS A Beef 9.01 The existing smallholder cattle industry should be consolidated. Extension should be directed towards improving the transfer of knowledge. Research should concentrate on solving those management problems which result in poor calving precertages, high breeder mortalities, pasture and facility deterioration and feral cattle. With reasonable financial management and technical performance, smallholder cattle projects can be financially viable, providing the project area is adequate, pasture quality is reasonable, water is available, and there is ready access to market. 9.02 Regular smallholder cattle censuses based on the 1978 National Smallholder Cattle Census (see Annex 1) are recommended. The data thus provided would act as a base for national DPI's involvement in the consolidation of the smallholder cattle industry through provincial debriefings. 9.03 Expansion should not be actively encouraged. But credit and extension assistance should be available to credit-worthy farmers who have met equity rcquirements in terms of land, improvement, labour and expertise. Any expansion in the eattle industry dacigned to replace meat imports should be through larger-scale partnerships oetw;en commercial development agencies and land owners. The cptions of - 57 - division into sma.llholdings, or the development of peripheral smallholders, should be kept open. The integration of cattle with existing or proposed grain and tree crops large-scale developments has potential for national herd expansion. 9.04 DPI's role should be project identification, feasibility studies, negotiation of finince, protecting the interests of the land-owning partners and ensuring their -participation in project development and implementation, as well as providing technical expertise as required. 9.05 The PNGDB is actively involved in developing large-scale joint-venture cattle ranching projects and has the expertise and infrastructure successfully to implement this expansion on a commercial basis. International credit for cattle expansion should be injected through the PNGDB and other Papua New Guinean commercially-oriented development agencies. B Pigs and Poultry 9.06 Smallholder commercial pig and poultry development should be limited to nucleus estate projects. Commercial interests should establish the parent companies and provide stock,- feed, financial and technical management, and marketing. Units should be developed close to sources of supply of feed and stock, and markets. DPI's role should be in identifying areas for development, feasibility studies, farmer selection, and ensuring that the industry's, farmers' and the Government's intere-sts are protected. Any international credit injected into the development of these industries should continue to be through the PNGDB who have the expertise, infrastructure and experience to establish commercial nucleus estate parent companies. Research should continue into improving production and performance in both industries as well as in the subsistence village pig and poultry sector with a view to improving village nutrition and income-earning opportunities. But credit-based commercial production should not be encouraged in the rural subsistence village where management expertise, labour and feed requirements are generally unavailable at the standard required for economic production.  WEST IRIAN (INDONESA) I j- r- - -- c> 0 i A CC/ z z z_ _ ez z - \ \ z <410 z, 0 00 z > z~ zI ~ l 9-0 I - IO