Malawi - Agricultural Sector Adjustment Credit Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 15641 PERFORMANCE AUDIT REPORT MALAWI AGRICULTURAL SECTOR ADJUSTMENT CREDIT (CREDIT 2121-MAI) may 23, 1996 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. Currency Equivalents (annual averages) * Percentage MK/US$ USS/MK April Over- Valuation 1985 1.719 0.582 0.36 61.0 1986 1.861 0.537 0.52 3.2 1987 2.209 0.453 0.41 10.4 1988 2.561 0.391 0.32 22.2 1989 2.730 0.366 0.28 30.7 1990 2.729 0.366 0.29 26.2 1991 2.803 0.357 0.26 37.3 1992 3.680 0.272 0.24 13.3 1993 (April) 4.200 0.238 0.18 32.0 1994 0.11 Source: 1985-89 World Currency Yearbook, 1990 to date, Currency Alert, various issues. MK 1 = 100 Tambala Government of Malawi Fiscal Year: April 1 - March 31 Abbreviations and Acronyms ADMARC Agricultural Development and Marketing Corporation AFTES Environmentally Sustainable Development Division of the Africa Technical Department (of the World Bank) ASAC Agricultural Sector Adjustment Credit (Cr. 1212) ASAP Agricultural Sector Adjustment Program (funded by USAID) CIMMYT Centro Internacional de Mejoramiento de Maize y Trigo (International Maize and Wheat Improvement Centre) EC European Community EDDP Entrepreneur Development and Drought Recovery project ESAF Economic Stabilization and Adjustment Fund GDP Gross Domestic Product IDA International Development Agency IMF International Monetary Fund Kw Kwacha MADIA Managing Agricultural Development in Africa MOU Memorandum of Understanding NPC Nominal Protection Coefficient OED Operations Evaluation Department PAR Performance Audit Report PCR Project Completion Report PER Public Expenditure Review PR President's Report PSIP Public Sector Investment Plan SAC Structural Adjustment Credit SAL Structural Adjustment Loan (Ln. 2026) SECAL Sector Adjustment Loan SFFRFM Smallholder Farmers' Fertilizer Revolving Fund of Malawi UDI Unilateral Declaration of Independence UK United Kingdom UNDP United Nations Development Program USAID United States Agency for International Development FOR OFFICIAL USE ONLY The World Bank Washington, D.C. 20433 U.S.A. Offilce of the Director-General Operations Evaluation May 23, 1996 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: MALAWI-Agricultural Sector Adjustment Credit (Cr. 2121-MAI) Performance Audit Report The Malawi Agricultural Sector Adjustment Credit (ASAC) (Cr. 2121-MAI) for US$70 million equivalent (plus US$5.2 million of IDA reflows) was approved in FY90. Attached is the Performance Audit Report (PAR) prepared by the Operations Evaluation Department (OED). The project followed three structural adjustment operations and reinforced their macro-conditionality (Ln. 2026-MAI, Cr. 1427-MAl and Cr. 1644-MAI). The credit was closed fully disbursed in December 1992. Both project design and implementation proved deficient. The fundamental problem of a partial and unpredictable liberalization in the marketing of smallholder crops continued. Project design (but not implementation) called for the marketing monopoly, Agricultural Development and Marketing Corporation (ADMARC), to make an initial payment to smallholders for all tobaccos based on 45 percent of the moving average of auction prices, followed by a bonus payment which would return any profits, above marketing costs and a reasonable return to ADMARC, to producers. Only for a small volume of burley tobacco licensed to smallholders was this done, and then only at the insistence of USAID's follow-on Agricultural Sector Adjustment Program (ASAP). The ASAC continued to support macroeconomic policy adjustments in line with earlier operations and added steps to improve the efficiency of land use on estates; to divest ADMARC of its non-marketing functions and ensure proper budgetary support for any non-commercial activities imposed upon it; to review the budget of the Ministry of Agriculture and support a re-direction of maize breeding towards high yielding flint hybrids. Most important of all, it envisaged a "new deal" on tobacco for smallholders. They were to get at least 45 percent of a moving average of market realizations for all tobacco, and the smallest smallholders were to gain the right to produce burley tobacco on a pilot basis. One important achievement stands out: the end of the estates' monopoly on burley tobacco production. This was an important, though modest, development. It did not call for any reduction in burley tobacco licenses to estates, but allocated the growth in marketings to smallholders. Since it involved a ten-fold increase in labor income per hectare, it represented a radical improvement for those smallholders affected. Unfortunately, this was less than two percent of smallholders. USAID's follow-on ASAP has developed this into a radical change in agricultural policy, starting with a free exchange rate, multiple channel marketing, removal of any effective limitation on smallholder burley production and the cap on payments to estate tenants has been removed. This rapid progress only throws into relief what was not achieved under the ASAC. The ASAC relied on promises of future action, and this probably contributed to its lack of success, whereas ASAP's insistence on action before disbursement was the key to its success. 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. Rapid political change characterized the period of project implementation. Concerned about human rights and governance issues, bilateral donors stopped all but humanitarian aid from mid-1991. The Bank chose to continue lending, with the objective of being an honest broker between the Government, donor community and the newly assertive political opposition. The resulting progress towards improvement in human rights and governance, and fears of acute economic contraction due to reduced aid flows, came to weigh more than originally foreseen in the decision to release the second tranche. The audit confirms the ratings given the project when the Project Completion Report was reviewed. Project outcome is rated as unsatisfactory with modest institutional development and likely sustainability. The findings of the audit will contribute to OED's forthcoming study of experience with agricultural sector adjustment operations. Attachment Robert Picciotto by Ulrich Thumm FOR OFFICIAL USE ONLY Contents Preface .............................3...... ................3 Basic Date Sheet ...........................5........ ................5 Evaluation Summary ..................................................7 1. Introduction ..................................................... 11 Identification .......................................... ........ 13 Macro Policies ...................................................... 15 Sector Policies ................................................. 15 Paradigm Shift ..................I8.............................18 Bank-Government Relations ......................................... 19 2. Design .............................................. .....21 Project Objective ........................................ ....... 21 Project Design ........................ .................. 21 3. Preparation to Effectiveness.......... ......................... 27 Preparation to Negotiations ........................................ 27 Board Discussion ................................................. 28 Effectiveness ........................................... ....... 30 4. Implementation .......................... ................. 31 Macroeconomic Reforms ...............................................31 Sectoral Reform ................................................ 31 5. Outcomes ......................... ................. 39 Macroeconomic Policy ............... ........................... 39 Sector Policies ........................ .................. 39 6. Ratings .......................... ................. 43 7. Findings and Issues ...................................... ..................45 8. Lessons .................................................. 47 This report was prepared by Wilfred V. Candler, Task Manager, who audited the project in September 1995. Benjamin Crow provided administrative support. The report was issued by the Agriculture and Human Development Division (Roger Slade, Chief) of the Operations Evaluation Department (Francisco Aguirre-Sacasa, Director). 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 wihout World Bank authorization. 2 Tables Table 1.1. Malawi: Structural Adjustment Operations, 1981/90...................... 15 Table 1.2. Nominal Protection Coefficients on Smallholder Export Crops ....... ......... 17 Table 3.1. Possible Scenarios for Transfer of Burley Growing to Very Smallholders............28 Table 4.1. ASAC and Government Proposals for First-payment for Burley Tobacco, 1990/91 Season.......................................... 33 Table 4.2. Malawi: Tobacco Production and Prices, 1970-91 ......................34 Table 4.3. Budgetary Transfers to ADMARC, Due and Paid .......... ... .......... 38 Boxes Box 1: An Acknowledged Problem as Early as 1982............................ 17 Box 2: The Smallholder Estate Nexus as Seen in the Late 80's ..................... 36 3 Preface This is a Performance Audit Report (PAR) on the Malawi Agricultural Sector Adjustment Credit (ASAC) in the amount of US$70 million equivalent (plus US$5.2 million of IDA reflows) with the objective of supporting the Government's ongoing adjustment program to achieve sustainable growth, food security and poverty reduction. A Dutch Grant of Dfls 5 million and a UK Program Aid Grant of 10.0 million pounds were also disbursed. The credit was approved on April 3, 1990, and became effective on April 18, 1990. It was closed fully disbursed in December, 1992. The PAR is based on the Project Completion Report(PCR),' the President's Report,2 the Development Credit Agreement, the transcript of the Executive Directors' meeting at which the credit was considered, review of Bank files, sector and economic reports, reports of other donors and discussion with Bank staff. An OED mission visited Malawi in September, 1995, and discussed the credit with officials from the Ministries of Agriculture, Finance and Lands and Valuation. Agribusiness interests were also interviewed. The PCR provides a full and frank account of the project's performance. It is well- balanced, with Part II, written by the Borrower, providing the more interesting policy insights. The draft PAR was sent to the Borrower, the Netherlands Directorate General for International Cooperation and the UK Overseas Development Administration. 1. Project Completion Report: Malawi, Agricultural Sector Adjustment Credit (Credit 212 1-MAI), Report no. 88/93 CP-MLW 23 PCR, March 24, 1994. 2. President's Report: Malawi Agricultural Sector Adjustment Credit (Credit 2121-MAI), Report P5189, March 1, 1990. 3. Development Credit Agreement (Agricultural Sector Adjustment Program), Credit 2121-MAI, April 11, 1990. s 5 Basic Data Sheet AGRICULTURAL SECTOR ADJUSTMENT CREDIT (CR. 2121-MAI) Key Project Data (amounts in SDR'000) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 127.74 170.00 133% Credit amount 52.60 56.53 107% Cofinancing 20.00 15.00 75% Cancellation -- 0.93 na Date physical components completed Economic rate of return na na na Institutional performance: Moderate Cumulative Estimated and Actual Disbursements FY89 FY90 FY91 FY92 Appraisal estimate (SDR'000) 26,000.0 52,600.0 56,600.0 -- Actual (SDR'000) 15,137.0 26,653.0 55,774.4a 56,530.3b Actual as % of appraisal 58.2 50.7 98.5 99.9 Date of final disbursement: April 2, 1993 a Including IDA reflows of SDR 4 million approved for Malawi in late 1991. b About SDR 70,000 remained undisbursed as of April 2,1993. Project Dates Original Actual Initiating Memorandum June 22, 1989 Negotiations October/November 1989 February 22, 1990 Letters of Development Policy March 12, 1990 Board Approval December 1989 April 3, 1990 Signing April 11, 1990 April 11, 1990 Effectiveness April 18, 1990 April 18, 1990 Closing Date December 31, 1991 December 31, 1992 6 Staff Inputs (staff weeks) FY88 FY89 FY90 FY91 FY92 FY93 FY94 Total Preappraisal 9.4 53.5 62.9 Appraisal 3.3 0.1 3.4 Negotiations Supervision 24.3 13.3 7.1 44.7 Other 6.Oa 6.0 Total 9.4 56.8 0.1 24.3 13.3 7.1 6.0a 117.0 Projection from PCR (The latest Implementation Resource Report (Supervision/OIS) only states Time Spent on Supervision as of 6101/93.) Mission Data Date No. of Staff days in Specializations Performance Types ofproblems (monthyear) persons field represented rating Identification/ 2-3/89 Preparation Appraisal 7/89 Supervision 1 10/90 1 12 AE 2 Delay in PSIP Supervision 2 1/91 2 22 AE, A 2 Policy, staffing Supervision 3 12/91 na na na na na Supervision 4 1/92 2 0 AE, E 2 Low tobacco price Supervision 5 8/92 3 6 AE (3) 2 na Completion 4/93 3 30 AE (2), F na-not available AE= Agricultural Economist, A=Agriculturalist, E=Economist, F=Financial Analyst 7 Evaluation Summary' Background 1. The Agricultural Sector Adjustment Credit (ASAC) was the sixth in a series of structural adjustment operations (Table 1.1). It supported both macro and sectoral policy changes, providing a "time-slice" of the financing required for Bank-supported adjustment programs. It responded in part to the observation in a 1989 Bank strategy paper that: "Malawi presents an apparent paradox: during most of the 1980s it embraced an adjustment and development strategy guided and supported by the Bank and IMF, but its economic performance has been disappointing. Per capita income and consumption have declined, macroeconomic imbalances have persisted and social indicators have remained extremely low, even by African standards." Or, as a senior Malawian official remarked, "we have experimented with so many things, and we are still a poor country." Design 2. The ASAC continued to support macroeconomic policy adjustments in line with earlier operations and added steps to improve the efficiency of land use on estates. The ASAC also intended to divest the Agricultural Development and Marketing Organization (ADMARC) of its non-marketing functions and ensure proper budgetary support for any non-commercial activities imposed upon it; review the budget of the Ministry of Agriculture; and continue to shift maize breeding towards high yielding flint hybrids with an increased budget. Most important of all, it envisaged a "new deal" on tobacco for smallholders. They were to get at least 45 percent of a moving average of market realizations for all tobacco and, on a pilot basis, very smallholders were to gain the right to produce burley tobacco. Since burley tobacco yielded more than ten times the labor income per hectare than could be obtained from hybrid maize, this promised a significant attack on rural poverty for 5,000 farmers by December, 1990, with later expansion to 90,000 farmers. This design called for smallholders to contribute significantly to export income and represented a paradigm shift from earlier, less successful adjustment operations which had been reviewed in the 1989 strategy paper. A follow-on project, "Entrepreneurship Development and Drought Recovery," Cr. 2396, (EDDP), provided continued support for the adjustment program. 3. A major crippling distortion had underlain the strategy being pursued in the Bank- supported adjustment program, up until the ASAC. The Bank had launched analytical work in the early 1980s in conjunction with the structural adjustment operations. By 1987, it was clear that business as usual was not going to lead to a sustained recovery, even compared to growth performance levels of the 1970s. The Bank's review of twenty years' experience led to the conclusion that without a concerted effort to reduce poverty through policy and structural reform, Malawi would never be able to achieve sustainable growth. This led, in the words of Bank staff, to a new paradigm for Malawi which emphasized growth through poverty reduction. This new paradigm was articulated in the Economic Report of 1990, which was discussed 1. This summary is somewhat longer than OED normally aims for, in deference to readers from Operations who felt the draft summary left too much of the study to be read in the main text. 5/22/9611:21 AM 8 formally with the Banda Government and its partners. Even under this new paradigm, price policy was still omitted from the Bank's diagnosis of the main causes of poverty. As shown by the Nominal Protection Coefficients (NPCs), the Bank's early adjustment operations did not address the root causes of pervasive poverty and, hence, left the smallholder sector with little or no incentive to expand export crop production. The major emphasis had been on macroeconomic stabilization. Malawi always managed to recover from recurrent economic crisis-thereby achieving short-term stabilization objectives-but this was not translated into sustained growth. Implementation 4. First tranche conditionality was satisfied within two weeks of Board approval, and had it not been for the new tobacco deal for smallholders, the conditions for second tranche release would have been completed on schedule in December, 1990. A supervision mission in September, 1990, identified that the Government appeared reluctant to implement the new tobacco deal as agreed. In the event, tranche release was delayed for ten months to give the Government time (and perhaps incentive) to adopt a tobacco policy satisfactory to the Bank. This was significantly softened from the original proposal, in that (a) for the first year, 1990/91, the initial tobacco payment for burley was only 27 percent of the three-year moving average of auction realizations, rather than 45 percent as agreed; and (b) the agreed bonus payment, based on actual ADMARC profits, was restricted to burley producers, rather than all tobacco sales as originally agreed. Also ADMARC was allowed to retain a larger proportion of its profits on smallholder tobacco sales (35 percent) than might have been inferred from the President's Report. Even the bonus payment was a commitment to act, not an action. In the event, it fell to the USAID funded Agricultural Sector Adjustment Program (ASAP) to ensure actual performance of the undertaking given under ASAC. 5. Project achievements included ending the total ban on smallholders growing burley tobacco, continued support for research and extension on flinty maize varieties and an announced (but not completely effective) moratorium on conversion of customary to state land. For the most part, pricing policy continued to impose high implicit taxation on smallholders.2 The audit concludes that this pricing regime essentially sidelined and impoverished the 72 percent of the population employed in the smallholder sector and thus explains the "paradox" of adjustment without progr mentioned in the Bank's 1989 Strategy Paper (para. I above). 6. Rapid political change characterized the period of project implementation. Concerned about human rights and governance issues, bilateral donors stopped all but humanitarian aid from mid-1991. The Bank chose to continue lending, with the objective of being an honest broker between the Government, donor community and the newly assertive political opposition. The resulting progress towards improvement in human rights and governance, and fears of acute 2. Nominal Protection Coefficients (NPC) were less than 1.0 for all export crops reported (Table 1.2). NPCs refer to the ratio of nominal producer prices to export parity prices (adjusted to the farm gate level). A NPC of 1.0 indicates that the product is neither subsidized nor taxed between export and the farm gate. NPCs of less than 1.0 indicate net taxation between the farm gate and export. In particular, an NPC of 0.13 (rice in 1990/91) corresponds to a taxing away 87 percent of the gross revenue which would correspond to an NPC of 1.0. Thirteen of the 20 entries in Table 1.2 involve taxing away more than 50 percent of export parity gross revenue. Nor is that all, since the NPC is in nominal terms and does not reflect the export tax implicit in an over-valued exchange rate. This implicit export tax has varied from three to 60 percent. In more than half the years from 1985 to 1993, it has been over 20 percent. These are nt income taxes or taxes on profits, they are taxes on gross revenue. 5/22/9611:21 AM 9 economic contraction due to reduced aid flows, came to weigh more than originally foreseen in the decision to release the second tranche. Ratings 7. Almost three years after the closing of the ASAC, agricultural policy in Malawi is dominated by changes made under the subsequent ASAP and EDDP. Cotton and tobacco marketing have been completely liberalized, the cap on the earnings of estate tenants has been removed, a floating exchange rate (which tripled the Kwacha value of exports) is in place, tobacco licensing has been removed defacto and is due to be removed dejure in three years and there are no limits on private trade in maize, with the Government attempting to defend producer prices in a range 1.25 to 2.50 Kwacha per kilo. If all of these subsequent changes could be attributed to the ASAC "Post hoc Propter hoc, " then it would have to be rated as highly satisfactory. However, basing performance on changes made (rather than promised) under the ASAC, the audit agrees with the PCR in rating the project outcome as unsatisfactory, institutional reform as modest and sustainability of the few completed reforms as likely. Lessons 8. The direct policy lessons from this credit are simple: I. To significantly reduce poverty in Malawi required that the (poor) smallholders be treated at least as well as the (better off) estates. II. Malawi was unable to achieve its full development potential so long as the smallholder sector was paid only a small fraction of export earnings. III. The key to food security is, as this project argued, to allow smallholders to earn a cash income sufficient to be able to afford either to use HYV maize and complementary inputs or to buy maize (i.e., for the land-scarce, food security is incompatible with very high taxation of cash crops). IV. The existence of two systems of tenure provided no reason for limiting the crops which could be grown or the marketing channels that could be used. 9. Lesson IV parallels the Government's conclusions in Part II of the PCR: "The Government recognizes that the estate subsector has been protected for a long time and that this has been at the expense of smallholders. It is high time these subsectors were given equal opportunity" (PCR, Part II, para 7.1.2). 10. Lessons I to IV are stated in the past tense since changes introduced under the auspices of the USAID-supported ASAP and the Bank-funded EDDP have ended the discrimination against smallholders. 5/22/9611:21 AM 10 11. A more subtle lesson is: V. The decision to fund a minimal lending program in order to maintain a policy dialogue may preclude significant progress towards policy reform. By analogy with the Heisenberg Uncertainty Principle, it may be necessary to choose between a policy dialogue and actual implementation of policy changes. 12. This is a general problem for Bank staff, which presents itself in principle whenever political or economic circumstances change significantly from what was foreseen at appraisal. Bank staff can either take the view that the specific conditions for tranche release have to be fulfilled for the tranche to be released (thus making actual policy change a condition for tranche release), or the staff can invite a wider discussion as to the appropriate policies in the light of changed political or economic circumstances (thus indicating a willingness to modify tranche conditionality or even waive it). The staff argue that in this case the importance of maintaining open communication with Government justified, after a ten-month interregnum, some relaxation of the tranche conditions. 13. Policy change should not be seen as an end in itself; policy is in the nature of an "input" to the operation of the economy. Two key output or "impact" variables are already showing the benefits of the ASAC/ASAP policy changes: i) Poverty Measure. A clear indication of poverty reduction is the difficulty the estate sector has in finding tenants. There has been active competition by estates for tenants (evidence of increased prosperity in the smallholder sector). This is increasing the amount tenants are being offered and, hence, reducing their poverty.3 ii) Production Response. The target volume of burley to be sold at auction in 1995 was 99,000 tons. Actual sales were 110,000 tons, and the target for 1996 has been raised to 120,000 tons. This is a 20 percent increase in targeted sales of Malawi's most important export, in one year. 14. The enforcement of estate leases with respect to forestry and the introduction of monitoring and evaluation procedures would have required significant outlays by Government, which were not detailed in the President's Report (PR). They were in a sense "unfunded mandates." Thus, even where the Bank is providing program support, and few, if any, restrictions are placed on the use of its funds, the audit concludes that: VI. For adjustment programs, any expenses the Government is expected to incur should be appraised and budgeted for in the PR in as much detail as in a Staff Appraisal Report (SAR) for an investment project. Without such a detailed description, there is a danger that borrowers will agree to conditions without fully appreciating the financial implications involved. 3. For a number of years, some estates have paid more than the very low mandated returns to tenants. This season has seen active competition. A. Duncan, "The Impact of Pricing Policy on Tobacco Tenancy in Malawi," Food Studies Group, Oxford, April 1990 (Mimeo). 5/22/9611:21 AM 11 1. Introduction "The Bank's early adjustment operations did not address the root causes ofpervasive poverty and elusive growth.... The turn-around in adjustment lending came with the 1989 Agricultural Sector Adjustment Credit-the World Bank's first explicitly pro-poor operation in Malawi. " 1.1 Malawi, a landlocked country in Southern Africa, has a population of 9.6 million (up from four million at Independence in 1967) and an area of 48,000 square miles (about as big as New York state). It is bordered on one side by Lake Malawi, but unlike most of its neighbors has no mineral wealth. Agriculture provides 85 percent of export income (and two-thirds of the remainder are processed agricultural products). In 1990 tobacco provided 72 percent of export income, and burley tobacco alone provided 44 percent. Under the colonial regime, a dualistic agriculture was developed. This involved an estate sector with written tenure and a traditional (or smallholder) sector with customary tenure. The estate sector was comprised predominantly of large farms owned by European settlers. 1.2 The dualistic structure was continued after independence with two differences. First, many of the European owners sold out to Africans, and substantial areas of land were transferred from customary to estate tenure. This land was transferred often in quite small parcels, so that in terms of size, there is now essentially a continuum between large smallholdings and small estates. The continuation of this dualistic structure after independence represented a conscious social policy of the new president. The policy has been variously interpreted as designed to create a land-owning, and, therefore, conservative, middle-class or as a key tool for political patronage and, hence, political control. 1.3 Quite different laws, in part held over from colonial times, applied to the two sectors. The estate sector was not subject to marketing restrictions, nor was it subsidized. It was, however, excluded from the production of Dark Fired and Sun-dried tobaccos. All marketings to and from the smallholder sector were channeled through ADMARC (the Agricultural Development and Marketing Corporation) which used "profits" (that is, implicit taxes) on the export of cash crops to subsidize the retail price of maize, fertilizer sales to smallholders and a host of loss-making agricultural processing, transportation and banking enterprises. The quite incredible implicit taxation (from 13 percent to 87 percent) levied on the smallholder sector by ADMARC is summarized in the Nominal Protection Coefficients (NPCs) listed in Table 1.2. A parallel policy applied to estate tenants for whom there were mandated maximums which could be paid for tobacco (Table 4.1). Between them, these price and wage policies largely explain the abject poverty of most Malawian smallholders and tenants.2 There was one other difference: burley tobacco production was restricted to the estate sector. With UDI (the Unilateral 1. "Malawi," Briefing paper for Mr. Wolfensohn, May 22nd, 1995, p. 41. 2. "Malawi: Growth Through Poverty Reduction," Report No. 8140-MAI, World Bank, March 1990, para. 3.01, describes the sectoral distribution of poverty, but, in the view of the audit, misses the crucial price/tax factor in identifying the causes of poverty. "This report suggests six major factors: (1) limited employment opportunities; (2) low physical productivity of labor and land; (3) low levels of human capital; (4) limited access to land and economic rents; (5) minimal income transfers; and (6) rapid population growth" (para. 3.14). In the audit's view, there was no shortage of income transfer; it just happened to be in the wrong direction. 12 Declaration of Independence), Rhodesia was essentially shut out of the world tobacco market, and Malawi enjoyed a substantial burley tobacco boom, which through the 1970s delivered 5.2 percent growth real in GDP, or three percent real GDP per capita. Superficially successful, this policy delivered "export-led growth" from the estate sector, while the traditional sector was expected to focus on "food security" and subsistence production. With the energy shock and stabilization of the demand for burley tobacco, this period came to an end. 1.4 This apparently successful policy of "export-led growth" contained serious internal contradictions, as the Marxists would say. The estate sector was dependent on an ample supply of cheap tenant labor (on average they were paid about 27 percent of the auction floor price for growing the tobacco) which required low prices (and poverty) for the smallholder sector to assure the supply of cheap labor. Low prices for smallholders, translated not only into cheap labor for estates, but also into high profits for ADMARC which were used to subsidize the consumer price of maize and to invest, without perceptible developmental impact, elsewhere in the economy. This chain of causation can also be run in the opposite direction, starting with the desire for ADMARC to throw off a surplus for use elsewhere in the economy, requiring high profits for ADMARC and low prices (and poverty) for the smallholder sector, implying minimal cash crop production by the smallholder sector, which forced reliance on another sector (the estate sector) to generate increased export income. 1.5 In either case, the result was that over 50 percent of the land and about 72 percent of the population was for all practical purposes sidelined from contributing to development by high implicit taxation. It is no surprise that this policy eventually collapsed; the wonder is that it did not do so sooner and that the Bank can have had three Structural Adjustment Credits (SACs) in Malawi without a frontal attack on the key structural problem in the economy. 1.6 As a Bank strategy paper put it in 1989: "Malawi presents an apparent paradox: during most of the 1980s it embraced an adjustment and development strategy guided and supported by the Bank4 and IMF, but its economic performance has been disappointing. Per capita income and consumption have declined, macroeconomics imbalances have persisted and social indicators have remained extremely low, even by African standards. These poor results have arisen from two factors: impact of severe shocks (especially regional security problems) on a fragile and resource poor economy and rigidities in the political and social system. In retrospect, Bank assistance strategy in the 1980s underestimated the depth of Malawi's structural rigidities and the limitations of implementation capacity."5 1.7 A less generous assessment would say that the Bank's policy advice to Malawi has never seriously addressed the debilitating and exploitive relationship between estates and the smallholder sector (para. 1.4). In consequence there was no supply response from the smallholder sector and, hence, no meaningful recovery. 3. As shown in Table 1.2, there was a significant increase in NPCs for smallholder crops from 1980/81 to 1984/85 representing price policy reforms introduced as part of the initial SAL. This was the high point of smallholder price policy reform and still left tobacco with an implicit tax of 22 percent and all other crops with tax rates in excess of 25 percent; in addition, an overvalued exchange rate, levied a further implicit tax on exports. 4. Structural Adjustment Loan (SAL), (Ln 2026) and Structural Adjustment Credits (SACs) (Cr. 1427, Cr. 1644), Supplemental SAC (Cr. A-9) and Industrial and Trade Policy Adjustment Credit (Cr. 1920). 5. Country Strategy Paper, Discussion Draft, 26th May, 1989. 13 Identification 1.8 The ASAC, which is the subject of this audit, was the sixth Bank-financed adjustment operation in Malawi, as listed in Table 1.1, and, as such, it financed a "time-slice" of the balance of payments support for the adjustment program. Key actions under the adjustment program have been summarized as: "(a) devalue the domestic currency and actively maintain an appropriate exchange rate; (b) strengthen the Export Promotion Council and liberalize imports; (c) keep interest rates positive in real terms; under ASAC (the project being audited), revise the Banking Act to enhance efficiency of financial institutions; (d) increase excise taxes, import duties, and user charges; under ASAC, convert excise taxes to an ad valorem basis and revise income tax structure; (e) increase funding for agriculture and key economic and social sectors; under ASAC, continue to hold down the fiscal deficit; (f) strengthen monitoring of public investments in agriculture to ensure economic and financial viability; under ASAC, exclude low-return investments and maintain balance between investment and recurrent expenditures and within recurrent expenditures between non-wage and wage operating costs; (g) reform Press Holdings and the Malawi Development Corporation to make agro- processing more efficient and generate more foreign exchange; (h) make the Agriculture Development and Marketing Corporation (ADMARC) more efficient and expand the role of private traders; under ASAC, develop explicit agreement for Government to fund fully the non-profitable activities of ADMARC required by Government, on a transparent basis; (i) reduce fertilizer subsidies and strengthen the payment system to the Smallholder Farmers Fertilizer Revolving Fund of Malawi (SFFRFM); under ASAC, differentiate subsidies to provide higher subsidy to higher analysis fertilizer, introduce small bags and liberalize the fertilizer market; (k) (sic) revise estate expansion plans and make them more efficient; under ASAC, halt transfer of land from customary tenure to estates, raise estate land rents, adjust rents at intervals of not more than three years and introduce rents differentiated by land quality; (1) carry out an annual review of smallholder prices acceptable to IDA; under ASAC, introduce a two-payment system for smallholder tobacco; 6. A large private holding company, with equity interests in every important economic sector, including majority shareholding in Malawi's commercial banks. 14 (m) under ASAC, allow smallholders to grow burley tobacco under customary tenure and license at least 5,000 smallholders to do so; (n) under ASAC, liberalize the seed cotton market; and (o) under ASAC, change research policy to develop high-yielding flint maize."7 1.9 Bank staff are insistent that the exploitive relationship between estates and smallholders was fully understood as a result of the 1989 Strategy Paper and earlier sector work.8 They describe the ASAC as a paradigm shift which rejected the earlier, almost exclusive emphasis on macroeconomic balance and accepted the necessity to provide adequate incentives to smallholders, while maintaining macroeconomic balance. Suffice it to say that the relationship was overlooked in the Bank's poverty report (footnote 8) and, at best, poorly articulated in the PR and presentation to the Board. 1.10 Project identification stemmed from Malawi's continued need for balance of payment support and a consideration of the policy issues which still remained to be addressed. Indeed, as is often the case, the amount and timing of the desired credit were decided bfore the policy changes.9 This plagued preparation of the credit since it was not possible to find sufficient agricultural policy changes which were acceptable to the Government. The obvious option of reducing the amount of the credit to match the (modest) policy changes promised was not acceptable to the Bank on the basis of macroeconomic considerations (notably Malawi's debt service capacity) and a desire to maintain a policy dialogue with the Government. Indeed, these considerations led to the amount of the credit to be increased even as the Government's inflexibility on agricultural policy was being demonstrated. Hence, the macro-policy content of this Agricultural SECAL was high. 1.11 There was a major overlap between the policy concerns of the earlier SACs and the ASAC. The SACs contained agricultural sector conditionalities (such as the operational efficiency of the marketing parastatal ADMARC, fertilizer subsidy reduction, steps to increase the efficiency of the estate sector and smallholder prices "acceptable to IDA"); and the ASAC continued and expanded macroeconomic conditionalities (such as exchange rate adjustment, a 7. "Malawi: Economic Reform and Agricultural Strategy," Graeme Donovan, July, 1994, AFTES Working Paper No. 10, World Bank, page 2. 8. "It is just not true that, prior to discussions with the Government delegations, the Government-Bank had not appreciated the extent to which the prosperity of the estates depended on the poverty of smallholders. The Bank staff was all along fully aware-often discussed with Government staff-of the blatantly obvious facts that (the ramifications of the estate-smallholder policies were also one of the main themes at a two-day national workshop that preceded ASAC preparation and in which the participants included several senior Bank staff): (i) estate development had been at the expense of smallholders, with tobacco policy being the central issue. (ii) ASAC policy reforms posed a major threat to the estates since the reforms would put pressure on estates to pay better prices to their tenants or tenants would otherwise sell estate tobacco to smallholders, erode 'rent' accruing to the estate cartel, reduce smallholders' need to become tenants, etc." See, however, also para. 4.10 and Box 2. 9. This may often be the case. However, it appears to fly in the face of Operational Directive 8.60, Adjustment Lending Poliev, which nominates inter alia "(c) the strength of the reform being supported" as among the determinants of the size of adjustment loans (para. 53). 15 revised Banking Act, revision of the tariff structure and steps to control the fiscal deficit).The credit would have been better presented to the Board as SAC IV than as ASAC since (a) this would have invited a debate as to the need for a fourth SAC when the 1989 Bank Strategy Paper had already noted the disappointing performance under the first three, and (b) it would have removed the implication that the proposed agricultural policy components were fully owned by Government and adequate to give renewed growth. Table 1.1. Malawi: Structural Adjustment Operations, 1981/90 Title Number Approved Amount Cumulative (S million) (S million) Structural Adjustment I Ln. 2026 June, 1981 45.0 45.0 Smallholder Fertilizer Cr. 1352 April, 1983 5.0 50.0 Structural Adjustment II Cr. 1427 December, 1984 55.0 105.0 Structural Adjustment III Cr. 1644 December, 1985 30.0 135.0 Industrial and Trade Policy Cr. 1920 June, 1988 70.0 205.0 Agricultural Sector Adjustment Cr. 2121 April, 1990 70.0 275.0 Macro Policies 1.12 The Banking Act was revised to provide for a market determined interest rate. Government was committed to further liberalization of imports and to a program of active exchange rate adjustment (but not to a market determined exchange rate). The requirement that ADMARC's non-profitable activities be fully funded by Government was a major macroeconomic reform, since without this provision, ADMARC provided a way for Government to borrow from the banking system "off-budget." Government was to consult the Bank before finalizing ADMARC's producer prices. 1.13 Exchange rate misalignment10 had not been a major concern over the adjustment period and was not a major issue for the ASAC. To the extent that the exchange rate differed from the market rate, it was over-valued, thus representing a tax on exports, for the most part agricultural exports, which varied from 13 percent to 37 percent. Sector Policies 1.14 Since independence, agricultural policy in Malawi has relied on the estate sector for export income and extraction of the "agricultural surplus" from the smallholder sector. Though to some extent a holdover of the colonial regime, this policy has been refined since independence. Key components were: i) Different marketing arrangements (and, hence, dramatically different marketing margins) for the two sectors, and ii) Limitations on the crops which could be grown. Smallholders were excluded from the production of burley tobacco, and estates were excluded from dark fired and sun-dried tobaccos. 10. See inside front cover. 16 1.15 Up until the ASAC, smallholders were forbidden to produce burley tobacco (hence the characterization of the ASAC, at the head of this section, as "the World Bank's first explicitly pro-poor operation in Malawi"). The estate sector, on the other hand, could market directly to the auction floor and obtained the full auction price. For practical purposes it was neither subsidized nor taxed so that its Nominal Protection Coefficient was 1.0. 1.16 Until 1987, the smallholder sector had to market its products through ADMARC and received a small fraction of market realizations, as shown by the NPCs in Table 1.2. A NPC of 0.21 corresponds to an export tax of 79 percent and an NPC of 0.13 to an export tax of 87 percent. The lowest "export tax equivalent" in Table 1.2 is for Groundnuts in 1987/88 when it dropped to "only" 13 percent. The farm gate price for maize was 48.6 percent of its economic value (i.e., an NPC of 0.49).11 Note that these are nominal protection coefficients, reflecting taxation between export and the farm gate (between imp=r and the farm gate in the case of maize). The taxation implicit in an overvalued currency is in addition to that revealed by the NPCs in Table 1.2. In 1987, there was a partial liberalization of non-tobacco and non-cotton smallholder products, with private traders allowed to operate, but subject to administratively determined export limits. 1.17 A (small) portion of ADMARC's excess profits obtained from the marketing of smallholder crops was returned to the smallholder sector via (a) a fertilizer subsidy and (b) lower consumer prices for maize, which benefited all consumers, including the 55 percent of smallholders unable to grow all their own food requirements. 1.18 It is not the mandate of this audit to evaluate the performance of the earlier SAL and SACs; however, it is impossible to gloss over the fact that after ten years of "structural adjustment" under the guidance of the Bank and Fund, and the accumulation of $275 million of adjustment related debt, major crippling distortions underlay the strategy being pursued. As the Government's contribution to the PCR notes, "Historically it (ADMARC) has paid low prices to smallholders, averaging some 20 to 30 percent of world market prices," (Project Completion Report (PCR), Part II, para. 7.22). Twenty to 30 percent of world prices, evaluated at an overvalued exchange rate, is equivalent to a tax rate in excess of 70 percent and 80 percent on gross revenue. The wonder is not that Malawi had done poorly under earlier SACs, but rather that it had not done worse. Box I illustrates that these comments are not original to this audit; they had been available in the academic literature from shortly after the inception of the first Structural Adjustment operation. 11. "President's Report, Malawi: Agricultural Sector Adjustment Program," (PR) March, 1990, World Bank, Report P-5189, Tables I and 2. 17 Table 1.2. Nominal Protection Coefficients on Smallholder Export Crops Year Rice Tobacco Groundnuts Beans Cotton Maize' 1980/81 0.21 0.16 0.32 0.31 0.47 1984/85 0.40 0.78 0.74 0.73 0.27 1987/88 0.20 0.26 0.87 0.68 0.50 1989/90 0.49 1991/92 0.13 0.45 0.44 0.71 0.27 'Maize estimate based on import parity. Definition: NPC = ratio of nominal producer prices to export parity (adjusted to farm gate level). Note 1: Export parity evaluated at official, not market, exchange rates. Note 2: NPCs have been calculated as a ratio of nominal producer prices to export parity prices (adjusted to farngate level), based on (a) ADMARC producer prices for Faya Grade I rice; Chalimbana Grade A groundnuts; white Haricot beans; a weighted average of Northern Division Dark Fired tobacco prices for various Grades and Grade A cotton; (b) average export prices for all varieties and grades, except in the case of tobacco, which is the average NDDF auction floor price and beans, for which average pulse prices had to be used; and (c) in the absence of data on actual fob prices, marketing costs have been assumed at 10 percent of the producer price for groundnuts; 20 percent for tobacco, beans and cotton and 33 percent for rice, following Sahn et al. Source: Export crops, "Malawi: Agricultural Sector Memorandum: Strategy Options in the 1990s," June, 1994, World Bank, Report No. 12805, Table 3.3, page 32; maize, PR, Tables I and 2, pages 25 and 26. Box 1. An Acknowledged Problem as Early as 1982 "...phase-out of British budgetary grants-in-aid. The one source of financial resources which was available was for the not uncommon solution of extracting a surplus from the peasant sector via government marketing boards... .However, ...in Malawi the policy adopted...of using ADMARC surpluses to finance development resulted in the imposition of a much heavier tax burden than had been experience in previous decades....The average gross profit per annum (by ADMARC) was Kw 20.2 million, which amounts to a gross extraction from the peasant sector of Kw 181.9 million over the 9-year period.' "The decline in real value of returns to peasant labor, which we have documented, is clearly an issue of humanitarian concern, but it is also crucial to an understanding of the nature of structural changes which have taken place in the Malawian economy. We argue that the structural change which has occurred in agriculture is a result of explicit government policy implemented primarily through control of the prices received by the peasants for their output. The government, operating through ADMARC, has kept the price paid for smallholder output artificially low and as a result has greatly reduced the net rate of return to smallholder labor." "This (rapid growth in estate agriculture) can, in large part, be attributed to four factors....Thirdly, there existed adequate supplies of inexpensive labor, a condition which was promoted by government policy" [emphasis added]. ' An approximate distribution of this profit per worker (full and part-year) in the peasant sector...results in a figure for gross resource extraction of 23 Kwacha per worker per annum. As a rough guide to the significance of the 23-Kwacha figure, it represents at least a third of the per capita incomes of the poorest 90 percent of the population. [Note: Footnote is from the original.] Source: "Structural Change in Malawi since Independence: Consequences of a Development Strategy Based on Large- scale Agriculture," Jonathan Kydd and Robert Christiansen, World Development, Vol. 10, pp. 355-375, 1982. 18 Paradigm Shift 1.19 Box 1 and Table 1.2 show that a crippling distortion underlay the strategy being pursued in the Bank-supported adjustment program, at least up until the ASAC, which is the subject of this audit. Until the 1980s, the myriad of issues around politics and development (now often termed governance) were essentially treated as off-limits by the World Bank. Political sovereignty was strongly respected and the World Bank prided itself on its ability to appraise projects fairly and technocratically in diverse political environments. In general, projects in Malawi were implemented more expeditiously than elsewhere in Africa. However, the benefits of these projects accrued to fairly limited geographic areas and segments of the population and, thus, did not succeed in achieving a lasting and widespread development impact. While Malawi developed a reputation for efficiency and responsibility, this masked a growing concentration in the formal sector and lack of trickle-down impact on smallholders. In an era in which World Bank success was often measured by meeting or exceeding lending targets, Malawi was a welcome and diligent client. As shown by the NPCs, the Bank's early adjustment operations did not address the root causes of pervasive poverty and, hence, left the smallholder sector with little or no incentive to expand export crop production. The major emphasis was on macroeconomic stabilization. Malawi always managed to recover from recurrent economic crisis-thereby achieving short-term stabilization objectives-but this was not translated into sustained growth. 1.20 The Bank had launched analytical work in the early 1980s in conjunction with the structural adjustment operations. By 1987, it was clear that business as usual was not going to lead to a sustained recovery, even compared to growth performance levels of the 1970s. The Bank's review of twenty years' experience led to the conclusion that without a concerted effort to reduce poverty through policy and structural reform, Malawi would never be able to achieve sustainable growth. This led, in the words of Bank staff, to a new paradigm for Malawi which emphasized growth through poverty reduction. This new paradigm was articulated in the Economic Report of 1990, which was discussed formally with the Banda Government and its partners. Even under this new Faradigm, price policy was still omitted from the Bank's diagnosis of the main causes of poverty. 1.21 Some progress had been made in improving smallholder prices even under the rejected paradigm. Thus, under the first SAL, ADMARC's producer prices had been raised (Table 1.2; compare 1984/85 and 1980/81), but later SALs allowed producer prices to revert. In 1987, the Government started to liberalize smallholder producer markets (except for tobacco and cotton). However, this was a partial and unpredictable liberalization.13 Thus, a key issue for the audit is to what extent the project was able to translate this new paradigm into results on the ground. 12. Para. 2.2, below and "Malawi: Growth Through Poverty Reduction," World Bank, Report No. 8140-MAI, March, 1990. 13. "Malawi: Agricultural Sector Memorandum: Strategy Options in the 1990s," World Bank Report No. 12805-MAI, March, 1995, para. 4.59 says "...traders were banned from the bean market in 1988 and, intermittently, from the groundnut market between 1990 and 1993. These embargoes, together with Government pricing policies, had a powerful effect....In addition, every year under the Control of Goods Act, limits are administratively placed on the quantities of various types of produce that private traders can export...." 19 Bank-Government Relations 1.22 Historically, the Bank has prided itself on its technocratic approach and its ability to do good projects even in the presence of defective regimes. Only in the case of the most highly dysfunctional regimes has the Bank stopped lending on political grounds, preferring to keep a minimal lending program in place and, hence, a basis for a dialogue on policy and governance. 1.23 The explicit pro-poor focus of the ASAC had inevitable political implications and led to an all-time low in the Bank-Government relationship, which may have contributed to Malawi's political awakening in the early 1990s. 1.24 With the end of both the Cold War and South Africa's apartheid regime, the forces for change in Malawi gathered strength. Recognizing the Government's dependence on external donors, a nascent opposition began a campaign to influence Malawi's bilateral partners to bring pressure on the Government to open up the political system. Unusually, the aid coordination process with the Bank at the center became the main lever for change as the bilateral donors stopped disbursement of funds crucial to Malawi's economy until Malawi adopted a more open political system. With the support of these donors, the Bank played the role of honest broker in a tripartite process, involving Government, the political opposition and external partners. This led initially to a successful multi-party referendum (1993) and later (1994) to open and free multi- party Presidential and Parliamentary elections. 1.25 The Bank, in its role as convenor of the donor Consultative Group, gained the confidence of both Government Ministers and the opposition leaders and was able to help steer a course through various economic and social perils. The result was that Malawi maintained sufficient economic stability to keep the process on track while accommodating the political pressure brought to bear by other partners. This product of the Bank's aid coordination had a greater impact on Malawi's development prospects than any loan or study. 1.26 This raises a general and difficult question for the Bank as lender of last resort: When should it pull out? In this case, Operational staff argue that it was probably good to be more lenient on conditionalities and allow a minimum flow of foreign exchange that would keep the economy afloat while other donors were keeping up the pressure for political change. The counterfactual, whereby all donors, including the Bank, would have pulled out of Malawi could have easily resulted, they say, in economic chaos, civil unrest and, eventually, could have precluded the smooth political transition that has been termed a model for Africa. 1.27 This is a reasonable view, responsibly advanced and in the mainstream of Bank thinking. However, faced with a really intransigent regime, it can (a) lead to the Bank lending on the promise of very modest policy changes, and (b) leave to others the responsibility for bringing about substantive policy change or even a change of regime.14 To second-guess the Bank's decision in this case would be too easy. We will never know what would have happened if the Bank had matched the intransigence of the Banda Government; it is sufficient to note that the Bank strategy left it bilateral donors to bring about the needed policy reforms, together with the risk of inducing economic chaos and civil unrest. 14. Bank interest is, of course, limited to the advocacy of improved policies. In the case of Malawi, however, there can be little doubt that the cessation of bilateral assistance, due to human rights concerns, hastened the introduction of free elections and the associated change in regime.  21 2. Design Project Objective 2.1 The project's objective was to reduce poverty through agricultural growth (PR, para 19). This was to be achieved by measures to improve land and labor productivity. Project Design 2.2 The ASAC postulated that: "The main causes of poverty were high population growth, low productivity of land and labor in agriculture, limited employment opportunities and low quality of human capital due to nutrition levels and limited access to education and health services. The Government and donors have initiated several targeted measures to alleviate poverty, including supplementary feeding programs for the vulnerable groups, free primary health services and indirect transfer payments through fertilizer subsidies. The Government and donors, however, are aware that, given the high proportion of the poor, the problem of poverty can be addressed effectively and in a sustainable manner only through growth; and that, to ensure effectiveness and sustainability, poverty reduction measures need to be integrated into the development process. The proposed Credit will support measures to improve land and labor productivity, thus reducing poverty through agricultural growth" (PR, para 19).15 2.3 The agricultural policy changes supported in the ASAC were: i) Output Prices. Under SAL I (approved June 1981), the Government had agreed to review annually the price of smallholder crops, using a price-setting methodology developed with IDA assistance, and to submit the recommendations to IDA for review before prices were announced. This consultation was to continue under ASAC, with a lead-time of at least 21 days. In the case of smallholder tobacco, a two payment scheme was to be introduced, whereby a first payment of 45 percent of a three-year moving average of auction prices would be paid on delivery of the tobacco to ADMARC, and a second (or "bonus") payment would be made at the end of the season to reflect actual auction prices, less marketing costs and a reasonable return to ADMARC. The aim was 15. Full paragraph quoted to avoid "quotation out of context." Again, note that prices paid to smallholders are not identified as contributing to their poverty. 16. The Bank developed "methodology" aimed to make Malawi self-sufficient in maize. ("Food secure" in politically correct language.) To this end a maize price was set which was expected to yield self-sufficiency and other prices were adjusted so as to yield comparable returns to maize. A probably unintended result was to divorce smallholder prices from border prices. Under the ASAC, the Bank was also concerned for ADMARC's commercial viability. Thus "In addition, the government was required to set prices for crops handled by ADMARC so that individual crop accounts at least break-even on average over a three year period" (PCR, para 5.14). This had not been a serious problem with export crops but was designed to ensure that the consumer maize subsidy was obtained from Government, rather than by cross-subsidization. 22 to bring smallholder tobacco realizations into line with border prices. As the PR noted, "in 1988, producer prices of smallholder tobacco had fallen to about 29 percent of export parity as realized through the local tobacco auction. Low producer prices have been one of several factors contributing to the decline in smallholder tobacco production." ii) Maize Marketing. An existing scheme by which traders or farmer clubs who delivered maize in lots of one ton or more to ADMARC depots was to be expanded in the Northern Region, and, for this region, the trading margin offered by ADMARC was to be increased from 14 percent to 20 percent over the official producer price for small volumes. iii) Burley Tobacco. Production quotas for burley tobacco were to be made available to at least 5,000 smallholders (out of 1.6 million) where at least 75 percent of the recipients farmed less than one hectare. Initially, this tobacco was to be sold to ADMARC, under the two-payment scheme, but the possibility of an alternative mechanism to give smallholders direct access to the auction floor was to be studied. iv) Seed cotton. The possibility of allowing the private sector to market seed cotton was to be studied. v) Fertilizer Pricing. Under SAC II (approved December, 1984), the Government had committed itself to phase out fertilizer subsidies completely by the 1988/89 season. This subsidy removal program was suspended in 1987, following a sharp rise in transport costs caused by the closure of transport routes through Mozambique. With smallholders being paid less than half the border price of imported maize, they had no incentive to use fertilizer on local maize at unsubsidized prices. An argument was thus advanced for a continued subsidy, which would be phased out over four years, with production incentives maintained by an increase in consumer prices for maize of 9.3 percent in real terms during 1990/91-1992/93.17 This fertilizer subsidy applied only for sales to smallholders. About 25 percent to 30 percent of subsidized fertilizer was estimated to be resold by smallholders to the estate sector. In addition, a discount of five percent was to be instituted for purchases of fertilizer in amounts of at least 200 Kg (PR, para 75). This was to encourage private traders and farmers clubs to engage in small-lot retailing. 17. Ironically, the calculations in Tables I and 2 of the PR, suggest that the wrong subsidy level was selected. With the subsidy and without charging for labor or capital, Table I shows that fertilized local maize returns Kw 13.7 per hectare more than unfertilized maize. However, if the opportunity cost of labor and capital is included, this converts to a loss of Kw 39.9 per hectare despite the fertilizer subsidy. 23 vi) Fertilizer Supply. An earlier European Community (EC) initiative had established a buffer stock of fertilizer to be administered by an independent trust, the Smallholder Farmers Fertilizer Revolving Fund of Malawi (SFFRFM).18 The ASAC provided for the development of a five-year rolling plan for fertilizer requirements and sound guidelines for operation of the buffer stock. In the face of a rapidly depreciating Kwacha, the revenue from one year was not sufficient for SFFRFM to replace the fertilizer sold. This had been exacerbated by a one year lag in Government subsidy payments to the SFFRFM. Under the ASAC, Government undertook to reimburse the SFFRFM on a quarterly basis and to allow it to borrow commercially as necessary to bridge any cash-flow shortfall. The basic problem of historical costing of fertilizer sales in an inflationary environment was, in fact, met by fertilizer donations from bilateral agencies which made up as much as 25 percent of annual imports. The ASAC also provided for fertilizer to be made available in 10, 25 and 50 kg packs to assist farmers unable to afford the standard 100 kg bag. vii) Proper funding ofADMARC by Government for non-profitable activities. This, together with the divestiture of ADMARC of a range of unprofitable enterprises (which with food and fertilizer subsidies had otherwise soaked up the excess profits accumulated from the marketing of smallholder crops), promised to reduce ADMARC to an agricultural marketing organization, albeit a monopoly parastatal, but at least shorn of its major loss-making activities. viii) Maize Breeding. Support was to be given to the ongoing replacement of an earlier strategy of trying to get smallholders to adopt high-yielding dent maize varieties (which have poor storage and pounding characteristics in Malawi) by a breeding program for high-yielding flint maize. This strategy, initiated in 1987, already enjoyed the support of CIMMYT. 18. Rather confusingly, the buffer-stock (physical stocks of fertilizer held in SFFRFM warehouses and the property of the government) is managed by SFFRFM in parallel with its own fertilizer supplies which are designed to be turned over annually. SFFRFM operates with a "fire-wall" between its management of the government's fertilizer in the buffer-stock and its operation of the revolving fund, intended to ensure that smallholder fertilizer supplies are adequate. Bilateral donations of fertilizer are either to Government (to be added to the buffer-stock) or to SFFRFM for distribution or sale to smallholders. The primary aim of shifting the financing of smallholder fertilizer from ADMARC to SFFRFM was to segregate fertilizer funds from ADMARC's other financial activities and thus ensure that fertilizer imports were not inhibited by any cash-flow problems that ADMARC itself might encounter. The distinction between the buffer-stock and SFFRFM can be illustrated by the following quote from the PCR which does not make the distinction: "5.16 (b) ...the EEC's fertilizer project...required Government to take additional measures including establishing a code of practice for drawing on (THE GOVERNMENT'S) buffer stocks, settling within three months of billing outstanding accounts owed to SFFRFM (FOR SFFRFM FERTILIZER DISTRIBUTED TO SMALLHOLDERS FREE OR AT SUBSIDIZED PRICES), ensuring that (THE GOVERNMENT'S) buffer-stocks would not be reduced below the equivalent of 90,000 tons and ensuring submission of timely financial reports by the Fund's management (OF THE EXPENSES AND REVENUES FROM THE DISTRIBUTION OF ITS FERTILIZER TO SMALLHOLDERS)." The capitalized inserts distinguish when, within this one sentence, the PCR is talking of the Government's buffer-stock and when of the revolving-fund's distribution of its own fertilizer to farmers. The above mentioned "fire wall" is designed to ensure that there is no cross-subsidization of the fund from the buffer-stock or vice-versa. (In practice SFFRFM does not bill Government for the cost of storing the buffer-stock, so there is a modest subsidy from the fund to the buffer-stock.) 24 ix) Extension Service for Estates. An advisory service, targeted particularly at smaller and newer estates, was to be established under the auspices of the Tobacco Association of Malawi, i.e., independently of the Ministry of Agriculture which serves the smallholder sector. x) Halting Transfer ofLand to the Estate Sector. The Government agreed to discontinue its policy of encouraging the transfer of land from traditional to estate tenure. (This met the Bank's concern at the small farm size and rapidly increasing land scarcity in the traditional sector.) xi) Intensification of Land Use on Estates. The PR states that only 14 percent of their available land is utilized by estates in any one season19 and that Government had agreed to review rents every three years (rather than seven) and to increase rents from K 10 per hectare to K 20 per hectare (i.e., from $4 to $8 per hectare). xii) Estate Woodlots. The requirement in estate leases that 10 percent of their land be under trees, so as to be self-sufficient for fuelwood, was to be enforced. No budgetary provision for the promised Government enforcement action was included in the financing plan. xiii) Review of Expenditures of the Ministry ofAgriculture. The Government agreed to review the expenditures of the Ministry of Agriculture, in the context of an agreed public expenditure review (PER), to ensure (a) that the overall resource envelope was not violated, (b) that the highest priority activities were funded and (c) that the ratio of salaries to non-salary expenditures provided sufficient resource for the staff to operate effectively. The PR also noted that donor assistance was needed to finance recurrent expenditures; many of which were in fact of a developmental nature. xiv) Reducing Food Insecurity. According to the PCR, 55 percent of farm households run out of food grain long before the next harvest, and it noted that Government was committed to increasing food security by increased household production rather than through food aid programs or transfer payments. Many of the previous provisions were cited as contributing to household production (smaller fertilizer packs, breeding higher yielding flint maize varieties, fertilizer subsidy, allowing cash generation via burley tobacco production). A UNDP-funded project, to be executed by the Bank, was also expected to test cost-effective ways of targeting food-deficient households. xv) Monitoring. A component was included to allow the Government to monitor the impact of the ASAC. As for the enforcement of estate woodlots, there is no indication of what this component was expected to cost, nor was budgetary provision made for these costs. 2.4 Macroeconomic Policy Changes. In addition Government undertook to continue and expand macroeconomic policies initiated under the earlier SACs. These included (a) a flexible exchange rate policy, (b) further trade liberalization, (c) elimination of prior approval for imports (with the exception of a small list of luxury imports), (d) an improved tax drawback scheme for 19. PR, para. 90. The PCR (para. 5.21) says that "16 percent of cultivable land in the estate sub-sector is not used." 25 exports, (e) continued tight control of the money supply and (f) solicitation of a shift of donor financing towards recurrent expenditures in key development sectors. The fiscal impact of better tobacco prices for smallholders was minimal, since ADMARC's excess profit had not reverted to the Treasury, but rather had been used by ADMARC to fund a number of unprofitable subsidiary enterprises.  27 3. Preparation to Effectiveness Preparation to Negotiations 3.1 Review of the files, and discussions with the Bank staff and officials involved, reveal a large measure of consensus on the policy content of the proposed credit. The Aide-Memoire of the preparation mission (February 15 to March 10, 1989) records agreement on a well fleshed- out skeleton of the policy basis for the ASAC. Subsequent discussions involved principally fine- tuning of this agreement. 3.2 Bank staff were basically satisfied with the efforts which had been made to achieve macroeconomic balance, though understandably disappointed that this macroeconomic balance had not been rewarded with renewed growth. In part this was attributed to exogenous factors such as the increased transport costs due to closure of the railway routes across Mozambique and severe droughts. It does not appear to have been attributed, at least explicitly, to the microeconomic problem of the very low returns provided to the smallholder cash-crop sector (para. 7.3). However, better treatment of this sector was the key innovation of the Bank's new paradigm for Malawi's development. 3.3 The Bank's proposal to allow smallholders to grow burley tobacco was not welcomed. After what the staff describe as an enormous effort, the Government agreed to license the expansion of burley tobacco to smallholders, as shown in Table 3.1. Initially, this would allow 4,000 smallholders, out of 1.6 million, to grow burley by project closing in 1992. This did not require reallocation of any quota from estates to smallholders. 3.4 Similarly, the Bank's proposed increases in rental fees was only designed to restore them to their real level in 1985 (i.e., to Kw 30 per hectare, when gross revenue from burley tobacco is about Kw 5,000 per hectare). Even allowing for fallow land, this change could not be expected to have any significant impact on land use. 3.5 More radical were the proposed changes in ADMARC's operations by which it would cease to cross-subsidize unprofitable ADMARC operations from an implicit tax on smallholder tobacco.20 These changes involved an initial payment of 60 percent of a three-year moving average of actual tobacco auction receipts, followed by a final payment based on actual receipts less marketing costs and a reasonable return to ADMARC and a streamlining of ADMARC's operations to reduce the number of unprofitable (but "developmental") operations. The initial payment was whittled down before negotiations to only 45 percent of the moving average, but more importantly, the possibility of smallholder "farmers clubs" being allowed to by ass ADMARC completely by having direct access to the auction floor was to be studied: they would make their own transport and agency arrangements and thus avoid completely any use of ADMARC's services. 20. Bank Memo, June 8th, 1989. This reform was to apply to all smallholder tobacco sales, not just burley. 21. Letter of Development Policy, para. 9. 28 3.6 In addition to the historically good relations established between Bank staff and Malawian officials, the smooth processing of this credit was no doubt assisted by the clear commitment of the Bank to extend the credit and the "carrot" getting bigger. Proposed as a $50 million credit, it was appraised as a $60 million credit and negotiated as a $70 million credit (SDR 52.6 million). Table 3.1. Possible Scenario for Transfer of Burley Growing to Very Smallholders (000 Tons) Year (Buying Season) 1990 1991 1992 1993 1994 1995 1996 Estimated Demand' 52.20 56.10 58.91 61.86 64.95 68.20 71.60 Estimated Demandb 60.00 60.00 60.00 61.85 64.95 68.20 71.60 Allocation to Estates 60.00 60.00 60.00 60.66 62.45 62.20 62.60 Allocation to Smallholders -- 0.2 0.5 1.2 2.6 6.0 9.0 Number of Smallholderse 2,000 4,000 12,000 25,000 60,000 90,000 TCC original estimates b TCC and export buyers revised estimates c'Assuming each smallholder is allocated 100 kg. Board Discussion 3.7 The staff introduced the President's Report by observing that, though notable results had been achieved in improving external and internal balance under the earlier three SACs, yet they had not achieve a significant broad-based supply response in agriculture. The ASAC was designed to remedy this problem. Having summarized several of the initiatives the staff concluded: "7. The smallholder's supply response could also be enhanced by allowing them to grow high-valued crops such as burley tobacco, which generated relatively large cash incomes from a small piece of land. However, by law, only estates were permitted to grow this crop. 22" 3.8 This response was an understatement. The PCR (para. 5.1) estimates that in 1990 a hectare of burley tobacco returned a labor income of Kw 3.65 per day for 1,260 days or a total labor income of Kw 4,599, versus a hectare of hybrid maize which returned Kw 2.59 per day for 163 days of a total of Kw 422. A more than ten-fold increase in labor income. Bank staff have recently expressed the view that smallholders would be limited in their switch to tobacco due to serious structural constraints: "Underlying the audit is the view that Malawi's smallholder agriculture faced no serious structural constraints....The audit's conventional price policy prescription ignore the facts that in Malawi: (i) most smallholders have one or less hectares of land, virtually all of which is under cultivation (ii) the little unutilized arable land comprises either marginal/fallow land or forest/woodland and cultivation of which cannot support sustainable development as it would accelerate degradation of the ecosystem 22. This and other quotations in this section come from the Staff Notes of the Board Discussion, SD90-17/1, dated May 16, 1990. 29 (iii) a short season for land preparation-using hand hoe- and a high proportion of farm family members below the economically productive age results in seasonable labor shortages (iv) the majority of smallholders will divert resources to cash crop only after ensuring enough own production of staple food crops, notably maize (v) smallholders have to devote most of the land and labor resources to the production of maize, largely because the variety-flint maize-which households can process and store properly is characterized by very low yields (vi) little resources will be released from maize in the absence of availability of a high yielding flint maize."23 This neatly summarizes key misapprehensions which allowed the importance of highly distorted prices to be downplayed. As smallholders have shown, there was no necessity for them to produce maize before allocating resources to more profitable crops, and land scarcity provides an urgent incentive to make a switch. In this case, this would have meant a switch to burley tobacco with a ten-fold increase in labor income or to other crops with a three- to four-fold increase (if implicit taxes were removed). 3.9 For the most part, Board members welcomed the ASAC and, in particular, congratulated the staff on the effective collaboration with bilateral donors which had characterized the preparation of the credit. Questions focused on poverty, competition in the estate sector and the Bank's earlier research project. 3.10 In response to questions about poverty and why farmers could not grow burley tobacco, the staff responded: "18. ...that the proposed project addressed the issue of poverty alleviation by promoting the development of high- yielding maize varieties and by actions to target the cultivation of burley tobacco by those farmers with less than one hectare of land," and "23. ...by law, smallholders were not permitted to grow burley tobacco. Although this was a very sensitive issue, the Bank had worked to ensure that smallholders would be allowed to grow the lucrative crop. The laws had been changed and approved by Parliament." 3.11 Had the staff chosen to reply in quantitative terms, they could have added that, at the end of the ASAC, it was projected that 5,000 of 1.6 million smallholders would be allowed to grow in aggregate two percent of burley tobacco, rising three years later to 90,000 farmers and 12.5 percent of the crop. 3.12 The staff were also asked whether there was any market mechanism involved in the land allocation within the estate sector and whether the officials who set the rents were not themselves estate leaseholders? It appeared to the questioner that some element of competition was necessary in order to achieve allocative efficiency. The staff responded that: 23. See also "Agricultural Pricing and Marketing Policy in Malawi: Implications for a Development Strategy," Robert Christiansen and Roy Southworth, MADIA Study, November, 1988, page I1, which takes much the same view. 30 "21. ...The Government was.. .committed to the introduction of differentiated rents to reflect variations in land value. However, this could only be done after it had carried out a land classification study." It was, of course, exactly the questioner's point that a market could be expected to set differential values much more quickly and efficiently than any bureaucratic procedure based on a technical land classification study. 3.13 The staff were also asked about the relation of the proposed emphasis on breeding high yielding varieties of flint maize and the research project that the Bank had funded in 1985. The response was: "27. ...in earlier projects, the emphasis had been on improving the physical facilities and the institutional capability for research. The research policy had focused on dent maize that generally offered the highest possible yield," but was less "suitable to the growing conditions and the storage and processing techniques already known to the smallholders in Malawi." A more accurate comment would probably have been that since the earlier project CIMMYT had developed a line of high yield flint germplasm, which was not available when the strategy for the research project was being developed. Effectiveness 3.14 The credit was approved on the 3rd of April, 1990, signed on the I1th and made effective on the 18th. This reflected good liaison between the Bank and borrower. Bank staff clearly defined for Government the steps needing to be taken for effectiveness. These steps reflected the agreements reached at negotiations (i.e., there were no surprises), and the officials, for their part, had prepared and shepherded the necessary legislation through Parliament and in front of the necessary signatories. In retrospect, it is evident that the conditions of effectiveness were in the nature of "opening doors" or permitting new actions to be taken. Second tranche release required that the permitted actions actually take place and understandably was less easily achieved. 31 4. Implementation Macroeconomic Reforms 4.1 No significant problems were encountered in the implementation of the macroeconomic policy reforms, with the possible exception of Public Sector Investment Plan (PSIP). Even in this case the delay encountered seems to have been due to the press of other tasks rather than any disagreement with the Bank as to the inherent need for a PSIP. The "active exchange rate policy" designed to reflect differences in the inflation rate between Malawi and its trading partners turned out to be a less than active policy. It resulted in a seven percent devaluation in March, 1990, and two further devaluations, of 15 percent and 22 percent in 1992. These latter devaluations left the Kwacha (Kw) 13 percent overvalued as compared to the free market; in other years it was from 25 percent to 32 percent overvalued (see inside front cover). By early 1991 prior approval for foreign exchange expenditures had been removed from all items, except a narrow list of luxury items. Both the overall budget deficit and current expenditures as a proportion of GDP were reduced from 1989/90 to 1991/92. The banking system was made more competitive, an auction system for official debt was introduced and the Reserve Bank of Malawi's rediscount rate was linked to the auction. Progress was made in developing a more effective duty drawback system for exports. 4.2 The initial PSIP presented to the Bank in October 1990 was deficient both in exceeding the fiscal targets set in the Policy Framework Paper and in that the outlays for the social sectors were found to fall short of assurances previously given to the Consultative Group (of donors) for Malawi. A revised review, responding to concerns expressed by the Bank and other donors, was received in February, 1991, and accepted, albeit with some concerns about the level and composition of transport sector investment. Sectoral Reform 4.3 Here too, the majority of policy changes were implemented as agreed. Fertilizer subsidies were reduced to only 1.4 percent of total Government expenditures in 1992/93 and totally eliminated on low analysis fertilizer, thus skewing assistance as agreed towards high analysis fertilizers. Price reviews were submitted to the Bank as agreed, though as the PCR observes, "it is not clear that this process has had an impact on the economic efficiency of the official pricing regime or on improving the smallholder incentive framework" (para. 5.10). Private trade in seed cotton was legalized, and a premium of 20 percent was paid to private traders or farm clubs for deliveries of at least one metric ton of maize. ADMARC has continued to divested itself of its non-marketing functions (a program that was under-way with USAID sponsorship prior to ASAC). A Memorandum of Understanding (MOU) was to be signed between the Government and ADMARC, by which ADMARC would be reimbursed for its non- marketing developmental and food security activities. It was also to redefine ADMARC's role as buyer and seller of last resort, though further reforms were still needed before the private trade in inputs and outputs could be expected to emerge. A new system of quarterly reimbursement of the SFFRFM for subsidized fertilizer has been introduced; and a system of rolling five-year projections of fertilizer requirements has been implemented in conjunction with the donors. New varieties of high yield flint maize are being developed, and three have been released. Legislation 32 to limit, albeit modestly, the transfer of land from customary tenure to the estate sector has been introduced. The rental rates for estate land have been tripled to retain their real cost at the 1985 level, and provision for rental collection from tobacco sales has been introduced. Little progress was made in targeting agricultural credit and fertilizer subsidies to food deficient households or on the unbudgeted activities of policing the requirement that estates plant woodlots and of monitoring. 4.4 The only serious challenges to the concept of the ASAC occurred (a) with respect to tobacco marketing, in general, and burley tobacco production and marketing, in particular and (b) the redefinition of ADMARC's role. Under the heading "Improved Smallholder Access to High Valued Crops," the Letter of Development Policy, on the basis of which the credit was advanced, undertook to (a) initiate a scheme for the growing of burley tobacco under customary land tenure, (b) to study the option of smallholders selling their tobacco directly to the auction floor, (c) ensure that smallholder burley licenses would be targeted to ensure that 75 percent of the recipients had less than one hectare and (d) ensure that by December, 1990, a total of 5,000 smallholders would have been licensed to produce burley on customary land. In addition, the letter undertook that by December, 1990, ADMARC would implement a two-payment scheme for smallholder tobacco, involving a pre-season announced price equal to at least 45 percent of a three year moving average of auction prices and a second payment that would be based on actual realization less ADMARC's overhead and marketing expenses attributable to smallholder tobacco and a reasonable return to ADMARC. 4.5 Fulfilling these undertakings were included in the pre-conditions for release of the second tranche of the credit. This had been expected to take place in December, 1990. A supervision mission in September, 1990, found that no action had been taken with respect to the above undertakings, and, in the follow-up letter, the Bank made suggestions as to how these requirements might be met and reminded the Government that second tranche release depended on the Bank being satisfied with progress in carrying out the program. 4.6 A follow-up supervision mission in December, 1990, had to report that the Government did not intend to apply to burley tobacco the agreed price policy for smallholder tobaccos and did not want to pre-announce the likelihood of a second payment and formula for the payment. This position was confirmed by a telex to the Bank on January 10, 1991: "It has been resolved that we separate the treatment between burley tobacco and other tobaccos. With regard to other tobaccos, it has been decided that smallholders will receive the pre-planting price based on at least a three year rolling average. In addition, they will get a bonus. Details of the bonus will be faxed to you separately after today's meeting to consider the issue. However, there will be no pre-announcement for the fear that tobacco prices are volatile, and it would therefore be inappropriate to raise false expectations which could result in mistrust." "Regarding burley tobacco which has been an estate crop until now, it has been decided that we test the market in the context of liberalization by helping smallholder burley growers to bring tobacco directly to the auction floor, using an agent. The smallholder will thus obtain a price set by the auction floor. The agent will only deduct fees, and the remaining proceeds will go to the smallholder. We believe this is a better arrangement than the original concept. 33 We are in the process of working out modalities of creating such an agent arrangement." 4.7 While no doubt put forward in good faith, this was a clear departure from the letter of development policy and the basis on which the credit had been approved. Gone was the study of the opportunity for non-burley tobaccos to bypass ADMARC's monopoly marketing channel, and gone too was an alternative marketing channel for burley tobacco pending evolution of the agency system. For many smallholders, the initial payment at harvest would be decisive; they would not be able to hold-out for a later payment, even if it was substantially bigger. 4.8 The Bank rejected this proposal as not responsive to the Development Credit Agreement, while expressing its receptivity to alternative proposals which would meet the spirit of the original agreement. Pending agreed actions, acceptable to both sides, the release of the second tranche would be held in abeyance. 4.9 In late May, the government sent a high level mission, led by the Minister of Finance and including the Governor of the Reserve Bank, to Washington to try to resolve the impasse. A background briefing note prepared by Bank staff identified the Government's real concern with the 45 percent formula, that it "would 'de-stabilize' the estate subsector since the resultant price would be higher than those paid by estates to their tenants. Instead of 45 percent, (the government proposed) the first payment should be 27 percent of the three-year moving average of auction prices," and that the second payment should return to producers half of ADMARC's profit (after deducting direct marketing costs and overhead) on selling their tobacco. (Memo, April 5, 1991). The relevant price comparisons are shown in Table 4.1, and Table 4.2 puts these prices in the context of the Government's general price policy for smallholder tobaccos. Policies which as previously remarked (para. 1.3) largely explain the abject poverty of estate tenants and most smallholders. Table 4.1. ASAC and Government Proposals for First Payment for Burley Tobacco Prices, 1990/91 Season 3-Year Average Using Using Maximum Auction Prices ASAC Formulaa Government Formulab Tenant Prices Grades K/kg K/kg K/kg K/kg 1 6.58 2.96 1.78 1.75 2 5.33 2.40 1.44 1.40 3 4.27 1.92 1.15 1.15 4 3.29 1.48 0.89 0.90 X 2.00 0.90 0.54 0.46 45 percent of three-year moving average of auction prices b 27 percent of three-year moving average of auction prices 34 Table 41. Malawi: Tobacco Production and Prices, 1970-1991 Year NDDF SDDF SunlAir Oriental Flue-Cured Burley Prod Price Price Prodo Price Pricec Prod' Price Price Prod Price Price Prod' PriceA Prod Price Priced 1970 9,985 58.60 22 56.35 1,841 59.33 74 39.20 4,676 83.67 5,675 62.57 1971 11,907 83.05 22 75.86 2,324 67.17 134 46.76 6,409 90.76 5,665 52.62 1972 13,462 62.26 22 55.95 2,916 61.20 805 39.02 8,648 89.15 5,453 55.56 1973 12,899 59.22 22 50.11 2,354 49.27 1,056 40.65 9,990 129.26 6,045 80.87 1974 9,232 95.53 22 85.87 1,709 84.85 2,392 38.14 10,522 148.90 5,372 109.46 1975 10,132 147.58 27 138.23 1,689 152.63 1,012 41.16 14,899 129.76 7,997 93.32 1976 38,136 178.13 30 143.32 1,871 175.82 1,301 47.57 16,171 147.82 7,212 103.35 1977 18,048 195.61 34 587 176.77 30 2,370 181.11 568 46.78 20,185 172.16 9,979 137.35 1978 15,360 106.95 46 1,042 79.41 44 3,293 71.87 568 46.78 21,773 171.25 11,567 115.72 1979 11,064 108.64 46 1,223 74.01 44 2,812 72.95 445 47.95 25,356 158.31 15,921 107.72 37 1980 8,616 124.49 46 841 88.28 44 1,344 85.93 335 51.64 26,301 100.95 16,686 117.74 37 1981 9,928 174.97 46 897 131.29 44 1,330 137.31 558 53.19 19,714 175.33 18,804 231.61 41 1982 6,521 343.69 52.08 604 199.64 43.69 1,182 267.17 44.42 615 66.47 22,103 212.99 27,602 216.24 45 1983 7,576 287.54 75.64 808 197.39 67.13 664 234.38 64.09 na 80.75 21,659 187.08 41,537 130.71 47 1984 14,669 215.38 83.74 1,496 174.96 77.52 1,939 229.60 70.24 334 80.31 24.912 222.52 29.979 172.49 48 1985 16,182 151.53 102.03 2,000 95.16 81.50 1,870 106.57 84.60 129 90.87 22,281 237.54 30,372 181.55 51 1986 13,178 225.79 101.52 1,279 172.63 71.48 954 170.45 80.67 121 92.10 21.034 302.75 30,190 291.40 52 1987 11,870 324.83 105.73 625 235.21 82.44 981 266.27 83.70 97 95.37 24,464 395.97 36,789 396.33 78 1988 7,369 449.52 111.60 792 388.65 88.30 560 399.29 103.70 163 129.43 20,743 528.27 45.544 524.47 88 1989 6,279 604.80 157.70 509 629.89 119.20 372 505.23 148.20 300 187.61 20,000 652.87 61,212 369.97 118 1990 10,966 436.76 250.71 403 365.21 186.50 1,396 382.29 214.30 214.65 21.819 652.87 64,019 517.25 118 1991 21,775 602.79' 250.71 593 577.86c 215.00 1,075e 701.44e 214.30 529e 782.77' 229.71 22,710' 1,093.6e 58.694 1,1 10.4e Notes: a Production data (in metric tons) for NDDF and SDDF for 1977-84 are from the report by AGMMARK (UK), entitled "Malawi: Study of Estate and Smallholder Tobacco Growing and Marketing; The World Market for Tobacco from Malawi" Draft, November 1989, Table 6. From 1985 to 1991, production data for these two tobaccos are from the Ministry of Agriculture Planning Division, "Crop Production, Price and Gross Margin Data on the Major Crops in Malawi." For all the other tobaccos, production data for 1970-1990 are from the Ministry of Agriculture Planning Division, "Malawi Agricultural Statistics: 1991 Annual bulletin," Tables 2.24 and 2.25. For 1970-76 (inclusive), the production recorded under NDDF is for all fire-cured tobacco, of which, in subsequent years, NDDF made up more than 92 percent. It is possible that the figure recorded for 1976 production of all fire-cured tobacco is in error. b Auction prices in Tambalalkg for 1970-90 are from the Ministry of Agriculture Planning Division, "Malawi Agricultural Statistics: 1991 Annual Bulletin," Table 7.5. The latter records that Oriental tobacco is not auctioned. ' Producer price data in Tambala/kg for the first four tobaccos for 1982-91 are from the Ministry of Agriculture Planning Division, "Crop Production, Price and Gross Margin Data on the Major Crops in Malawi;" for NDDF from 1970 to 1981 and SDDF for 1977 to 1981, the producer prices are an estimated weighted average of official prices for various grades, using weights in the aforementioned publication and official prices from the Ministry of Agricultural Planning Division, "Malawi Agricultural Statistics: 1991 Annual Bulletin," Table 7.5. For Oriental tobacco from 1970 to 1981, rroducer price data arc from the latter report, Table 7.5. Officially stipulated maximum prices at which estates may buy from tenants; Source: A. Duncan. "The Impact of Pricing Policy on Tobacco Tenancy in Malawi." Oxford food Studies Group. April, 1990. Appendix 3. ' Price data for 1991 are from the Tobacco Exporters Association of Malawi; 1991 data in the "Production" column for Sun/Air dried, Oriental, Flue-Cured and Burley tobacco are.actually Exports.of the respective tobaccos in that year. 35 4.10 Prior to these discussions, it seems doubtful whether either the Government or the Bank had fully appreciated the extent to which the prosperity of the estate sector depended on the poverty of the smallholder sector: only if incomes in the smallholder sector were (artificially) depressed would tenants come forward to work for little more than a quarter of the auction price of the tobacco they produced. Some Bank staff dispute this judgment (see footnote 14), pointing to the Country Economic Memo and various papers written for the Bank supported MADIA research project (Manager's Agricultural Development in Africa) as evidence. Box 2 provides relevant quotations in support of the staff view, but these do not, in the audit's view, carry much weight compared to the misdiagnosis already quoted in footnote 8. 4.11 The Ministerial mission was successful in reaching agreement with the Bank that: i) For 1990/91, the initial payment would be 45 percent of the three-year moving average for all smallholder tobaccos except burley; for burley, it would be 27 percent; ii) In subsequent years it would be 45 percent for all smallholder tobacco marketed through ADMARC; iii) The second payment would give smallholders 65 percent of the profit ADMARC had made on selling their tobacco after deducting direct and proportionate overhead expenses; and iv) From the 1991/92 season on, smallholders could sell their tobacco to ADMARC, or an Agent though a farmers club, or to an estate (i.e., a multiple channel marketing system). 4.12 Given that the 1990/91 season was for all practical purposes over24 and given the quite exorbitant marketing margins customarily charged by ADMARC, this was an acceptable compromise. 4.13 In the event, the first payment was about Kw 1.5 per kilogram for burley tobacco which sold for as much as Kw 7.0 at auction.25 Needless to say there was considerable resentment from smallholders. Approximately half the smallholder producers, when faced with this low initial payment and no assurance of (or even notification of government's intent to provide) a second payment, decided to sell their tobacco to an estate for resale at auction. There is no information as to the prices at which these sales to estates took place. After some prompting, a second payment was announced on December 20th. Even in this case, the announcement did not include the amount of the second pa Ment: smallholders were to wait to be informed by ADMARC as to the extent of their payment. The payment was in fact based on Kw 2.75 per kg; almost twice the initial payment. Burley tobacco sold through the auction by estates (some of it originating from smallholders) did not, of course, qualify for the bonus payment. Despite this disappointment, smallholders applying for burley tobacco licenses increased from 8,000 in the 1990/91 season to 11,000 in the 1991/92 season. 24. Quota Licenses for 3,000 tons of smallholder burley tobacco were issued in 1990/91 (PCR, Table 7). 25. Memo, 20th November, 1991. 26. Daily Times, Malawi, 20th December, 1991. 36 Box 2. The Smallholder Estate Nexus as Seen in the Late 1980s From the Country Economic Memorandum, "Malawi: Growth Through Poverty Reduction," World Bank Report No. 8140-MAI, March, 1990. * "...[E]xtension of licenses to grow certain high value crops to smallholders, in particular burley tobacco, is expected to provide additional income for some poor households..." (page iii). * "Smallholder share in tobacco export value has declined...primarily because of the increasingly poor price incentives. Until recently, ADMARC, as the monopsonistic buyer of smallholder tobacco, passed through, on average, considerably less than world price to farmers" (para. 2.9). * "...[P]oorer farmers generally concentrat[ed] on traditional maize and root crops and with limited adoption of improved maize varieties and other cash crops such as groundnuts. The resulting low productivity of smallholder labor spills over into the behavior of other labor market for estate production, services and industry, placing no upward pressure on wage levels" (para. 3.16). * "Limited access to especially lucrative opportunities also contribute to poverty: for example, the lack of access to burley tobacco licenses for smallholders until recently" (par. 3.18). From MADIA Research Reports Report #6, "Smallholder and Large-Scale Agriculture in Africa: Are There Trade-offs Between Growth and Equity?" (1989) * "Smallholders are required to sell their output to...ADMARC at government-determined prices, which have tended to be a third and a half of the price that estates received for the same type of tobacco. These distributions reflect Malawi's development strategy of stressing estate agriculture" (page 8). * "This paper has demonstrated that Malawi's policy of requiring smallholders to grow only lower-quality tobacco and to sell their production only at government-established prices... There is no excuse for continuation of this policy" (page 16). Report #9, "Structural Adjustment, Agricultural Development and the Poor: Lessons from the Malawian Experience" (1989) * "Malawi is a well-managed economy which has suffered from tremendous external shocks....Its problems of adjustment, however, have been made difficult by dualism-within--dualism--a structure by which Malawi's agriculture sector is sharply divided by legal restrictions" (page 16). * "...[A] strategy for sustainable, broad-based growth in Malawi...requires inclusion of the poor and involves ...policies... aimed at...overcoming the dualism rooted in the estate strategy and its attendant income and asset disparities" (page 16). Report #11, "Markets, Marketing Boards, and Cooperatives in Africa: Issues in Adjustment Policy" (1989) * "...[I]n most MADIA countries there was not net taxation of agriculture....While the government of Malawi taxed smallholder tobacco, it subsidized smallholder maize and fertilizer" (page 7). By contrast the Aide Memoire of the December 1990 supervision mission makes a clear connection between smallholder-price and estate tenants: "...[Tlhe Government has decided that the pre-planting price based on at least 45 percent of a three year rolling average of auction prices will not apply to burley tobacco and that the applicable price will be the same as paid to estate tenants....[S]ince 45 percent would result in a price higher than the one payable to tenants, estates would find it difficult to retain tenants..." (para. 2). 4.14 The above second payment was for burley tobacco, while the two price scheme agreed for ADMARC was for all tobacco: "(b) Level of Second payment for Smallholder Tobacco ADMARC will pay to smallholder farmers selling tobacco through ADMARC a second payment of not less than 65 percent of net revenues after allowing for ADMARC's marketing 37 costs and overheads related to the marketing of smallholder tobacco. This payment formula will apply for any smallholder tobacco sold through ADMARC in the 1990/91 marketing season and thereafter; (c) Timing of Second Payment to Tobacco Growers All smallholder farmers selling tobacco through ADMARC will receive a second payment from ADMARC in October-November. This timing will apply for the 1990/91 marketing season and thereafter."2 4.15 It appears from the files that this undertaking by the government to provide a second payment to non-burley tobacco producers was never announced publicly by the government and was overlooked by the Bank.28 4.16 Significant cofinancing was provided by other donors (15 million pounds from the United Kingdom, and five million guilders from the Netherlands), and support for the continuation and expansion of the policies, introduced under ASAC, was provided by a follow- on USAID-funded ASAP, a singularly successful project. Its success can be traced in part to insistence on actions rather than on promises as a condition for tranche release-an insistence perhaps inspired by the Bank's experience with difficulty in getting follow-through after a tranche release. ASAP also benefited from a more open and flexible policy environment following the change in Government. EDDP also pushed for policy reform in parallel with ASAP. The audit has not formed a view as to the performance of the follow-on project. 4.17 ADMARC. The PR made provision for "further actions to ensure that the non- commercial functions that the Government wants ADMARC to undertake are adequately funded in a transparent manner. The Government and ADMARC will enter into a formal agreement...." (para. 80). It is worth quoting quite extensively from the PCR's account of the outcome of the above undertaking, since the audit found an even less satisfactory situation than described. "5.14 (c) Though the MOU was signed only in December 1993, its implementation began before then. In 1990/91, Government reimbursed ADMARC Kw 15.0 million for its losses in carrying out development and food security programs. In 1992/93, though the Government had reportedly budgeted Kw 20.0 million for this purpose, it made no transfers to ADMARC, due in large part to revenue generated from the handling of commercial/drought relief maize (MK 20.6 million)." "5.15 Evaluation. ADMARC and Treasury should be given credit for the development of the draft MOU agreement. Nonetheless, this area of the ASAC 27. Letter from Ministry of Finance to the Bank, August 6, 1991. 28. The Aide-Memoire for a supervision mission in November, 1991, makes reference to the second payment in terms of "Government agreed that ADMARC will pay to smallholders selling tobacco through ADMARC a second payment..." (Aide-Memoire, November 25, 1991). However, the Bank's follow-up letter refers to "Second Payment for Burley Tobacco Smallholders" (letter, November 25, 1991, emphasis added in both cases). It is hard to comprehend why the scope of the government's undertaking was thus narrowed by the Bank, if done consciously, or ho it could have been done unconsciously. In the event, ADMARC initially refused to make the second payment for non-burley tobacco, but was persuaded to make a payment under the conditionality of a later credit. 38 adjustment program has been prolonged and delayed beyond the Bank operation due to difficult and lengthy negotiations. There appear to have been two major causes in this delay, namely the cost to Government of ADMARC's interventions on its behalf and ADMARC's lack of confidence that it would indeed be reimbursed in full for the losses incurred...." 4.18 The PCR may well be correct that the MOU was finally signed in December, 1993; however, the document itself is undated, with only a date-stamp from Treasury for February 21, 1994. The stamp has the appearance of being a record of official receipt of the MOU, rather than as being an integral part of the document. Since the project closed at the end of 1992, either date agrees with the PCR's observation that this area of the ASAC was delayed beyond the Bank operation. While the Bank may have continued to have some leverage, the project, per se, did not. 4.19 The MOU (para. 2.11) calls for Government payments of MK 28.5 million in 1990/91, MK 35.4 million in 1991/92 and an estimated MK 20.2 million in 1992/93. It also lays out the principles for reimbursement in future years. Nowhere does it provide for offsetting against Government liabilities-profits that ADMARC may have made in other areas of its operations. It took four years to negotiate the agreement, and, after it was signed, the Government treated it as "a scrap of paper" with the Bank's implicit endorsement. 4.20 To emphasize the point, Table 4.3 compares money due ADMARC under the MOU with amounts actually paid. Thus, ADMARC had reason to be skeptical whether it would be reimbursed in full; and given the magnitudes of the sums involved, no wonder the Treasury was reluctant to assume responsibility for costs which had traditionally been off-loaded off-budget onto the backs of the poorest of the poor. Table 4.3. Budgetary Transfers to ADMARC, Due and Paid (Millions MK) Year Due Paid 1990/91 28.5 15.0 1991/92 35.4 0.0 1992/93 (Est.) 20.2 0.0 4.21 The Bank has to be given credit for achieving, in this instance, its transparency objective. What it reveals, unfortunately, is the Bank once again mistaking an agreement on paper for results on the ground.29 29. The audit says "again," since even within this project the Bank was surprised that the "unfunded mandates" to enforce the forestry provisions of estate leases and to do monitoring and evaluation were not implemented. 39 5. Outcomes Macroeconomic Policy 5.1 Macroeconomic policy under the ASAC was essentially a continuation of the policies implemented under the three earlier structural adjustment operations. Despite the three devaluations, the official exchange rate remained well above the market rate. Domestic macroeconomic indicators performed well in 1991, with a GDP growth of 7.8 percent (up from a "good" 4.8 percent in 1990). The volume of merchandise exports increased by 51.4 percent over 1990. Despite this, the import boom associated with a more liberal import policy resulted in the external account deficit rising from 7.8 to 12 percent of GDP in 1991. This is not a very surprising result, given the over-valued exchange rate. 5.2 Provision for the Bank to review the Public Sector Investment Plan (PSIP) was effective in holding total planned expenditures to levels earlier agreed with the donor Consultative Group (and hence more or less consistent with expected aid flows). 5.3 The gains made in 1991 were not continued into 1992 due to the worst drought in a century and a halt to non-humanitarian aid from the donors. The drought was well beyond the government's control (other countries in Southern Africa were similarly affected) and resulted in maize production dropping from 1.6 million tons to 0.66 million tons for the 1992/3 season with similar reductions in other crops. The elimination of non-humanitarian assistance traced to the donors' dissatisfaction with Malawi's record on Human Rights. Sector Policies 5.4 The biggest policy impact of the ASAC was clearly in giving smallholder's access to the market for burley tobacco. There were three prongs to this policy initiative. The first was to allow smallholders to grow burley; the second was to allow fanner clubs to bulk their production and sell it directly to the auction floor; and the third was an agreed profit-sharing formula for pricing burley tobacco sold through ADMARC. By 1992/93, 27,400 smallholders were marketing 7,000 tons of burley tobacco. A very significant outcome of this policy change has been to lay to rest the canard that smallholders cannot produce burley tobacco of an acceptable quality. In practice, the quality of smallholder tobacco was better on average in some years (1990/91 and 1992/93) than that produced by estates. 5.5 A companion improvement for smallholders was an agreed formula which gave them 45 percent of a three-year moving average of auction receipts for non-burley tobaccos, albeit at the cost of ADMARC continuing to be the monopsonistic buyer. 5.6 In practice, the above significant achievement of the ASAC has been made redundant as the liberalization of tobacco marketing started by ASAC and vigorously exploited by ASAP and EDDP has begun to bite. With intermediate traders changing their tobacco prices on a weekly 30. Never a very appealing argument, even on theoretical ground, since the vast majority of estate tobacco is produced by small tenant farmers, each cultivating less than a hectare. 40 basis in the light of auction receipts, there is no place for formula pricing. In order to compete effectively, ADMARC too has had to move to a system of weekly price setting. 5.7 Other important outcomes have been: i) ADMARC's divestiture program for non-marketing activities, started under the earlier SALs, was essentially completed. ii) The emphasis in maize breeding on hybrid flint varieties (initiated under guidance from CIMMYT prior to the ASAC) has yielded three varieties with high acceptance by farmers so that use is currently constrained by supply of seed rather than farmer acceptance. 5.8 Less important outcomes include: i) Government's submission of its draft annual price review for Bank comment three weeks prior to its finalization (increasingly a non-event in a world of market competition). ii) The encouragement of "private trade in grain" by paying a 20 percent premium for deliveries of over one-ton of maize to ADMARC depots by traders or farmers' clubs, iii) Packaging of fertilizer in 10 Kg, 15 Kg and 25 Kg bags for the convenience of smallholders unable to finance the normal 100 Kg bags. Demand for 10 Kg and 15 Kg bags has been minimal, and they have been discontinued. iv) Private trade in seed cotton was legalized.31 v) A system of five-year rolling fertilizer projections was introduced by SFFRFM, and formulae for drawing down the fertilizer stock were established. vi) Fertilizer subsidies were to be reduced from 30 percent (economic subsidy) in 1990/91 to 25 percent in 1991/92 and 20 percent in 1992/93. The total value of these subsidies was not to exceed two percent of total Government expenditure in 1990/91, 1.6 percent in 1991/92 and 1.3 percent in 1992/93. These objectives were essentially met, with the subsidy entirely removed on low and medium analysis fertilizers, and the actual subsidy reduced to 1.39 percent of Government expenditures in 1992/93. In project design, this was seen as primarily a macroeconomic policy consideration, but is probably correctly interpreted as primarily an agricultural sector policy, given the modest amount of total expenditure involved (see, however, para. 5.10) vii) The conditions under which land could be transferred from customary title to the estate sector were made more restrictive. 31. Under ASAP, ADMARC has introduced a custom ginning service, enabling farmers clubs to choose between selling their lint or seed cotton. 41 viii) Land rents were raised for the Estate sector from Kw 10 per hectare to Kw 20 and to Kw 30 in 1993/94 so as to maintain the rents at their 1985 real value. ix) A separate estate extension service to be financed by the Tobacco Association of Malawi was established. (These latter three changes were designed to improve the efficiency of land utilization in the estate sector.) x) To encourage private (and farmer club) marketing, fertilizer was made available at a "wholesale" price 20 percent below retail for orders of one ton or more. A fertilizer market liberalization policy was finally announced in May 1993. Some on-selling of fertilizer imported by SFFRFM by the private sector has occurred. To date, no importation by the private sector for sale to smallholders is known to have occurred. 5.9 Unresolved issues included: i) The promised reorganization of the inspectorate staff of the Department of Lands to allow them to enforce the requirement that at least 10 percent of estate land be planted to woodlot has been approved. It remains to be implemented. ii) A parallel United Nations Development Program (UNDP)- funded project executed by the Bank to test ways of targeting agricultural credit and fertilizer subsidies to the very poor demonstrated the difficulty of such programs but did not show how these difficulties could be overcome at reasonable cost. iii) As agreed under ASAC, a review of expenditures by the Ministry of Agriculture was made in 1990; however, the results were felt to be unsatisfactory by the Bank. Of particular concern was the revealed trend to reduced amounts of non-salary expenditures per employee with the result that they would become increasingly office-bound. Despite urging from the Bank, Government has not yet produced a report or plan which adequately tackles this problem. 5.10 After over a decade of urging by the Bank the Government has finally eliminated its fertilizer subsidy targeted to smallholders in the present 1995/96 season.32 Simultaneously, the Bank has agreed33 that funds from the Agricultural Services Project (Cr. 2514) can be used to fund fr= distribution of 23,000 tons of fertilizer to very smallholders. It would appear that the Bank's arguments against the subsidy targeted to smallholders was not against such subsidies per se, but rather that the subsidy was not selective enough. 32. This, in conjunction with the floating of the Kwacha, has tripled the cost of fertilizer to smallholders. 33. At the urging of the IMF, which made free distribution of inputs a condition for continuation of assistance of its ESAF (Economic Stabilization and Adjustment Fund). 42 5.11 As brought out in para. 3.8, in 1990 the labor income from burley tobacco per hectare was Kw 4,599 versus a labor income per hectare for maize of Kw 422 per hectare. For the Bank to support, even by the provision of "free" inputs, a program designed to increase the attractiveness of the lower income alternative is perverse: Inputs which are free to the farmer but increase national indebtedness. The files read as if subsidy elimination was an issue for macro- policy, not directly related to better targeting. 43 6. Ratings 6.1 The PCR rated the project outcome as unsatisfactory, institutional development as modest and sustainability of the completed reforms as likely. On balance the audit confirms these ratings. 6.2 Borrower performance is rated as satisfactory, since almost all project conditionalities were complied with or changed by negotiation. Significant failures included the monitoring and evaluation component and the implementation of estate woodlots. Bank performance is rated unsatisfactory. In the narrow context of the project, even the quite modest target of returning to smallholders ADMARC's profits on sale of their tobacco was abandoned. In the larger context of project design, the Bank did not persuade Government to permit the unrestricted export of cash crops and thus to address the primordial problem of excessive implicit taxation of smaltholders and tenant farmers. As noted in paras. 4.4 to 4.12, the Bank performed well in holding back the second tranche until an acceptable pricing policy for smallholder burley was agreed. Actual implementation of the bonus payment, however, had to await further negotiations by USAID in the context of the ASAP.34 6.3 As suggested elsewhere in this audit, operational staff involved with the project have a different perspective on the release of the second tranche. They believe that the project was overtaken by events, and the conditions, particularly political conditions which applied during appraisal and at Board approval, were no longer applicable. Holding up release of the second tranche for ten months served notice that the exclusion of smallholders from burley production was no longer an acceptable policy. Small though the number of farmers and quantity of tobacco were, this signaled the beginning of the end of the Estates' capacity to exploit tenants. Moreover, the more assertive position being taken by bilateral donors (quite unforeseen at the time of Board presentation) risked a complete break between donors and the Banda government. Without in any way suggesting that the other donors should soften their increasingly tough position, the Bank itself agreed to the release of the second tranche as a means of keeping communications open between donors and the Government and hence keeping open the possibility of a "velvet" transition. 6.4 In this view, the transition to a political regime which would remove all constraint on smallholder tobacco production, and allow ADMARC to be bypassed, was much more important than the amount ADMARC was to pay in one year for a relatively small quantity of tobacco. 6.5 Clearly, the second tranche was not subject to the "blue-print" conditionalities and legal covenants presented to the Board; rather, it was used to facilitate the "process" of policy (and implicitly political) reform. Knowing what they know now, the concerned staff say they would have no hesitation in endorsing the original presentation of the project to the Board, or the way it was modified from a "blue-print" to "process" approach during implementation. In short, even with hindsight, and even having taken account of the point emphasized in this audit, they would argue that both project outcome and Bank performance were satisfactory. 34. As Part II of the PCR says (i.e., after project closing and long after release of the second tranche), "8.4 Government is still having problems getting a second payment from ADMARC to tobacco farmers prior to the beginning of the agricultural season so that farmers can buy inputs."  45 7. Findings and Issues 7.1 OED methodology, by prescribing topics which must be covered, providing for peer review and involving a comprehensive annual review of OED's work, goes a long way to constraining the subjectivity, which is often attributed to evaluation. Less than three years after project closing, agricultural policy in Malawi has been transformed, and the first fruits of improved policy are being harvested. If the audit were to accept the argument "Post hoc, Procter hoc, " then the ASAC could only be rated "highly satisfactory." 7.2 Regretfully, based on OED methodology, the audit cannot take this view. The relevance of the credit's original goals was modest indeed. Even so, they were not met. The lion's share of credit for policy change has to rest with (a) the changed political climate in Malawi and (b) the extraordinarily effective impact of the ASAP and possibly EDDP. By contrast, the Bank in the context of ASAC gave away the bonus payment to smallholders for non-burley tobaccos (footnote 34), accepted a one-year delay in the 45 percent formula for smallholder burley, released the second tranche on the basis of an agreement to pay the burley bonus (rather than waiting for actual payment), and acquiesced to the ADMARC MOU being violated almost before it was signed (paras. 4.17 and 4.21). 7.3 A fundamental development policy problem in Malawi (paras. 1.3 to 1.5) rests in the dualism of its agricultural policy. Under the cloak of "export led growth" from the estate sector, it effectively "side-lined" 72 percent of the labor force and 50 percent of the land.35 Prohibited from producing some crops and subject to egregious taxation, there was no way that the smallholder sector, over half of whom lived below the level of absolute poverty, could better themselves. The Bank defined the problem as "remaining structural adjustment constraints in agriculture" (PR, para. 18) instead of tackling frontally excessive agricultural taxation or even the soundness of agricultural price policies on which the Bank had committed itself to be consulted. The only thing "structural" about the agricultural price policy was the excessive profits it was designed to provide for ADMARC, which were then used to cross-subsidize an extensive range of loss-making activities. It says a lot about the real comparative advantage of Malawian smallholders that, even subject to this tax regime, they still produced small quantities of cash crops. No production system could be expected to thrive under such heavy taxation. 7.4 The ASAC chipped away at the low returns to smallholders by giving some smallholders access to burley tobacco licenses and ensuring reasonable prices for them either by bypassing ADMARC or by ensuring that 65 percent of the profit made by ADMARC on smallholder burley tobacco was returned to producers. Other smallholder tobacco growers benefited from the 45 percent of a moving average of market prices provision, but ADMARC was allowed to retain most of the additional profits. For other smallholders, it did not address export limitations but sought a commitment by Government to reduce fertilizer subsidies. To describe such an adjustment program as a "poverty reduction project" is a travesty of the truth, support of hybrid maize breeding, fertilizer planning, restraint on transfer of land to the estate sector, etc., notwithstanding. 35. Sahn, D. E. et al, "Policy Reform and Poverty in Malawi," Cornell Food and Nutrition Policy program Monograph NO. 7, December, 1990, Table 3, p.10 (Quoted in Graeme Donovan, "Agriculture and economic reform in Sub-Saharan Afica"). 46 7.5 The key finding of this audit, with respect to the ASAC and indeed the Bank's decades long development policy dialogue, is that it missed the big issue: whether the smallholder sector was to be given the opportunity to contribute to Malawi's development. 47 8. Lessons 8.1 The lessons from this credit are clear: I. To significantly reduce poverty in Malawi required that the (poor) smallholders be treated at least as well as the (better off) estates. II. Malawi was unable to achieve its full development potential so long as the smallholder sector was paid only a small fraction of export earnings. III. The key to food security is, as this project argued, to allow smallholders to earn a cash income sufficient to be able to afford either to use HYV maize and complementary inputs or to buy maize (i.e., for the land-scarce, food security is incompatible with confiscatory taxation of cash crops). IV. The existence of two systems of tenure provided no reason for limiting the crops which could be grown or the marketing channels that could be used. 8.2 Lesson IV parallels the Government's conclusions in Part II of the PCR: "The Government recognizes that the estate subsector has been protected for a long time and that this has been at the expense of smallholders. It is high time these subsectors were given equal opportunity" (PCR, Part II, para. 7.1.2). 8.3 Lessons I to IV are stated in the past tense since changes introduced under the auspices of the USAID-supported ASAP have ended the discrimination against smallholders. This success of ASAP (as compared to the ASAC) rests very largely on the clear decision by USAID that ASAP disbursements would be tied to actions, not promises. It is true that the change in political regime with the retirement of the President for Life meant that USAID was no longer dealing with an intransigent political regime. Under USAID's management, intransigence would have led to suspension of disbursements. 8.4 A more subtle lesson is: V. The decision to fund a minimal lending program in order to maintain a policy dialogue may preclude significant progress towards policy reforms. By analogy with the Heisenberg Uncertainty Principle, it may be necessary to choose between a policy dialogue and actual implementation of policy changes.36 8.5 The dilemma implicit in the above lesson as to whether to employ a "hard-nosed" approach and to insist on significant policy reforms as a basis for lending or a softer approach of lending on the basis of modest policy reform to maintain a policy dialogue (and the capacity to 36. For an analogous argument that it is in the interest of both borrower and donor to have aid conditionalities which antulMfilled, but where lending continues, see P. Mosley, "The Failure of Aid and Adjustment Policies in Sub- Saharan Africa: Counter-Examples and Policy Proposals," University of Reading (England), Department of Agricultural Economics and Management, Discussion Papers in Development Economics, Series G, Vol. 111 (1995/96). 48 act as honest-broker between Government, other donors and the political opposition) was discussed in paras. 1.22 to 1.27. Even (or especially) with hindsight, Bank staff insist that, in this instance, the softer option of maintaining a policy dialogue was the right choice. The audit does not disagree. In a reversal of the old saw: "the operation was unsatisfactory, but the patient thrived," not because of the changes achieved under the project, but because of the pressure of the political opposition and other donors, and the Bank's ability to work constructively with all parties. 8.6 Policy change should not be seen as an end in itself; policy is in the nature of an "input" to the operation of the economy. Two key output or "impact" variables are already showing the benefits of the ASAC/ASAP policy changes: i) Poverty Measure. A clear indication of poverty reduction is the difficulty the estate sector has in finding tenants. There has been active competition by estates for tenants (evidence of increased prosperity in the smallholder sector). This is increasing the amount tenants are being offered and, hence, reducing their poverty.37 ii) Production Response. The target volume of burley to be sold at auction in 1995 was 99,000 tons. Actual sales were 110,000 tons, and the target for 1996 has been raised to 120,000 tons. This is a 20 percent increase in targeted sales of Malawi's most important export, in one year. 8.7 The enforcement of estate leases with respect to forestry and the introduction of monitoring and evaluation procedures would have required significant outlays by Government, which were not detailed in the President's Report (PR). They were in a sense "unfunded mandates." Thus, even where the Bank is providing program support, and few, if any, restrictions are placed on the use of its funds, the audit concludes that: VI. For adjustment programs, any expenses the Government is expected to incur should be appraised and budgeted for in the PR in as much detail as in a Staff Appraisal Report for an investment project. Without such a detailed description, borrowers may agree to conditions without filly appreciating the financial implications involved. 37. For a number of years, some estates have paid more than the very low mandated returns to tenants. This season has seen active competition. A. Duncan, "The Impact of Pricing Policy on Tobacco Tenancy in Malawi," Food Studies Group, Oxford, April, 1990 (Mimeo).    IMAGING Report No: 15641 Type: PPAR