Malawi - Smallholder Fertilizer Project

Do:umcnt of The World Bank FOR OFFICIAL USE ONLY Report No.8534 PROJECT COMPLETION REPORT MALAWI SMALLHOLDER FERTILIZER PROJECT gCREDIT 1352-MAI) APRIL 18, 1990 Agriculeure Operations Division Southern Africa Department Africa Regional Office flits document has a restricted dishrbton nd may beue by reipetony in th efra of their offichii duties. Its contents mayl not otherwise be discosed withu World Bak _h_  ABBREVIATIONS ADD Agricultural Development Division ADMARC Agricultural Development and Marketing Corporation ASA Annual Survey of Agriculture FINCOM Finance Corporation of Malawi FU Fertilizer Unit RAF igh Analysis Fertilizer JMC Joint Management Committee MA Management Accountant MITCO Malawi International Transport Company MOA iMnistry of Agriculture NRDP National Rural Development Program PCR Project Completion Report SFRF Smailholder Fertilizer Revolving Fund  FOt OM CIA USE ONLY THE wowsD SANK Washtngton. D.C. 20433 U.S.A. ON" cW Dict4vGwGal April 18, 1990 MEMORANDM WTO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Malawi Smallholder Fertilizer Project (Credit 1352-HAI) Attached, for information, is a copy of a report entitled "Project Completion Report on Malawi - Smallholder Fertilizer Project (Credit 1352-MAl)" prepared by the Africa Regional Office. No audit of this project has been made by the Operations Evaluation Department at this time. Attachment This document hs a rsict distribution and may be used by tecipients only in the perfofmance | of theit offlcial duties. Its contents may not otwise be disclosed without Wold Bank authorization.  FOR OFFICIAL USE ONI MAL~AW~I SMALLHOLDER FERTILIZER PROJECT (CR. 1352 - HAI) PROJECT COMPLETION REPORT TABLE OF CONTENTS PaRe Preface i Evaluation Summary ii PART I A. PROJECT BACKGROUND . Introduction ................ ..... 1 B Background ..................................... 1 P Project Formulation and Objectives ............................. 1 P Project Design and Organization ................................ 2 B. PROJECT IMPLEMENTATION Establishment of the Fund and overall Implementation .......... . 3 , Procurement and Distribution ....................................... 3 , Financial and Physical Performance of the Fund ............... .. 4 , Institutional Support ..................................................... 5 , Policy Dialogue .............................................. . 6 Monitoring and Evaluation . ............ . ............* ..... .. . . 6 C. PROJECT RESULTS . Fertilizer Procurement and Sales ............................... 7 Financial Results ............................................ 7 . Agricultural Impact ........... ................. ............... 8 . Social Impact .............................................8...... . Economic Impact ........................................... ...... 9 . Sustainability ..... ............................................ 9 Borrower's Performance ............................................... 10 IDA - IFAD Performance ......................................... 10 D. CONCLUSIONS . Lessons to be Derived ... ....................................... 11 ATTACHMENT As Bank Comments 12 ATTACHMENT B: IFAD Comments (1) 16 IFAD Comments (2) 18 PART II (No comments) This document has a restricted distribution and may be used by recipients only in the performance of theit official duties. Its contents may not otherwise be disclosed without World Bank authorization.  PART III . TABLES TABLES 1 - Related Bank Loans andlor Credits 21 2 - Project Timetable 22 3 - Credit Disbursements 23 4 - Project Implementation 24 5 - Project Costs and Financing 25 6- Project Results 27 A. Direct Benefits H B. Economic Impact 28 7 - Status of Loan Covenants 31 8 - Use of Bank Resources 31 A. Staff Inputs 31 B. Missions PART III - ANNEX 32 FIGURES 1. Project Staff Imp'ementation 32 2. Gross Margins for Various Crops 35 TABLES 1. Exchange rates 36 2. Schedule of project disbursements 37 3. Financing contributions 38 4. Schedule of application of funds 39 5. Revolving fund 41 6. Schedule of interest and subsidy 42 7. Fertilizer procurement and sales 43 8. Fertilizer demand and sales - Appraisal and actual 45 9. External transport costs 47 10. Port distances and freight costs 48 11. Fertilizer consumption (nutrient) 49 12. Fertilizer consumption by product 50 13. Current fertilizer prices and subsidy levels 51 14. Projected and actual subsidy levels 52 15. Fertilizer sales forecasts 54 16. Cropped hectarage 1982183 - 1988189 55 17. Distribution of farm holdings and cultivated area by farmsize 56 18. Maize cultivation as a percentage of cultivated area by farmsize 57 19. Fertilizer use by farmsize and crop 58 20. Smallholder fertilizer and maize prices 59 21. Nutrient price/maiz, price ratios for smallholders 60 22. Comparison of nutrient costs of fertilizers 61 23. Gross margins for various crops 62 24. Maize cultivation and production 63 25. Seasonal credit and fertilizer sales 64 APPENDICESs 1. Economic Analysis 66 2. Monitoring of Fertilizer Delivery 75  (i) K4AWI SMALEOLDER FERTILIZER PROJECT PROECT COMPLETION REPORT This is the Project Completion Report (PCR) for the Smallholder Fertilizer Project in Malawi for vhich IDA Credit 1352-MAI in the amount of SDR 4.6 million was approved on May 5, 1983 to supplement IFAD Loan 120-4W in the amount of SDR 8.5 million and IFAD Grant 1.0 million which were approved on May 24, 1983. The Credit was closed on March 31, 1988, the original closing date, while the IFAD Loan and Grant closed on March 31, 1989, one year behlnd schedule. At closing the Credit showed a balance of SDR 18,388 which was cancelled on September 30, 1989, while disbursement of the IFAD Loan and Grant for commitments entered into before the closing date continued until September 30, 1989 when the Loan showed a balance of SDR 5.672 and the Grant SDR 113,541. The PCR Part I and Part III Annex were prepared by the FAO/IBRD Cooperative Program, while the Bank and IFAD provided comments on the PCR (Attachments A and B respectively). The Agricultural Division of the Southern Africa Department prepared the Preface, Evaluation Summary and Part III tables. The Bank sent the Borrower the Evaluation Summary and Parts I and III of the draft PCR with the request to prepare Part II but no reply was received. Preparation of this PCR was started by the Cooperative Program team which visited Malavi during February 1989 and is based, inter alia on information gathered during the Mslawi visit, the Staff Appraisal Report, the Development Credit Agreement, the IFAD Financing Agreement, supervision reports, the Mid-Term Review prepared by IFAD, correspondence between the Bank, IBAD and the borrower and studies financed by USAID on institutional aspects of the fertilizer agency and on the impact of fertilizer sub'tidy on farmers and the economy. The mission also reviewed the accounts and audit reports of the fertilizer agency.  (ii) MALAWI SHALLEOLDER FlRSTLIZER PROJECT PROJECT COMPLETION REPORT EVALUATION SUMMARY Obiectives Ci) In densely populated Malawi, additional agricultural production must come from intensified farming. Fertilizer is the most importaut input for sustaining and increasing agricultural production. In the early 808, fertilizer supplies were jeopardized by financial problems of the Agricultural Development and Marketing Corporation (ADMhRC) responsible for importiug and distributing fertilizer to smallholders. Consequently, the pro4ect was designed to increase the productivity of the smallholder agriculture through improving MQA'8 policy making ability, sufficient and timely financing for fertilizer procurement. and technical support to the subsector. These objectives were to be achieved by improving procurement and distribution procedures; establishing a fertilizer revolving fund into which all revenues would be paid; strengthening ADMAICs and Ministry of Agriculture's (MOA) capacity to deal with fertilizer and related issues5 end helping to reforn policy issues particularly with respect to input and crop pricing (para 3). Of the SDB 14.1 million IDA and ITAD finance, SDR 13.1 million were for the purchase of fertilizer (foreign costs) and SDR 1.0 million for technical assistance. Imulementation ineerience ii) The project was identified by IFAD and appraised by a joint BankI/FAD team. Identification to effectiveness took only 16 months (para 1) The Government established the Fertilizer Revolving Fund promptly and operated it effectively (para 7). Fertilizer procurement and transport from port to depot were handled efficiently by contractors, but transport from depot to consumer and distribution did not improve during the project (para 9-11). (iii) The Fertilizer Unit (FU) in the Ministry of Agriculture (MOA) was established 15 months late. Despite transport and commnication problemL, the Unit did good work notably improving demand projections, introducing *Eigh Analysis Fertilizer (tAF)v to save transport costs, and introducing small bags to make fertilizer use accessible to low income farmers (para 21). (iv) The Management Accountant to operate the Fund was appointed on time (appointment was an effectiveness condition). The three other TA posts were appointed with a 12-15 months delay. Three of the four appointments were considered to be effective, but the transport adviser left no impact. Short term TA for monitoring and evaluation was not used. Disbursements of the IFAD loan and IDA credit were made as scheduled, but the TA Grant (72 of total assistance) was drawn down more slowly than projected due to delays in TA appointments and was not fully disbursed (Part III Table 3).  (iii) Results (v) Fertilizer sales in terms of nutrients were close to appraisal projection (down somewhat due to a low maizetfertilizer price ratio (from 1985 through 1987), but returning to normal when the ratio increased (para 14). Donor grants were needed to help finance the supplies. While the Fund has grown to MK 95 million (para 31 and Part III Annex Table 3). partly due to the EEC buffer stock project and other donor grant contribution, this is not enough to cater for the growing fertilizer demand and rising costs. (vi) Despite higher than projected maize yield response to fertilizers and increased fertilizer consumption, national agricultural production of the smallholder sector did not increase. This may be partly due to falling yields on unfertilized land and partly to the drop in hybrid maize area. Hybrid maize mainly grown for sale, was adversely affected by the maize price relative to other cash crops and by temporary maize marketing problem. (vii) The policy dialogue with the Government was, at an early stage, taken over as part of structural adjustment lendiag and there have been regular consultations on crop and input pricing. Fertilizer subsidy elimination, projected at appraisal, was postponed due to steep cost increases following the closure of Hozambican railways. (viii) Overall, the project succeeded in its main objective of maintaining the fertilizer supply for the smallbolder sector. There were two disappointments. The first, the lack of improvement of fertilizer distribution due partly to unsatisfactory monitoring of fertilizer movements, but mainly due to the ageing and dwindling transport fleet and inadequacy of the Government prescribed transport rates (para 12). The other is the lack of monitoring and impact evaluation that persisted, despite repeated consultants' efforts funded by IFAD to redress this defect. Sustainability (ix) 1hile the resources of the Fund built up to beyond appraisal projections, partly because of support from USAID, EEC and other donors, it is not able to meet the rising fertilizer demand in the face of increasing fertilizer and transport costs. New capital injections will be needed for the foreseeable future, while the fund cannot recover the full cost plus a margin for expansion (para 38). The direction of the Fund and of the Fertilizer Unit benefitted from effective technical assistance. The handover of Fund management to local staff and replacement of technical assistance to manage the buffer stock was a major issue during the last two project supervisions and had not been finalized at the time of PCR preparation; The process will receive attention from the ongoing sector adjustment operation and is likely to attract EEC support. The Fertilizer Unit, in KOA, has reverted to a one person unit that cannot provide all the services requested by the Ministry. With possible staff changes that are cowmon in the Ministry, the long term effectiveness of this unit may be doubtful.  (iv) Findings and Lessons Learnt (x) The project restored reliable fertilizer supplies and established an institution that, with further external assistance, should be able to meet Malawi's groving needs for fertilizer. Introduction of HAP fertilizer to reduce transport costs and of small packs to give low income farmers access to fertilizers count among other project benefits. However, the long term objective of increasing national produetion was not achieved. Fertilizer alone is not enough for that purpose and policy issues and the role of other agricultural services that affect production are being dealt with by other Bank operations. The project fell short in attempts to improve fertilizer distribution (partly because of shortage of vehicles in Malawi), and to introduce effective monitoring and evaluation. (xi) To meet the fertilizer supply crisis, project preparation had to be brief and this factor might have justified earlier mid-term evaluation to correct shortcomings. In this case there were monitoring and evaluation problems that IFAD attempted to overcome by repeated consultant assistance to the project, but in retrospect it seems that the Ministry of Agriculture Unit did not have the capacity for this work in addition to its existing commitments. The most important lesson, however, seems to be the difficulty of satisfactory institution building by a one time only project. As has been found with research aud extension, support should be continued over several project cycles.  MALAWI SMALLHOLDER FERTLIZER PROJECT (CR. 1352 - MAI) PROJECT COMPLETION REPORT PART I - PROJECT REVIEW FROM THE BANK'S PERSPECTIVE Proiect Identity Names Smallholder Fertilizer Project Credit No.: 1352-MAI RVP Unit, Africa Region Countrys Malawi Sectors Agriculture A. PROJECT LACKGROUND Introduction 1. The Smallholder Fertilizer Project was identified and prepared by IFAD from February to July 1982, appraised in October 1982 jointly by IDA and IFAD, and negotiated in March 1983. The four-year project was estimated to cost US$29.97 million to which IDA would contribute by a credit of US$5 million, and IFAD by n loan of US$9.28 million and a grant of US$1.05 million. The project became effective in June 1983; Credit 1352-MAI closed on 31 March 1988 but IPAD loan 120-MW and grant 109-MW continued to 31 March 1989. Background 2. Agri.culture in Malawi engaged about 85? of the working population, and accounted for 401 of GDP and 89Z of export earnings in 1982, at the time of project appraisal. With rural densitiws of 130 to 200 persons per km2 of arable land, and an average farm size of 1.2 ha, fallow lands have virtually disappeared in smallholder agriculture. Government objectives are to maintain self-sufficiency in staple foods while expanding agricultural exports and improving rural incomes. Fertilizers started to play a strategic role in the early 1980s, with about 202 of cropped land regularly receiving fertilizers, as against only 10? in 1977. Marketing and input delivery services for smallholders are provided by the Agricultural Development and Marketing Corporation (ADMARC), a parastatal, which operates a nation-wide network of stores and marketing sheds. Proiect Formulation and Obiectives 3. At the time of project appraisal, while research, extension and credit services were being supported under the National Rural Development Program (NRDP), further improvement in smallholder productivity was constrained by insufficient fertilizer supply and late delivery. The situation became serious during the 1981/82 season because of ADMARC's growing financial difficulties, and progressively worsened in the subsequent seasons due to growing insecurity on the Mozambican routes used  -2 for the transport of all fertilizer imports. Non-comercial obligations contributed to ADMARC's financial difficulties, since it had to bear the costs of storing surplus maize and of subsidizing fertilizers. 4. The Project was designed to increase the productivity of the smallholder sub-sector through improvements in MOA's policy making ability, the restoration of sufficient and timely financing for fertilizer procurement and technical support to the sub-sector. It was envisaged that these objectives would be met through the following componentsx i) improvement of ADMARC's procedures for procuring and distributing fertilizers; ii) establishment of a fertilizer revolving fund in the Reservi Bank of Malawi; iii) strengthening of the institutional capacity of the MinWstry of Agriculture (MOA) and ADMARC; and iv) the 'setting in motion of a process to bring about a reform of all relevant policies,' through agreements on subsidies and regular consultations on resource allocations, crop and input pricing and measures to strengthen agricultural research. Prolect Desian and Organization 5. The main thrust of the project was to establish a separate fund sufficient to ensure the timely availability of the required finance. This Fund would be managed by a small unit within ADMARC, and ADMARC would continue to be responsible for the physical implementation of procurement and distribution; however, procurement was to be effected only through its subsidiary the Finance Corporation of Malawi (FINCOO), and following ICB procedures. A Joint Management Committee (JMC), consisting of representatives from MOA, the Ministry of Finance, Department of Statutory Bodies and ADMARC, would monitor the stgtus of the Fund. 6. It is now apparent that there were certain shortcomings in project desigr. These relate, in particular, to the following: i) there was no clear indication on how fertilizer distribution was to be improved; ii) the fertilizer unit Was set up under the extension service which had no authority and very few means to carry out adequately its wide-ranging functions; iii) for M&E activities, there was no clear delineation cf responsibilities between the different institutions involved (ADMARC. Smallholder Fertilizer Revolving Fund (SFRF), the National Statistical Office, and WOA); and nor were the information requirements clearly defined; and  - 3 - iv) the agriculturist TA was given far too many tasks for the period allocat-A 'der the project. S. PROJECT IMPLEMENTATION Establishment of the Fund and Overall Implementation 7. The Project became effective on 24 June 1983 with the appointment of the Management Accountant (MA) and the paying in of the initial deposits, which were set as coaditions of effectiveness. Tenders for the 1983-84 season had been offerqd prior to project effectiveness. The Management of the Fund, atrengtheted with personnel seconded from ADMARC (Part III Annex Fig. 1), became fully operational with the initiation of new tendering and accounting procedures for tne following season. 8. Disbursements from IDA and IFAD loans had been completed by end March 1987, but IFAD grant financing technical assistance was extended to 30 September 1989, chiefly because of the late appointment of technical assistance staff and their counterparts (Part III Annex Fig. 1). IDA supervisions were, however, continued until end 1918. Procurement and Distribution 9. Annual tenders are established on the basis of an aggregation of local demand estimates prepared by staff of the Agricultural Development Division (ADD) almost two years ahead of delivery; this global estimate is then revised by HOA an: adjusted to match supply possibilities. The Finance Corporation of Malawi (FINCOM) has successfully handled the procurement of fertilizers, complying with Bank/IDA ICB procedures. FINCOM evaluates tenders and provisional awards are made by the Joint Management Camittee. It charges a fee of 12 for its services and arranges insurance for the external transport element. 10. Since the 1984-85 season, the Fund has tendered separately fcr the external transportation of fertilizers. The Malawi International Transport Company (MITCO) was awarded the first contract to coordinate external transport and has subsequently won the award each year. Due to the volumit of exports at its disposal which can te matched against imports, MITCO has been able to charge transport costs below or equal to those anticipated at appraisal, thereby offsetting the impact of the Beira port closure (Part III Annex Table 9). Although shorter routing (Dar-es-Salaam, Beira via Harare, and later, NACALA) will be progressively used more frequently, it is likely that freight costs will not be reduced substantially (Part III Annex Table 10). MITCO has experienced no difficulty in obtaining the requisite foreign exchange for external transport. 11. The demand estimates compiled by ADD staff (Part III Annex Table 10) are used to prepare schedules for distribution from the 2 ADMARC main depots to its 93 parent markets. There have been every year instances  - 4 . of late, excess or under supply in many markets. Due to lack of effective monitoring - by ADD staff as envisaged at appraisal, it is difficult to demonstrate any significant progress in this respect. Monitoring even failed to clearly identify major issues. Deviations from delivery schedules may in rare cases be made deliberately by ADMARC to remedy inaccurate demand estimates. In most cases, however, .1ey would appear to stem from transport difficulties and lack of control by ADMARC over transporters. 12. The transport situation has, in fact, deteriorated further during the 1988-89 season, due to increased competition for the transport of food aid for the estimated 600,000 Mozambican refugees now in the country. These exceptional circumstances should not, however, hide the structural causess an ageing and dwindling fleet, and insufficient Government-fixed transport rates. 13. Procurement and distribution operations are now undergoing significant changes with the introduction of the EEC buffer stock scheme. Initially intended to serve only as a strategic reserve in case of total closure of supply routes, it is now envisaged to use it also to stagger imports, thus reducing pressure on transport, and allowing the Fund to benefit from advantageous spot prices. For the Fund, this will result ins i) an increase in the number of annual tenders and in the number of primary delivery points into Malawis imported fertilizers will be delivered directly to the three buffer stock stores and to a number of parent markets in addition to the two ADMARC main depots through which all fertilizer had to pass until 1988; ii) its direct involvement in store management and fertilizer handling through the buffer stock stores; and iii) the establishment of more formal relationships with ADMARC, particularly regarding delivery schedules and monitoring mechanisms. Financial and Physical Performance of the Fund 14. By 31 March 1988, contributions to the Fund amounted to MR65.ll million, i.e. about twice the Appraisal Report estimates (Part III, Annex Table 3). Inspite of this increase, which was due to contributions from USAID and to a mucb larger than projected contribution of the Government, the Fund faced serious liquidity problems during the first three years of the project. About two thirds of the current assets were represented by debts owed by ADMARC, ADDS, and the Government in regard to subsidy payments (Part III Annex Table 4). The Government then came to the rescue of the Fund, which also succeeded in obtaining faster payments from debtors. As a result, debts amount to less than one third of the current assets since 1986-87. 15. Quantities of fertilizer made available every year have always been maintained at levels close to or above those projected at appraisal. Quantities bought by the Fund were lower than envisaged, but donor  fertilizers, which amounted to 20-252 of total sales from 1985-86 to 1B87- 88, made the difference (Part III Annex Table 8). Fertilizer sales increased in the first year but were then static for three years before resuming a strong growth when maize prices increased in 1987 and 1988. Sales were constantly and significantly lower than quantities theoretically available for sale. This may be partly due to a depressed demand related to low maize profitability in 198586 and 1986/87 (Part III Annex Fig. 2), but also to the fact that part of the fertilizer was received too late in many market places. 16. With a view to reducing transport costs per unit of nutrient imported, the project has gradually introduced 'High Analysis Fertilizers" (HAF) which represented over 552 of sales in 1988/89. Thanks to demonstration and training activities carried out by the project, these are now generally well accepted, although traditional fertilizers are still preferred by many farmers. 17. In April 1988, the Fund was separated from ADMARC to become a Trust Fund administered by a Board of Trustees and empowered to receive grants and to enter into contracts for distribution as well as for procurement; it was then given the responsibility to manage the buffer stock project and operate the three regional warehouses, which will receive the bulk of incoming fertilizers (see para. 13). Institutional Sunoort 18. The project was to strengthen institutional capacity through the provision of technical assistance: a management accountant for the Fund (4 years), a transport management adviser to ADMARC (2 years), an agricultural economist to the Planning Division of WOA (1 year) and an agriculturist to the extension service of MOA (4 years). 19. The Management Accountant, appointed before project effectiveness, initiated the new tendering procedures, established a system of accounting and prepared reports providing details of the Fund's activities; the accounts have been audited annually by independent auditors. His appointment as General Manager of the new Trust Fund in 1988 is seen as evidence of his satisfactory performance. His secondment to the Fund under IFAD assistance has been extended until September 1989. His counterpart was appointed only in 1986. 20. When the Transvort Manatement Adviser was recruited, over a year behind schedule, arrangements had already been made to overcome external transport difficulties. There is no evidence that he played a significant role in internal transport, which still remains one of the weakest links in the whole supply system. It is only since October 1988, when a new transport specialist was recruited under FAO assistance to ADMARC, that a thorough study of the transport system has been undertaken, which should provide the basis for decisive action to tackle this issue. 21. The Aariculturist was appointed only in early 1985, together with his counterpart. They established a Fertilizer Unit (FU) dealing, as envisaged at appraisal, with three major fields of activity: technical  recommendations, fertilizer demand estimates and distribution scheduling, and monitoring of fertilizer movements. These tasks far exceeded the envisaged staffing and the means that MOA could realistically provide. In spite of this and the lack of transport and office support, the FU has been instrumental in promoting the use of High Analysis Fertilizers through trials, training and demonstrations - although further work is still needed to improve efficiency of application and farmers' acceptability - and the use of smaller bags to make fertilizers more accessible to low income farmers. 22. With an input limited to 12 months, the Pricing Adviser was only able to participate in one pricing review consultation (1985-86). He was closely involved with the World Bank pricing methodology using the econometric model. Since his departure, the Planning Division has recruited a donor-financed pricing economist, and a further adviser was provided in 1987 under the World Bank financed Agriculture, Extension and Planning Support Project. Policy Dialonue 23. The Joint Management Comvittee (para 5) was not formally established until March 1985 although most of the members had met regularly beforehand. Meetings have subsequently occurred on a quarterly basis, albeit with a frequently varying membership. 24. It was intended that fertilizer prices would be kept under constant review to ensure that subsidies would be completely phased out by 1985-86 according to a schedule based on the routing and CIF prices prevailing during negotiations. With the closure of the Beira rail line in October 1982 and the Nacala line in January 1984, however, fertilizer prices rose and the Government decided to postpone the subsidy reduction program. These issues were then discussed under the structural adjustment program (see comments in Part III Annex Table 14). Prices of farm produce were also to be reviewed with a view to moving them towards farmgate parity prices. Since 1984, annual price consultations with the World Bank and IMF have occurred as part of the SAL agreement. In fact, the issue of pricing policy was taken over by the SAL project. Monitorina and Evaluation 25. The project intended to develop a comprehensive system tos (i) monitor the financial situation of the Fund; (ii) monitor fertilizer movements from procurement to selling points and sales; (iii) monitor and evaluate fertilizer use, and evaluate project effects. Since the data collection system of MOA was considered at appraisal to be well developed (e.g. The Annual Survey of Agriculture (ASA) program), no additional staff or budget were provided, other than expert consultancies. 26. The financial status of the Fund, through its accounts and regular reports, was efficiently monitored and corrective action taken when needed. The same occurred with procurement, for which regular data were contractually provided to the Fund by FINCOM and MITCO. In both cases, responsibilities were well defined and no extra data collacction was needed.  27. Data on fertilizer movements beyond the main ADMARC depots were never known to the Fund in sufficient time and appropriate format to allow effective monitoring and corrective action; besides, the Fund had no authority over ADMARC or transporters to take such action. The three M&E consultancies provided to the project concentrated on data collection, recommending changes in the distribution of tasks between the various envisaged partners, in a vain attempt to overcome staffing and equipment constraints. Above all, M&E suffered from inadequate planning (Part III App. 2). In particular, data collection was not focused to address key pre-identified issues; little effort was made to obtain distribution data from the only direct source - ADMARC - through appropriate changes in field staff reporting format and in their data processing system; and there were no time limits in procurement and delivery schedules. 28. Specific surveys to study fertilizer use and evaluate project effects were either not analyzed (rapid fertilizer survey) or not undertaken (socio-economic surveys) due to lack of staff and equipment. The analysis by the FU of 1982183 to 1984/85 ASA data did however provide valuable information on doses, and indications of maize yield response to fertilizers. C. PROJECT RESULTS Fertilizer Procurement and Sales 29. Despite continuous severe cash flow difficulties, due to late payments, and steeply rising fertilizer and transport costs, the project has been successful in supplying and distributing the amounts of fertilizer deemed necessary at appraisal. This achievement is largely attributable to donor fertilizer grants and to close collaboration between the funding institutions and the Government. 30. It is. however, clear that larger quantities could have been sold, had the project been able to improve distribution, or had maize prices been maintained at desirable levels (i.e. aimed at ensuring self- sufficiency in maize under normal climatic circumstances). The amount of unsatisfied demand cannot be estimated since the market never seems to have been saturated. Financial Results 31. During the fiscal year 1988, contributions to the Fund have increased from MK65 million to MX95 million largely due to the EEC financed Fertilizer Buffer Stock Project. This rapid increase has just been adequate to cope with the fast growing demand in fertilizers and the large increases in costs. As fertilizers' cost and demand are expected to increase further, it is very important that adequate financial resources be made available to ensure adequate and timely fertilizer availability.  Agricultural Impact 32. Because of generally mediocre soil fertility and relatively intensive cultivation methods, maize response to fertilizer is generally high, even with certain local cultivars. Fertilizer is mostly applied to maize (50Z on local varieties, 32Z on composites and hybrids) and often at levels considerably below those recommended. Farmers appear to adjust applications according to personal judgements regarding soil fertility and expected yield response, and in line with their available resources, and are probably far less ignorant in this field than extension staff would often admit. 1/ Therefore, the failure of the project 21 to modify the nationwide, uniform fertilizer recommendations to applications adapted to the various agro-climatic conditions may have less impact than feared. However, farmers' adjustment is based on an insufficient knowledge of plant needs, fertilizer characteristics and local soil deficiencies, and therefore such type of training and research would be needed to enhance fertilizer efficiency. 33. Despite the increase in fertilizer consumption, maize productivity - and therefore smallholder productivity - did not improve (Part III Annex Table 24). Declining soil fertility due to lack of fallow periods may not have been adequately compensated by the application of additional fertilizer. Moreover, the share of hybrid maize stagnated at about 5? of total area sown whereas it was expected to reach 121 in 1987/88. Hybrid maize is not grown for home consumption, due to unsatisfactory keeping, pounding and eaving characteristics, and is therefore highly sensitive to market conditions, which became unfavorable from 1985 to 1987 because of low prices and market insecurity (ADMOC was relieved from its obligation to buy all maize). Social Imoact 34. Only 23? of smallholders used fertilizers in 1i84-85, with great variations by region and farm size. This figure may have increased to 301 by now, but the general trend remains that fertilizers are used predominantly on larger farms for commercial cultivation of maize and other crops, while most farmers with less than 1 _za do .aot use fertilizers, chiefly for lack of resources, even though fertilizers could often reduce the amount of food those farmers need to buy on the market. 35. The magnitude of the food deficiency problem among the poorer smallholders is way beyond the scope of a fertilizer project. The efforts made to supply fertilizers in smaller bag sizes are however commendable (and much appreciated by farmers), since this was probably the only activity that the project could undertake to directly benefit those farmers, with limited risks of 'leakage3 to large farms. In 1987/88, 33? of fertilizers were supplied in small or medium bag sizes (10 to 25 kg), which increased to 871 in 1988/89. 1/ Much of the resistance to RAF derived indeed from extension staff disbelief or inability to demonstrate correct utilization. 2/ Due to the FU having had neither the authority nor the means to undertake or contract out the requisite trials.  - 9 - 36. Using the appraisal assumptions regarding the benefits of the project (see Part III App. 1), and assuming that 302 of the smallholders use fertilizer, the number of smallholders benefited may be estimated at 360,000. Economic Imvact 37. Following the approach taken in the Staff Appraisal Report, the project economic rate of return is very high, at over 1002, and even higher than calculated at appraisal (over 502). As project costs had been underestimated, this would result from larger than anticipated benefits. At appraisal, yield increase resulting from the use of fertilizer had been estimated at 14 kg of maize per kg of nutrient, while the actual achievement, based on the work of the FU, would be around 22 kg. The same methodology as at appraisal was used to estimate the ERR. However, it is questionable whether all benefits can be ascribed to project fertilizers and qualifications on this subject are elaborated in Part III, Appendix 1. Nevertheless, it is evident that the project is economically viable. The switching values indicate that the rate of return would still be satisfactory should the economic benefits be divided by more than 10. Furthermore, the project has played a significant role in the mobilization of foreign assistance, which is a constant key factor for improved fertilizer supply. Sustainability 38. The Fund's ability to continue to procure the required amounts of fertilizers, in the context of a fast rising demand and of rising fertilizer prices and transport costs, is entirely dependent on new regular capital injections. Assuming the Treasury would pay estimated subventions quarterly, and ADMARC and ADDs within three months from the time of sale, it is likely that the Fund could finance its projected procurements for 1989-90 from its existing resources. Additional resources would nevertheless be necessary from 1990-91 onwards. 39. The project made provisions to ensure the institutional sustainability of the Fund as it was originally designed. However, these proved inadequate when the changes in the Fund's status and role took place in 1988 (see para 17). Four new managerial positions had to be created and were filled through technical assistance. Despite efforts for the rapid nomination of counterparts, the time left for training before technical assistance expatriates are due to leave (September 1989 for the General Manager, early 1990 for the others) may be extremely short given their lack of experience in business operation. Indeed, in view of the more commercial orientation of the Fund, recruitment for management positions should be opened to candidates with adequate commercial experience and who may be outside the public administration, as already recommended by recent supervision missions. 40. The FU functioned mainly by virtue of the strong input provided by the Agriculturist. At present, it is staffed by only one person. It hbs now become obvious that the tasks assigned to it (para 21) are not  - 10 - compatible with the institutional set-up. Its head can have no more power and means than other subject matter specO.alists in the extension service. The FU could therefore only handle - if properly staffed and better equipped - technical and training tasks relating to fertilizer recommendations, in addition to acting as a focal point in the Ministry for all information on fertilizers. Clearly, there is a need to carefully review the role which the FU is expected to perform in the future, and staff it accordingly. Borrower's Performance 41. Throughout the project duration, the Government has made every effort possible to ensure its viability, particularly in providing the Fund with the resources needed to meet supply requirements. Given the close collaboration maintained year after year with supervision missions to address such problems, it is perhaps surprising that the statutory changes in the Fund in 1988 were made without first consulting the Bank or IFAD. This could have helped identify in time the present staffing issues (para 39). 42. As long as the Fund was part of ADMARC, it always enjoyed considerable support in respect of staffing and material needs. Cooperation with ADMARC regarding monitoring of distribution was, however, less satisfactory. A strict adherence to administrative rules precluded the FU from receiving sufficient staff and means from MOA - particularly regarding telephones and transport - to fulfil its numerous tasks, even though these had been increased by NOA. 43. Increases in fertilizer prices almost immediately relieved the Government of its obligation, under the project, to phase out subsidies (this program, together with the pricing policy dialogue, was thereafter transferred to other projects). IDA - IFAD Performance 44. The project preparation process was completed in a remarkably short time. This was critical for the success of the project, but was perhaps the reason for some shortcomings in project design (para 6) which were never successfully addressed later on. The initial supervision missions played an important role in the recruitment of T/A and counterparts and in monitoring the financial status of the Fund. The cashflow problems were constantly highlighted by timely supervisions, which ensured that they were effectively addressed. The supervisions were, however, less successful in correctly identifying major issues in fertilizer distribution and in monitoring and evaluation, or in securing efficient work conditions for the FU. D. CONCLUSIONS 45. The project succeeded in reestablishing Malawi's capacity to maintain fertilizer supply, following years of erratic supply and despite the closure of the traditional direct routes. Central to this was the establishment of the Fund which, with further external assistance, is now  - 11 - developing into a poteat but still vulnerable tool. The project had a further impact on fertilizer use through the promotion of HAF and of smaller bag sizes which helped to contain costs and increase accessibility to lower income farmers. 46. The project, however, failed to increase overall smallholder productivity since gains in fertilizer use were negated partly T.y declining soil fertility, partly by the stagnation of hybrid maize cultivation, due not only to the absence of suitable hybrid varieties, but also to policies which discouraged commercial maize cultivation. 47. Institutional strengthening was not adequately addressed, but on an ad hoc basis. The project also failed to improve, or even to clearly identify, distribution problems, due to inadequate monitoring. Information regarding fertilizer utilization remains scanty and inadequate to formulate further strategies to widen fertilizer use. 48. Fertilizer is very important for the agricultural economy of Malawi. Decisive steps have been taken under this project - and the subsequent buffer stock project - to improve supply. This support needs to be continued, particularly from the institutional point of view, to put the Fund on a sustainable footing. 49. Further progress in reaching the long-term objective of the project, smallholder productivity improvement, will depend on Malawi's ability to maintain the integrity of the Fund and further strengthen its capacity to control procurement and distribution. While research is needed to improve the genetic material and fertilizer recommendations. little would be achieved without a sound policy environment which needs to be based on a better knowledge of existing production systems. Lessons to be Derived 50. From the IDA-IFAD point of view, the lessons which can be learned include the following: i) since time for project preparation was not sufficient for detailed preparation of all components and since supervision missions are bound to focus on immediate priorities, it would have been useful to envisage a detailed review after, say, two years; ii) late appointments of TAs and unsatisfactory performances in transport and M&E expert services seriously affected project implementation and achievements. Alternative arrangements could have been sought to perform, in a timely manner, the tasks considered essential to achieve project objectives; iii) the failure of M&E to provide relevant information on fertilizer distribution and utilization was not primarily due to insufficient means but inadequate identification of M&E objectives and data requirements.  - 12 - PART I ATTACHMENT A MALAWI SMALLHOLDER FERTILIZER PROJECT PROJECT COMPLETION REPORT BANM CONTS A. Introduction (i, This Project Completion Report (PCR) for Halawi Smallholder Fertilizer Project (CR1352-MAI) was prepared by Staff of the FAO/World Bank Cooperative Program assigned to the Agricultural Division of the Southern Africa Department for that purpose. It constitutes one of the early attempts at the new style of Project Completion Report. This note expands the project background section of the PCR slightly and presents conments to reflect the views of the last task manager on the PCR findings. These conments should be regarded as footnotes that amplify PCR material or offer additional interpretations of the significance of the PCR findings. B. Prolect Backaround and DescriRtion (ii) With intensification the main way to increase agricultural production in Malawi, fertilizer is a crucial production factor in attaining the national objectives of food self-sufficiency and higher export crops production. For the estate sector, fertilizers are imported and distributed by the p=ivate sector, at full cost. For smallholders, the Agricultural Development and Marketing Corporation (ADMARC) was the sole importer and distributor. It has a large country-wide network of stores and sales (also produce purchasing) points. Fertilizers for smallholders are subsidized and, because of the price differential, some of the 'smallholder fertilizer' leaks to the estate sector. The extent of leakage cannot be quantified accurately, but is believed to be about 202 of ADMARC sales. The price differential is an obstacle to private sector distribution to smallholders which is under consideration. The procurement/distribution scenario has remained the same during the project except that the ADMARC Unit established by the Project to import fertilizers has become an autonomous trust fund, which continues to depend on ADMARC for internal distribution. The Ministry of Agriculture (MOA) remained responsible for demand estimates, recomaending fertilizer types and application rates, research, extension and pricing recommendations. (iii) Before the Project, the Government became concerned at increasingly erratic fertilizer supplies for smallholders. Supply difficulties were found to be due to inadequate finance and weak procurement and distribution procedures, and the Government requested IFAD and InA's help to overcome the problems. The resulting project, to be completed over four years comprisedt Fertilizer Purchases US$ 28.67 million Technical Assistance 1/ US$ 1.30 million Total Cost US$ 29.97 million 1/ Described and evaluated in PCR paragraphs 18-22  - 13 - C. Comments on Specific PCR Findinps (iv) Proiect Deuign (a) Re Paragraph 6(i) - The project provided for improved fertilizer demand estimates, procurement and transport. However, it could not overcome the transport fleet problem (paragraph 12) that impeded distribution. (b) Re Paragraph 6(ui) - In the organizational structure of the Ministry of Agriculture (MOA), the Fertilizer Unit was correctly placed, i.e., one of the subject matter units to advise the Department Head. However, the lack of operating resources that constrained the work of the unit provides a lesson for designing similar operations in the future. (c) Re Paragraph 6(iii) - It seemed clear at appraisal that ADMARC would monitor progress of procurement, distribution and performance of the Smallholder Farmers Revolving Fund (SFRF), while the MOA Central Evaluation Unit would evaluate the impact of the project on farmers and agriculture. It is true that fertilizer distribution monitoring and impact evaluation did not function despite four subsequent attempts by IFAD consultants to overcome the problems. It is doubtful whether additional resources at appraisal would have enabled the project to overcome the difficulties. (v) Implementation (a) Distribution - Re Paragraph 11 - It is agreed that monitoring of distribution must be improved and perhaps some consultancy effort specifically for this issue should have been used, but in the endt (i) the majority of fertilizer did reach the consumers more or less on time; and (ii) due to transport problems (paragraph 12), there might have been little that could have been done despite excellent monitoring. (b) Policy Dialogue - Re paragraph 24, while it is correct that the subsidy issue was dealt with by Structural Adjustment Operations, the consequences of activities and decisions in connection with the subsidy affected the SFRP. For instance, the USAID capital injection under its Subsidy Removal Program helped to alleviate the Fund's financial problems during the ADMARC liquidity crisis. The subsidy issue remains central to the Fund's viability, being an important factor determining fertilizer demand and resources available to the Fund. The Government agreed in principle with the desirability of subsidy removal, but  - 14 - became cautious when faced by lagging maize production, the need to import maize - particularly to cater for the influx of refugees, and fflat" fertilizer sales. In the medium- term, the Government feels that the value to cost ratio should be at least 2:1 for local maize. A subsidy for that purpose may be needed to offset rising fertilizer prices and is considered preferable to raising output prices in view of the hardship this would cause to food deficit households. Total subsidy removal should be possible when a hybrid variety acceptable to consumers for subsistence is widely used. Cc) Monitoring and Evaluation (paras. 25-28) - As indicated earlier, it should be noted ttat IFAD provided four consultancy efforts to overcome the problems, which proved quite intractable to even such intensive efforts to remedy defects. This is an interesting lesson in the face of Halawi's reputation for having established at effective agricultural survey organization. (vi) Proiect Results - Fertilizer Sales (nara. 30): Financial Results (para. 31) - Limited resources for fertilizer purchase was the main constraint to expanding sales. Despite project funding, SFRP became increasingly dependent on donor grant finance for expanding sales. This dependence is as great today as in the past, despite recent success in higher egearing, of the Fund's resources, due to the buffer stock and improved financial management. (vii) Sustainability (a) Fund Finances (oara. 38) - Adequacy of financial resources has become a major focus of the ongoing sector adjustment discussions. So far, the Government has preferred to request donor grant assistance to available alternatives such as IDA credits or bridging finance from local banks; (b) Fund Manamement (nara. 39) - Appointment of well qualified and experienced stsff to take over from technical assistance staff caused concern during the last two project supervisions. The Sector Adjustment Operation is also focusing on this aspect of the Fund's operations. (c) MOA (Dara. 40) - The Fertilizer Unit has always operated as an SMS unit to the Chief Agricultural Development Officer. Resources made available to it rather than its place in the organization chart determines the scope of its work, but a review of its most effective size and work program is important. (viii) Conclusions (Daras 45-47) The main Project objective, that of improving the fertilizer supply to smallholders was attained, and overall we consider this project to have been a success. Progress was also made in strengthening  - 15 - institutions, particularly in the development of the Fund. EEC assistance in this respect in the later project stages must be acknowledged. However, there is the delicate hand over of technical assistance management/leadership to local staff. This had not been accomplished for the Fund by the Project Closing Date, while the MOA Fertilizer Unit reverted to a one person opsration, considered inadequate to handle the workload. Paragraph 47 points to distribution problems, but the main cause was "the aging and dwindling (transport) fleet and insufficient Government - fixed transport rates" (para. 12) which could not be remedied by the Project. (iZ) Lessons to be Derived (para. 50) Failure of monitoring and evaluation measures feature in almost every section of the PCR. The failure has to be acknowledged and is regrettable, but it should be reiterated that IFAD made four attempts, through consultants, to redress the defects, but the problems defied solutions. The MOA Monitoring and Evaluation Unit seemed to lack capacity to carry out project evaluation in addition to its other tasks and the work should have been contracted out. On a more general theme, the most important lesson seems to be the difficulty of satisfactory institution building by a one-time only project. As has been found with research and extension, project support needs to be sustained over two or three project periods at least. In the case of Malawi's fertilizer institutions, the ongoing Sector Adjustment Operation is providing an institutional and financial safety net.  - 16 - PART I ATTACHMENT B MALAWI SMALLHOLDER FERTILIZER PROJECT PROJECT COMPLETION REPORT IFAD COMMENTS (1) Date: 14 June 1989 TO: Mr. R. Ali Chief, AF6AG World Bank, Washington FAX No. 477 6391 Through: Mr. S. Olowude Deputy Direct'jr, PF FROMs C. Dickerson Project Controller Africa Division IFhAD, Rome SUBJECT: MALAWI: Smallholder Fertilizer Proiect IFAD Loan No. 120-MV. IFAD Grant No. 109-MV IDA Credit 1352-MAls Completion Renort The report has been referred to departments of IFAD for comment. These comments are based on initial review in the Africa Division and we shall hope to inform you within 10 days if other comments are forthcoming. The report is well structured, succinct and generally fair in its conclusions. It is well supported in its tabular and appended material. In its main thrust to ensure fertilizer supplies to smallholders in Malawi, the projects achievements are well noted. It might be well to reflect that had the project not come into being when it did. Malawi's fertilizer supplies, and Malawi's ability to address the market in view of closure of main transport routes would have been seriously undermined. The project enabled Malawi to cope quite well with what otherwise could have been a most difficult situation. Under the circumstances of starting the project, there was not time for detailed research and planning (e.g., with regards to the domestic distribution network). That being said, it is true to say that better use could have been made of IFAD's grant resource for studying transportation distribution, and responsiveness could have been better towards proposals on monitoring and evaluation. Certainly the Fertilizer Unit has never been accorded the high priority in staff support/facilities that it merited. Such general policy while reflected, could perhaps be strengthened. Concern now must be that achievements are consolidated and that firm statements of those achievements are used to attract continued support from donors, to maintain fertilizer supplies, but also to ensure that institution strengthening is continued with a view to improving the efficiency of fertilizer procurement, distribution and use, especially by the smaller of smallholders. There are areas where continued support can contribute highly to development.  - 17 - The following are specific paragraph cosments, generally on matters of emphasis, covered at one point or another in the reports PARAGRAPH S In this sumnary paragraph, it may be well to reflect. as a counterweight to the shortcomings listed under paragraph 6, that the "main thrusto as per this paragraph, did in fact prove efficacious, leading to substantial improvements in supply and timeliness. It could be noted that because the project responded to a serious situation in fertilizer supply (paragraph 3), time did not permit a long preparatory phase - in fact the project went from initiation to effectiveness in 16 months. That the necessary appointments of technical assistants were not made on time was due to lack of response on the part of the Borrower in using IFAD's grant resources, (although this could be taken "as read'). Parazranh 6 (i): (Also paragraph 20). In Annex 6 (TOR 2) to the SAR, there is a term of reference for a Transport Management Adviser. This sets out that the incumbent was to review the present transportation system and effect improvements. The important task was never carried out. There was a two year delay in appointment and subsequently the appointee did not fulfil the requirements and left no report. Had the required work been done there would have been scope to introduce improvements. Paragranh 6 (ii) and 25: It is noted that IFAD did provide three successive technical inputs on M&E. However, it was found that Government did not decide to react to the recoammendations despite follow-up in successive supervisions. The reports appear to have been ostensibly ignored. It may be noted also that a comprehensive mid-term evaluation report was produced (somewhat late), but it did not appear to solicit the response on the part of the Borrower to IhAD. Paragraph (iii) and Uiv)s These comments are valid, taking into account the importance of the assignments, sufficient strength was not established in the respective fertilizer and pricing units, a matter that at the time reflected lack of response on the part of the Borrower. In retrospect, allowing that time did not apparently permit detailed plan formation the respective fertilizer and pricing units, and also that of monitoring should have been obliged to come up with detailed staffing and operational proposals early in the project period, as a basis for progress. In the event successive supervisions attempted to address the issues, but without satisfactory responses. Paragraph 33: This paragraph may raise questions on the efficiency of fertilizer use - which is not the intent. Impact assessment is of course difficult under the variable rainfall conditions and other incentive- sensitive factors such as price. This is clarified under paragraph 37.  PART I 18 - ATACHMENT B IFAD COM S (2) IIMU1 lFW FIDA PICA IF IN NARNFD FONDO JJl _ ~FUND1; 9 INTRNMNL N"KOKA Js tj a AGRfA L?TU4AL 0 V10M LOmM3T 03 EAPRROLM JDlELOPMENT AGRICOLE AGICQOLA , 107. V1 del S0*o 0814 nomq, ny -elW OM MAD. tOl. 4501 - FAX FACUlMIE TRANSIISION Fax No 47t8164 pt0 1 of 2 o"o 13 March 1,90 To Mr. Graham Donaldson Frofs Mr coliuDickeron Chief, Agriculture, InfraAtrltcture Project Coatrollere and Human Rsourcees Divisonv On4rations,