Document of The World Bank FOR OFFICIAL USE ONLY Report No: ICR00004739 IMPLEMENTATION COMPLETION AND RESULTS REPORT 5101-LR ON A CREDIT IN THE AMOUNT OF SDR 9,700,000 MILLION (US$15 MILLION EQUIVALENT) TO THE REPUBLIC OF LIBERIA FOR THE LIBERIA SMALLHOLDER TREE CROP REVITALIZATION SUPPORT PROJECT June 28, 2019 Agriculture Global Practice Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective {June 20, 2019}) Currency Unit = USD US$ 1.38 = SDR 1 FISCAL YEAR July 1 - June 30 Regional Vice President: Hafez M. H. Ghanem Country Director: Henry G. Kerali Senior Global Practice Director: Juergen Voegele Practice Manager: Marianne Grosclaude Task Team Leader(s): Abimbola Adubi ICR Main Contributor: Jeanette M. Sutherland ABBREVIATIONS AND ACRONYMS AfT Agenda for Transformation BCR Benefit Cost Ratio CBL Central Bank of Liberia CDA Cooperative Development Agency CPF Country Partnership Framework CPS Country Partnership Strategy EIRR Economic Internal Rate of Return ERR Economic Rate of Return EPA Environmental Protection Agency EPO-LIBINCO Equatorial Palm Oil-Liberia Operations Inc ESMP Environmental and Social Management Plan ESMT Environmental and Social Management Team FCV Fragility, Conflict and Violence FFS Farmer Field School FM Financial Management FO Farmer Organization GDP Gross Domestic Product GoL Government of Liberia HIPC Heavily Indebted Poor Country ICR Implementation Completion Results Review IRR Internal Rate of Return ISR Implementation Status and Results Report LASIP Liberia Agriculture Support Investment Program LTCS Long-Term Credit Scheme MARCO Morris American Rubber Company MFDP Ministry of Finance and Development Planning MoA Ministry of Agriculture NPV Net Present Value M&E Monitoring and Evaluation MIS Management Information System PAD Project Appraisal Document PFIs Participating Financial Institutions PIU Project Implementation Unit SMART Specific, Measurable, Achievable, Relevant, and Time-bound SRC Salala Rubber Corporation STCRSP Smallholder Tree Crop Revitalization and Support Project ToC Theory of Change TABLE OF CONTENTS DATA SHEET ...............................................................................................................................1 I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES ....................................................... 6 A. CONTEXT AT APPRAISAL .........................................................................................................6 B. SIGNIFICANT CHANGES DURING IMPLEMENTATION .............................................................. 10 II. OUTCOME .................................................................................................................... 13 A. RELEVANCE OF PDOs ............................................................................................................ 13 B. ACHIEVEMENT OF PDOs (EFFICACY) ...................................................................................... 14 C. EFFICIENCY ........................................................................................................................... 20 D. JUSTIFICATION OF OVERALL OUTCOME RATING .................................................................... 21 III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME ................................ 24 A. KEY FACTORS DURING PREPARATION ................................................................................ 24 B. KEY FACTORS DURING IMPLEMENTATION ............................................................................. 26 IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME .. 26 A. QUALITY OF MONITORING AND EVALUATION (M&E) ............................................................ 26 B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE ..................................................... 28 C. BANK PERFORMANCE ........................................................................................................... 29 D. RISK TO DEVELOPMENT OUTCOME ....................................................................................... 31 V. LESSONS AND RECOMMENDATIONS ............................................................................. 31 ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS ........................................................... 33 ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION ......................... 47 ANNEX 3. PROJECT COST BY COMPONENT ........................................................................... 49 ANNEX 4. EFFICIENCY ANALYSIS ........................................................................................... 50 ANNEX 5. BORROWER, CO-FINANCIER AND OTHER PARTNER/STAKEHOLDER COMMENTS ... 61 ANNEX 6. TABLES REFERENCED IN THE REPORT.................................................................... 63 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) DATA SHEET BASIC INFORMATION Product Information Project ID Project Name P113273 LR Smallholder Tree Crop Revitalization Support Project Country Financing Instrument Liberia Investment Project Financing Original EA Category Revised EA Category Partial Assessment (B) Partial Assessment (B) Organizations Borrower Implementing Agency Ministry Of Finance and Development Planning Ministry of Agriculture Project Development Objective (PDO) Original PDO “To increase access to finance, inputs, technologies and markets for smallholder tree crop farmers in Liberia, and to develop a long term development program for the tree crops sector” Page 1 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) FINANCING Original Amount (US$) Revised Amount (US$) Actual Disbursed (US$) World Bank Financing 15,000,000 15,000,000 13,732,883 IDA-51010 Total 15,000,000 15,000,000 13,732,883 Non-World Bank Financing 0 0 0 Borrower/Recipient 0 0 0 Total 0 0 0 Total Project Cost 15,000,000 15,000,000 13,732,883 KEY DATES FIN_TABLE_DAT Approval Effectiveness MTR Review Original Closing Actual Closing 05-Jun-2012 13-May-2013 03-Nov-2015 31-Dec-2016 30-Nov-2018 RESTRUCTURING AND/OR ADDITIONAL FINANCING Date(s) Amount Disbursed (US$M) Key Revisions 28-Jul-2014 3.64 Change in Legal Covenants 11-May-2016 7.51 Change in Results Framework Change in Components and Cost Change in Loan Closing Date(s) Reallocation between Disbursement Categories 10-Oct-2018 13.31 Reallocation between Disbursement Categories KEY RATINGS Outcome Bank Performance M&E Quality Moderately Satisfactory Moderately Satisfactory Modest Page 2 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) RATINGS OF PROJECT PERFORMANCE IN ISRs Actual No. Date ISR Archived DO Rating IP Rating Disbursements (US$M) 01 26-Dec-2012 Satisfactory Satisfactory 1.03 02 25-Jun-2013 Satisfactory Satisfactory 1.04 03 12-Mar-2014 Satisfactory Satisfactory 2.04 04 15-Jun-2014 Moderately Satisfactory Moderately Satisfactory 2.80 05 10-Dec-2014 Moderately Satisfactory Moderately Satisfactory 3.64 06 18-Jun-2015 Moderately Satisfactory Moderately Satisfactory 5.01 07 28-Aug-2015 Moderately Satisfactory Moderately Satisfactory 6.02 08 20-Dec-2015 Moderately Satisfactory Moderately Satisfactory 6.26 09 18-Jun-2016 Moderately Satisfactory Moderately Satisfactory 7.82 10 23-Nov-2016 Moderately Satisfactory Moderately Satisfactory 8.68 11 15-Jun-2017 Moderately Satisfactory Moderately Satisfactory 11.14 12 21-Jan-2018 Moderately Satisfactory Moderately Satisfactory 12.53 13 26-Oct-2018 Moderately Satisfactory Moderately Satisfactory 13.31 SECTORS AND THEMES Sectors Major Sector/Sector (%) Agriculture, Fishing and Forestry 80 Agricultural Extension, Research, and Other Support 10 Activities Crops 35 Public Administration - Agriculture, Fishing & Forestry 35 Public Administration 10 Sub-National Government 10 Page 3 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Transportation 10 Rural and Inter-Urban Roads 10 Themes Major Theme/ Theme (Level 2)/ Theme (Level 3) (%) Private Sector Development 100 Jobs 100 Finance 15 Finance for Development 15 Agriculture Finance 15 Urban and Rural Development 58 Rural Development 58 Rural Markets 15 Rural Infrastructure and service delivery 30 Land Administration and Management 13 Environment and Natural Resource Management 90 Climate change 64 Mitigation 64 Renewable Natural Resources Asset Management 26 Biodiversity 13 Landscape Management 13 ADM STAFF Role At Approval At ICR Regional Vice President: Obiageli Katryn Ezekwesili Hafez M. H. Ghanem Country Director: Yusupha B. Crookes Henry G. R. Kerali Senior Global Practice Director: Jamal Saghir Juergen Voegele Practice Manager: Karen Mcconnell Brooks Marianne Grosclaude Page 4 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Task Team Leader(s): Oliver Braedt Abimbola Adubi ICR Contributing Author: Jeanette M. Sutherland Page 5 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES A. CONTEXT AT APPRAISAL Context 1. Fragility, conflict, and violence (FCV) have affected Liberia after two devastating civil wars, one each in the 1990s and early 2000s, but the country was experiencing relative peace and stability at the time of appraisal and was going through its second democratic election, since the end of the conflict in 2003. The incumbent Administration was re-elected in November 2011 and took office in January 2012. 2. The country was also experiencing steady real growth rates, and real Gross Domestic Product (GDP) growth for 2011/2012 was forecasted to be about 7 percent. Reduced inflation – from 7 percent in 2008/2009 to 5 percent in 2013 – was expected. In 2010, the country reached the completion point of the Highly Indebted Poor Country (HIPC) initiative, which provided the basis for substantial debt relief and new public and private investments in all sectors. Despite these positive indicators, the transition from post-conflict recovery to long-term development was plagued with challenges and high levels of poverty persisted. 3. As is the case today, the agriculture sector played a significant role in the country’s economy. The sector accounted for more than 60 percent of GDP and provided employment to about two-thirds of the 4 million population. Rice and cassava were the main food crops, while rubber, oil palm, and cocoa comprised the majority share of the country’s agricultural exports. Despite favorable agro-climatic conditions and available labor force, however, the tree crops sub-sector was performing sub-optimally. With no replanting and maintenance activities for 25 years due to wars, tree crop plantations were very old and productivity was low. International demand for tree crop products was high, but with limited production, processing factories were operating below capacity. In the case of palm oil, the country had turned from an exporter to a net importer. The tree crops sub-sector was facing additional supply-side challenges. These included relatively small-scale farms – most of which were small (of less than 2 hectares) and medium-sized (of less than 40 hectares); remote location of farms, compounded by poor condition of rural roads and limited market linkages (with both financial institutions and buyers); few formal, market-oriented farmer organizations or cooperatives; and limited capacities in the public sector, namely in the Ministry of Agriculture (MoA). 4. Within this context, the agriculture sector was one of the key sectors targeted in the country’s vision and strategy documents, including: Liberia Rising 20301, Liberia’s long-term vision for socio-economic transformation and development and Agenda for Transformation (AfT)2, and the five-year development strategy towards fulfillment of the vision. Through the AfT, the Government committed to prioritize improving the results of smallholder agriculture by aiming to increase agricultural productivity, access to markets and availability of inputs, as well as reducing constraints related to infrastructure, finance, skills, regulations, and property rights. The Liberia Agriculture Sector Investment Program (LASIP)3 included among its focus, the rehabilitation and expansion of rural roads, agriculture infrastructure and technology development, rural finance services, market and enterprise 1 Liberia Rising 2030, January 2013 2 Agenda for Transformation: Steps Toward Liberia Rising 2030, April 2013 3 The 2010 Liberia Agriculture Sector Investment Program (LASIP) was a five-year plan and was developed to meet country commitments under the Comprehensive Africa Agriculture Development Program (CAADP) Page 6 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) development and institutional strengthening to improve public service delivery to the agriculture sector. The project was developed based on priorities outlined in the Bank’s Country Assistance Strategy 4 (CAS) and following the recommendations of the Diagnostic Trade Integration Study (DTIS) for Liberia, which identified tree crops as offering “the best opportunity for strong and shared growth and poverty alleviation both in the short and longer term”5. Theory of Change (Results Chain) 5. The original design of the Liberia Smallholder Tree Crop Revitalization Support Project (STCRSP), envisioned as a pilot project, did not include a formal Theory of Change (ToC). A retrospective ToC has been reconstructed (Figure 1), summarizing the key challenges and defining ex post the causal pathway from project activities to project development objectives – as described in the PAD. The project aimed at increasing access to finance, inputs, technologies, and markets for smallholder tree crop farmers in Liberia and strengthening institutions to improve public service delivery to the sector. The project’s interventions in the tree crops sector were aimed at alleviating binding constraints to the development of the sub-sector – within a very challenging FCV context and to pave the way for the development of a large-scale program for increasing productivity in the sub-sector. Accordingly, the project activities focused on: (i) providing technical training to farmers and Farmers’ Organizations (FOs) through Farmer Field Schools (FFS); providing start-up grants and facilitating long-term credit for cooperatives to purchase inputs; providing equity matching grants; facilitating market linkages with concessionaires and other buyers; constructing post-harvest facilities (e.g. storage, bulking, primary processing); and rehabilitating feeder and farm access roads (Component 1); and (ii) supporting applied tree crop agricultural research (complementary to other initiatives, particularly the West Africa Agricultural Productivity Program – WAAPP, P094084); formulating tree-crop master plans for cocoa, coffee, oil palm, rubber; formulating a national policy and strategy for the development of FOs and cooperatives; building capacities of the technical staff of the MoA and of the Cooperative Development Agency (CDA); training on land use rights validation and farm mapping methodologies (Component 2). 6. The project was restructured but the PDO remained unchanged. However, while the ToC was essentially preserved not all key assumptions were validated when tested during implementation and the results chain was affected. The implicit critical assumptions of the reconstructed ToC were as follows : (i) counterpart co-financing, particularly from farmers and concessionaires and for which project activities were budgeted, would materialize; (iii) farmers’ priority investment needs were for seeds, fertilizers, and other agro-chemicals; (iv) the long-term credit scheme would be available for farmers, at least until the new tree crops matured; (v) access to quality inputs would be readily available in Liberia; and (vi) concessionaires and other off-takers would be willing to participate in out-grower and market arrangements. 4 Joint Country Assistance Strategy for the Republic of Liberia for the Period FY09-FY11; March 31, 2009; Report No. 47928-LR. 5 Liberia Diagnostic Trade Integration Study 2008, as quoted in the STCRSP PAD, p.4 Page 7 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Figure 1: Theory of Change Page 8 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Project Development Objectives (PDOs) 7. According to the credit agreement, the project development objectives (PDO) were “to increase access to finance, inputs, technologies and markets for smallholder tree crop farmers in Liberia, and to develop a long- term development program for the tree crops sector.” Similar wording was reflected in the project appraisal document (PAD). Key Expected Outcomes and Outcome Indicators 8. The key outcome indicators (KOI) initially designed to measure the project performance, as outlined in Annex 1: Results Framework and Monitoring of the PAD, were as follows:  KOI One: Area of smallholder tree crop farms rehabilitated, replanted or planted under the project (ha), disaggregated by (i) rehabilitation; (ii) replanting; and (iii) new planting;  KOI Two: Incremental yearly net cash flow in project areas (US$/ha). Disaggregated by: (i) high input rehabilitated cocoa; (ii) medium input rehabilitated cocoa; (iii) rehabilitated coffee; (iv) rehabilitated oil palm (farmers’ run plantation); and (v) rehabilitated oil palm (out-grower scheme);  KOI Three: Long term credit delivered to oil palm and rubber out growers under the project (US$’000);  KOI Four: Farm access roads rehabilitated (km);  KOI Five: Direct project beneficiaries, of which are female (number/%);  KOI Six: Long-term, large-scale tree crop development program formulated and approved by the Ministry of Agriculture (Yes/No). Components 9. The project consisted of three components: (i) Component 1: Smallholder Tree Crops Revitalization (US$16.9 million, including an IDA contribution of US$9.5 million6). This component aimed at revitalizing the production and marketing of major tree crops (cocoa, coffee, oil palm, and rubber) in selected counties. The planned approach was to assist smallholders to resume production through rehabilitating, replanting, or new planting of tree crops; building capacity of smallholders and their organizations; and strengthening value chains. Two models – high and medium input7- were to be piloted on rehabilitated cocoa farms, and a low input model tested on coffee and oil palm rehabilitations. Three sub-components were designed to deliver on these objectives. 6 The total expected co-financing (US$8.1M, most of which was expected to be in-kind) did not materialize. The project did not collect data about in-kind contributions to the project. 7 The medium input rehabilitation model consisted of under brushing, reducing shade, weeding and pruning cocoa trees and applying fungicides to fight black pod disease, and application of insecticides. The high input rehabilitation model consisted of the same operations combined with the application of specialized cocoa fertilizers. It was expected that productivity would improve both in the short term (with rehabilitation) and the medium/long-term (with planting and replanting) for farmers to quickly resume production (STCRSP PAD, p.22). Page 9 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) a. Sub-Component 1.1: Cocoa/Coffee Revitalization (US$12.2 million, including an IDA contribution of US$6.5 million). The planned approach under this sub-component was to provide support at each link of the cocoa and coffee value chains to facilitate certification of farmer organizations and their access to preferential markets; b. Sub-Component 1.2: Smallholder Oil Palm Revitalization (US$1.4 million, including an IDA contribution of US$1 million). The planned activities under this sub-component included support for rehabilitating and replanting oil palm under two different models: (i) re-launching production and modern processing in one existing smallholder plantation, and establishing small-scale processing units to be run by cooperatives; and (ii) establishing an out-grower model linked to a concessionaire; providing small grants to purchase inputs and small tools; providing long-term credit; and developing farm access roads; c. Sub-Component 1.3: Smallholder Rubber Revitalization (US$3.2 million, including an IDA contribution of US$2 million). The planned activities under this sub-component included support for replanting on existing smallholder rubber farms and new planting on smallholder farms under out-grower schemes linked, in three counties, to concessionaires and, in one county, to a large Liberian farm. Additional planned activities were organizing farmer organizations; rehabilitating feeder roads; providing grants and long-term credit; validating land use rights and boundary demarcation; and land surveying. (ii) Component 2: Institution Building and Preparation of Future Large-Scale Tree Crop Development Program (US $3.3 million, including an IDA contribution of US$3.1 million). This component comprised three sub-components: (i) Institution Building (US$1.5 million), aimed at strengthening the main public and private institutions involved in project planning, coordination, and implementation; (ii) Preparing a Future Large-Scale Smallholder Tree Crop Development Program (US$0.5 million), aimed at preparing and validating master plans for the targeted tree crops and a proposal for a follow-up large scale operation; and (iii) Project Preparation Advance (US$1.2 million). (iii) Component 3: Project Coordination and Management (US$ 3.0 million, including an IDA contribution of US$2.5 million). This component comprised two sub-components: (i) Strategic Planning, Coordination, Management and Implementation Support (US$2.3 million) and (ii) M&E and Knowledge Sharing (US$0.7 million). B. SIGNIFICANT CHANGES DURING IMPLEMENTATION Revised PDOs and Outcome Targets 10. The project had three level-2 restructurings in June 2014, April 2016, and October 2018. PDO indicators and outcome targets were revised during the second restructuring, based on the mid-term review (MTR). Revised PDO Indicators 11. Of the six PDO indicators, one was dropped and replaced with a new indicator, and two were moved to the intermediate level. Outcome targets were revised for three of the indicators. After revision, the PDO indicators were reduced to five. The changes are summarized in Annex A6.7. The KOI indicators resulting from the restructuring8 are summarized below. They include three original indicators (KOI One, KOI Four, and KOI Five); 8 Second Restructuring Paper, May 2016 Page 10 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) and two new indicators: KOI Two (formerly an output indicator), and KOI Three (newly introduced):  KOI One: Smallholder tree crop farmers with access to finance, input, markets, technologies;  KOI Two: Area of smallholder tree crops farms rehabilitated, planted, or replanted under the project (6,056 ha), disaggregated by rehabilitation (4,506 ha), replanting (1,250 ha), new planting (300 ha);  KOI Three: Yield of major rehabilitated tree crops, disaggregated by high input cocoa (1.30 mt/ha), medium input cocoa (0.50 mt/ha), coffee (0.50 mt/ha), oil palm (5 mt/ha);  KOI Four: Long term, large scale tree crop development program formulated and approved by the Ministry of Agriculture;  KOI Five: Direct project beneficiaries (3,100, of which 15 percent are female). Revised Components 12. There were no changes in the description of the components. Other Changes 13. Revision to intermediate indicators: Four intermediate results indicators of Component 1 (yield of high input rehabilitated cocoa, yield of medium input rehabilitated cocoa, yield of rehabilitated coffee, yield of rehabilitated oil palm) were merged into one indicator – yield of major tree crops, disaggregated by each tree crop –and moved to the PDO level, as shown in Error! Reference source not found.. The intermediate result indicator for Component 2, Coaching of MoA and CDA staff, was dropped. 14. Revision to long-term credit scheme (LTCS): Based on a request from the Government of Liberia for an alternative arrangement, the following changes were made to the LTCS under the first project restructuring 9:  The management of the LTCS was delegated to the Ministry of Finance and Development Planning (MFDP) and its administration to another agent instead of the Central Bank of Liberia (CBL). The Liberia Bank for Development and Investment (LBDI) was selected as the agent bank.  Transfer of IDA funds for the on-lending of the long-term credit to the participating financial institutions (PFIs) was made through the MFDP instead of the CBL.  Flexibility was allowed in the choice of concessionaires, farmers’ organization and target counties for the implementation of the rubber and oil palm sub-components. 15. Revised project cost estimates and funding requirements: The total cost of the project was revised from US$23.1M to US$19.5M and the IDA allocation was revised from US$15M to US$13.7M. See Annex 6.1 for a summary of the cost revisions. 16. Extension of the project closing date: The project closing date was extended as part of the second project restructuring. The original closing date was December 31, 2016; the revised closing date was November 30, 2018. 9 Restructuring Paper on a Proposed Project Restructuring of Smallholder Tree Crop Revitalization Support Project; July 28, 2014; p.4 Page 11 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) 17. Reallocation of costs across disbursement categories: The reallocation consisted in a reduction in Expenditure Categories 3, 5, and 7 and an increase in Expenditure Categories 1, 2, 4, and 6, under the third project restructuring10 (See Annex 6.2). Rationale for Changes and Their Implication on the Original Theory of Change 18. First restructuring (July 2014): The original design for managing the LTCS required the Ministry of Finance and Development Planning (MFDP) to lend to a Wholesale Apex Bank/Institution, which would accredit one or PFIs to partner in the delivery of the project’s credit program11. The Central Bank of Liberia (CBL) was selected as the Apex Bank. Due to difficulties among the MoA, MFDP, and CBL in reaching an agreement to implement the long- term credit line – including the CBL’s refusal to have its account audited, and in signing a Subsidiary Loan Agreement and a Subsidiary Credit Facility Agreement – all preconditions for IDA disbursements under the rubber and oil palm sub-components to be disbursed – there was an eight-month delay in project implementation and a subsequent disbursement lag12. The LTCS design was revised to accommodate an alternative agent bank, working with PFIs, and flexibility in the choice of participating concessionaires and farmers . 19. Second restructuring (May 2016): The rationale specified in the mid-term review Aide Memoire and the Second Restructuring Paper for revising the PDO indicators and their targets were to: have specific, measurable, attributable, relevant, time-bound (SMART) indicators; better reflect the project intervention going forward; adjust to the actual funding available; and take account of the two-year project extension (see Error! Reference source not found. for more details). The intermediate indicators on yields were moved to PDO indicators because yield was considered as a medium-term outcome and therefore more appropriate at the PDO level to give it prominence. A new indicator on access to finance, input, technology, and markets, reflecting the PDO statement, was introduced. The intermediate indicator on Coaching of MoA and CDA staff was dropped, as it was felt that the coaching of the staff was no longer relevant given the high staff attrition rate13. The estimated total project costs and funding requirements were revised after accounting for the two-year extension of the closing date and the revision of project targets. The exchange rate variation also contributed to the need for project cost adjustments – US$1,194,106 was lost due to the depreciation in currency14. Finally, the delays experienced in achieving project effectiveness and those due to the Ebola outbreak, and the changes in the design of the long- term credit scheme were the reasons provided for extending the project closing date. 20. Third restructuring (October 2018): According to the restructuring paper, due to the increased cost of consultant services, the existing allocations in some categories were insufficient to finalize the remaining activities. A re- allocation across disbursement categories was necessary for the project to finalize the remaining activities. 21. Apart from the budget reductions, budget reallocations, and revisions to the design of the LTCS, the design of the components remained the same and, therefore, there were no major implications to the Theory of Change, even with the adjustments in the PDO indicators. 10 Restructuring Paper on a Proposed Project Restructuring of Smallholder Tree Crop Revitalization Support Project; October 10, 2018 11 Project Appraisal Document, STCRSP; May 9, 2012; p.86 12 Restructuring Paper on a Proposed Project Restructuring of Smallholder Tree Crop Revitalization Support Project; 2014; p.4 13 Aide Memoire, STCRSP Mid-Term Review Mission; November 2015; Annex 3, p.26 and Bank TTL. 14 MoA-PIU Page 12 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) II. OUTCOME A. RELEVANCE OF PDOs Overall Rating: High 22. The Project development objectives were reflective of the country’s vision and strategy, as indicated in the context at appraisal section. The sector strategy at the time included smallholder tree crop development, the rehabilitation and expansion of rural roads, and revitalizing agriculture research as priorities. The sector strategy at the time of project closing – LASIP II, had sustained productivity growth among its strategic objectives and cocoa, oil palm, and rubber among the list of priority crops. These three crops are also identified as a priority in the current Administration’s national development agenda – the Pro Poor Agenda for Prosperity and Development. 23. The PDOs were also aligned with both the CAS FY09-FY11 and the Country Partnership Strategy (CPS) FY13- FY1715, which was in effect for most of the project’s operation. The development objectives of the CPS included narrowing the infrastructure gap to increase access and connectivity and improving productivity in agriculture. Key targeted outcomes included the increased share of rural population with access to all season roads, the increased share of roads in good or fair conditions, the increased use of new technologies, and increased agricultural production. One of the CPS’ performance indicators was selected from the STCRSP PDO indicators: “area of smallholder tree crop farms rehabilitated, replanted, or planted under STCRSP increased from 0ha to rehabilitation of 3,700ha, replanting of 1,250ha, and new planting of 850ha.” 24. The objectives of increased access to finance, inputs, technologies, and markets also remain consistent with the current World Bank Country Partnership Framework (CPF) FY19-FY2416 and with the findings of the Systematic Country Diagnostic (SCD)17. Both documents have identified these constraints as underlying factors of the overall low productivity plaguing the agriculture sector. The CPF noted that smallholder farming households, which comprise most of the agricultural sector, are not competitive due to – among other factors – “limited use of improved technologies, poor transport infrastructure, and the extremely limited use of modern cultivation methods.” The CPF further noted that there is a need to link numerous smallholder farms that rely largely on traditional inputs and farming methods to produce subsistence crops for household consumption with large commercial concessions that employ improved inputs and modern farming techniques to produce cash crops for export, including cocoa and coffee. At the same time, the SCD found that access to seed, fertilizer, machinery, transport infrastructure, and markets in Liberia is lower than in comparable countries. 25. Thus, the Project Development Objectives were, and remain relevant to Liberia, and consistent with the Government’s development objectives and current Bank Country Partnership Framework. The relevance of the project development objectives is therefore rated as high. 15 Country Partnership Strategy for the Republic of Liberia for the Period FY2013-2017; July 1, 2013; Report No. 74618-LR 16 Country Partnership Framework for the Republic of Liberia for the Period FY19-FY24; October 26, 2018; Report No. 130753-LR 17 From Growth to Development: Priorities for Sustainably Reducing Poverty and Achieving Middle-Income Status by 2030, Systematic Country Diagnostic; May 21, 2018; Report No. 113720-LR Page 13 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) B. ACHIEVEMENT OF PDOs (EFFICACY) Overall Rating: Pre-restructuring-Modest; post-restructuring-Substantial 26. While the PDO was not changed after the second restructuring, there were revisions in the scope of the key associated outcome targets – the targeted area of trees rehabilitated, newly planted, replanted and the number of beneficiaries were reduced (see Annex 6.3) – and therefore a split rating is applied to the pre- restructuring and post-restructuring results. Assessment of Achievement of Each Objective/Outcome 27. The PDO had twin objectives: to provide access to finance, input, markets, and technologies; and to develop a long-term, large-scale program to support the tree crop sector. The key associated indicator for the first objective is “a compound statement with multiple outcomes [that] needs to be “unpacked” so that each promised outcome can be assessed separately”18. Accordingly, the outcome indicator “number of beneficiaries with access to finance; inputs; technologies; markets” is unpacked for the purpose of assessing the achieved project outcomes as described below. Table 1: Reported PDO Outcome Indicators Indicator Original Target Formally Achieved Pre- Achieved Post- Revised restructuring restructuring phase (%)* phase (%) 1. Smallholder tree crops farmers with 3,025 No change 3,750 123** 123 access to finance, input, markets, technologies (No.) 2. Area of smallholder tree crop farms 11,300 6,056 5,828.57 52 96 rehabilitated, replanted, or planted under the project (ha), disaggregated by: (i) 7,100 (i) 4,506 (i) 4,194.57 (i) 59 (i) 93 (i) Rehabilitation (ii) 2,200 (ii) 1,250 (ii) 1,004 (ii) 46 (ii) 80 (ii) Replanting (iii) 200 (iii) 300 (iii) 630 (iii) 315 (iii) 210 (iii) New Planting 3. Yield of major rehabilitated tree crops No change 77*** 77*** (mt/ha), disaggregated by (i) High input cocoa (i) 1.30 (i) 1.00 (i) 77 (i) 77 (ii) Medium input cocoa (ii) 0.50 (ii) 0.46 (ii) 92 (ii) 92 (iii) Coffee (iii) 0.50 (iii) 0.30 (iii) 60 (iii) 60 (iv) Oil palm (iv) 5.00 (iv) 1.50 (iv) 30 (iv) 30 4. Long term, large scale tree crop Yes No change Partially Partially Partially development program formulated and approved by the Ministry of Agriculture Direct project beneficiaries (of which % 5000 (20) 3100 (15) 3850 (900) 77 (90) 124 (153) female) Source: STCRSP ISR, No.13, Government’s project completion report Note: * Share of end of project achievement to original target; ** Indicator introduced at post restructuring; ***Weighted average calculated by the ICR 18 ICR Guidelines, September 2018 Page 14 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) 28. Due to delays in project effectiveness and the outbreak of the Ebola virus that covered 2014 and 2015, there were insufficient time to register results during pre-structuring to enable assessment on the outcome indicators. Nevertheless, some activities were undertaken, namely the identification of farmers, implementation of some trainings for cooperatives, and validation of farmers’ land. Even though the yield target achievement was substantial, given that the area and beneficiary base on which these were achieved (77% of the beneficiary base and 52% of the expected area) was significantly higher in the pre-restructuring phase, Efficacy for the pre-restructuring period is rated as modest. The following section assesses achievement post-restructuring. 29. PDO Outcome 1. Number of farmers with access to finance, inputs, technologies, and markets: Overall, this outcome was partially achieved. While access to inputs was fully achieved, there were partial achievements, to varying degrees, on access to finance and technologies, but a lack of achievement on access to markets. 30. PDO Outcome 1.1. Access to finance increased: While the reported data suggests that the project achieved this PDO outcome – ICR findings show that access to finance was limited to the one-off assistance provided by the project, whether in kind or in cash, while the provision of long-term credit did not materialize (with one exception noted below). The project aimed to increase smallholders’ access to finance through: (i) the provision of equity matching grants of at least US$128,00019 to a total of 40 cocoa and coffee farmer organizations to increase their working capital to allow FOs to pay at least 70 percent of the farm gate price to their members20 with the balance to be paid after the crops were sold to exporters and processors; (ii) the establishment of long-term credit for three out-grower schemes; and (iii) facilitation of smallholders and FOs’ access to adapted financial services delivered by participating financial institutions and rural microfinance institutions (MFIs). The project proposed to finance the establishment of (a) rural MFIs in the districts of intervention; (b) approximately 12 banking windows at section/clan level to be opened by participating commercial bank; and building the capacity of (c) PFIs’ executive and lending staff in key areas such as agricultural loan appraisal, cash-flow based lending, value chain financing and commodity collateralization and the use of innovative delivery methods such as mobile phone and affiliated banking windows. 31. At project end however, matching grants for a total amount of US$62,390 – less than 50 percent of the minimum intended, were provided to 12 cocoa cooperatives. The 12 cooperatives together include 119 farmer organizations21 in the three targeted cocoa and coffee producing counties22. So, fewer funds were distributed to more FOs. The project database lists 2,974 farmers as having access to matching grants 23, although there is no systematic data on the management and performance of the grants. 32. Under the long-term credit scheme, one rubber cooperative in Montserrado county, comprising 150 farmers, received access to US$111,526, in the form of an in-kind loan24. None of the beneficiary farmers had access to 19 The grant amount was estimated at US$3,200 per new FO and US$3,800 per existing FO 20 The grants were also to allow cooperatives to buy from other farmers, Project Completion Report 2013-2018, PIU; pps. 8, 19 21 FOs comprised anywhere between 100 to over 500 farmers 22 Bong, Grand Gedeh, and Nimba. 23 See Annex, A6.5 24 The PFI disbursed payment to a supplier for inputs provided to the farmers. The farmers are expected to repay the PFI the costs of the inputs Page 15 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) long-term credit or finance prior to the intervention (as indicated by the baseline data), and the project acknowledged that farmers would need financing from commercial banks at the end of the project to maintain the tree crops established under the project. 33. Access to finance is viewed as a combination of matching grants and the in-kind credit provided to farmers, based on the reported data and the project database, and many, if not most, farmers had access to one or the other of these instruments or both25. In addition, the amounts of the matching grant in local currency available to farmers were substantially increased because of depreciation of the Liberian dollar vis-à-vis the US dollar 26. The database shows a total of 3,518 project-supported farmers (not 3,750, but more than the target of 3,025), with 55 percent of the targeted 3,025 farmers with some access to credit, and 105 percent of targeted beneficiaries with access to some form of matching grants. The project database reports that 1,652 cocoa farmers from Nimba county received credit, when project documents27 and the participating financial institution confirmed that only one organization (the rubber cooperative already mentioned), benefited from the long- term credit scheme. The number of farmers formally reported as receiving credit and/or matching grants may be slightly lower. Many farmers – at least from two out-grower schemes – are still waiting for the project to fulfill its promise on access to long-term credit. However, there were no financing arrangements in place – neither through MFIs, MFDP, the CBL, nor through the commercial banks as envisioned by the project – to ensure farmers’ access to credit for maintaining their farms with tree crops supported by the project after it closed. 34. PDO Outcome 1.2. Access to inputs increased: The achievement on this outcome was significant to farmers. The project aimed to provide planting materials, seedlings, fertilizers and agro-chemicals to smallholder farmers through: (i) small grants paid to third parties to provide inputs, and high yielding, hybrid cocoa seedlings to cocoa/coffee farmers 28; (ii) seed gardens and village nurseries, (iii) out-grower-concessionaire schemes in which concessionaires would provide these inputs to oil palm and rubber farmers on a credit basis; (iv) the long-term credit scheme in which the credit was paid directly by the participating financial institution to third parties for inputs ordered by the loan beneficiaries; and (v) facilitating linkages with input suppliers, who were expected to provide inputs on credit to farmers. 35. No results were reported on establishing seed gardens29 and relevant out-grower concessionaire linkage. One input supplier was contracted to provide some agro-chemicals to farmers and training to 67 farmer field school facilitators. Fifty-two and 26 nurseries were set up in 2016 and 2017 respectively 30. There were informal linkages arranged with WIENCO, a Dutch Cocoa company, which provided technical training and input on credit to cocoa farmers under the rehabilitation scheme. Project data on WIENCO’s input and related farmers’ benefit is also absent. According to the project database, all 3,518 farmers listed (116 percent of the target) received at least 25 Matching grants were provided to cocoa farmers; the long-term credit scheme targeted rubber and oil palm farmers 26 The Liberian dollar depreciated against the US dollar from an average of 1US to 80 LBD to 1 US to 150 LBD during project implementation. 27 Project Completion Report, p.25; Project Impact Evaluation Report, p.14 28 The grant was also used to purchase small tools, protective gear – such as gloves, masks, rain boots, and the like 29 Seed gardens were intended to produce seeds for farmers to access. Seeds purchased from the seed gardens were to be used to establish nurseries to generate seedlings for planting. 30 PIU semi-annual report, July 1, 2015-December 31, 2015 Page 16 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) one of the following: small tools, fertilizers, and chemicals. These results suggest that the project was successful in providing a mix of inputs to farmers. 36. PDO Outcome 1.3: Access to technologies increased. Farmer field schools, training of cooperatives, and the establishment of processing mills, beneficiary farmers have been exposed to improved ideas and/or mechanism for cultivating tree crops and processing their harvests. The project aimed to provide smallholders with increased access to technologies via the following activities: (i) training in good agricultural practices and techniques through farmer field schools; (ii) supporting applied tree crop agricultural research, through the institutional strengthening component, including the establishment and maintenance of cocoa and coffee seed gardens, refurbishing laboratories for soil testing, biological analyses, support to adaptive research on tree crops; and (iii) piloting cocoa and coffee processing through the establishment of a small-scale cocoa and coffee processing unit. 37. Farmers received training on good agricultural practices through farmer field schools or technical advice provided by concessionaires and input suppliers. The project constructed two palm oil processing stations at the farmer managed plantation in Grand Gedeh and installed each with two motorized mills with the processing capacity of at least 40 drums of fresh fruit bunches within 45 minutes in total for the four mills. The project reported that the increased oil production at the factory enhanced the plantation’s income generation capacity although quantitative data on the pre and post processing activity incomes was difficult to obtain. Sixteen cocoa cooperatives were reported to have received moisture testing meters31. No results were reported on the applied tree crops agricultural research activities. Apart from the data provided by the implementing partner on the number of cocoa cooperatives trained (see Annex 6.8), data on the total number of farmers that benefited from many of the trainings and use of improved techniques provided is unavailable. 38. PDO Outcome 1.4: Access to markets increased. From the evidence available, the project did not achieve its objective of providing beneficiaries with satisfactory access to markets in terms of physical access and formal market arrangements. In the absence of reliable data, however, the extent to which the project has come close or short of its target cannot be determined. The project aimed to increase smallholders’ access to markets though the following activities: (i) the rehabilitation of 30 km of market access feeder roads in critical points to ensure easy marketing of cocoa and coffee and/or distribution of inputs to strengthened farmer organizations; (ii) the facilitation of formal market arrangements between smallholder farmers and concessionaires, exporters, and other off-takers, and (iii) the facilitation of linkages with financial and input markets. 39. At the end of the project, just five kilometers of roads (about 17 percent of the target) were rehabilitated 32. Anecdotal evidence, gathered from the ICR mission, suggests that some 48 beneficiary farmers in 9 communities serviced by this stretch of road were able to take their produce to market, including in the rainy season 33. There were no results recorded on formal market arrangements to guarantee the beneficiary farmers of a market once their tree crops are ready for harvesting34. Further, out-grower schemes were not established, or at least never 31 Government’s project completion report 32 Due to the reduced cost allocation to this activity, Mid-Term Review Aide Memoire, p.6 33 Project Completion Report 2013-2018, MoA & Interview with the PIU 34 The ICR mission found that under the cocoa rehabilitation scheme, some farmers were able to sell their crops to WEINCO, an implementing partner. Others reported that the yield of their rehabilitated of cocoa tree crops did increase, but they were forced to sell to middlemen at low price in the absence of guaranteed markets. Representatives from one rubber cooperative reported that in the absence of a formal market arrangement with the large estate selected by the project, they are pursuing marketing agreements on their Page 17 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) formalized with concessionaires or the large estate35, so these farmers do not have a guaranteed market for their produce. Other cooperatives, including those that received technical support from the concessionaire, EPO, expressed concern about market prospects for the crops once their trees are ready for harvesting36. Although limited in scale, the matching grant allowed some farmers to sell some of their crops directly to their cooperatives. 40. PDO Outcome 2: Area of smallholder tree crop farms rehabilitated, replanted or planted under the project increased. In the aggregate, the project substantially achieved its target (96 percent) for this key performance indicator. Also, the aggregated targets hide variations in achievement across different crops and models 37. There were some inconsistent figures in project documents regarding the disaggregated targets. The rehabilitation was understandably the focus of the project, which was testing models in a difficult FCV context. Even if slightly below target, the project’s achievement in rehabilitation was significant, considering that the FCV challenges were compounded by reduced project financing and exacerbated by the unprecedented Ebola outbreak in the region, which delayed farm improvement activities for about a year and a half38. The project database does not provide data on replanted or newly planted tree crops. The establishment and rehabilitation of tree crops were helped by the access to inputs and finance noted above. 41. PDO Outcome 3: Yield of major tree crops on (rehabilitated) high input cocoa, (rehabilitated) medium input cocoa, rehabilitated coffee, rehabilitated oil palm increased. The ICR calculated an aggregate yield achievement that is substantial. The ICR calculated the overall achievement of the yield indicator using the Paasche Index (PI) formula39. The weighted averages calculated by the ICR are 0.63 mt/ha for the target yield and 0.48 mt/ha for the yield achieved for a project overall achievement of 77 percent40. Whereas the establishment and rehabilitation of tree crop plantations helped by some access to finance and credit, is a one-time event, the attainment of increased yields requires continuous effort and care. As reported earlier, most farmers did not receive that continuous support from the project in terms of credit (to hire labor) and access to market (to provide revenues to maintain their farms). This likely explains the lower than expected results in the yield indicator. Nevertheless, in light of extenuating challenges the project faced, including a reduced budget, which compromised key project deliverables (access to credit and markets) impacting this indicator, a 77 percent increase in yields, particularly for trees that were abandoned for decades during the wars, is significant to the farmers. own with some exporters. However, they expressed concern that exporters are not following regulations that mandate them to purchase directly from cooperatives, and instead purchase from farmers offering lower prices – thus encouraging side-selling by farmers. 35 While technical assistance was provided to farmers by these enterprises, they did not perceive themselves to be in an out-grower relationship. 36 Meeting with cooperatives during ICR mission, February 2019 37 Project Impact Evaluation, p.9. Also see Annex A6.6 on data inconsistencies 38 While the official Ebola period lasted 15 months, it took a couple months for project activities to resume to normal once the country was declared Ebola free. 39 The PI was developed to assess the percent change in price over two periods, but the formula is just as appropriate to measure the percent change in the yield achieved over the target. To this objective, one substitutes prices with yields, and substitutes end-of-period goods quantities purchased with end-of-project crops areas achieved. Based on the PI formula, one can also derive the weighted average yields, similarly using end-of-project achieved areas to weigh both target and achieved yields (see Annex A6.4 for details). Here the Paasche Index is: (Sum (achieved area x achieved yield) / Sum (achieved area x target yield)) x 100. 40 An IFAD supported project of the same name, Smallholder Tree Crop Revitalization Support Project, which was effective from 2012- 2017, focused on rehabilitation of cocoa and coffee farms. Rehabilitation activities included under brushing, pruning, de-shading, and sanitation (removal of diseased plants, black pods etc), similar to the medium input model used by the Bank supported STCRSP. At close, the project reported to have rehabilitated 10,314 ha of cocoa farmland with an 82% achievement of yields – from a 2013 baseline of 0.08 mt/ha to 0.38mt/ha in 2017. Page 18 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) 42. PDO Outcome 4: Long-term, large scale tree crop development program formulated and approved by the Ministry of Agriculture. The project did not develop the long-term, large-scale program (STCRSP II) as originally intended, but given the comprehensiveness of the master plan prepared and its approval by the MoA, there is a chance that the second phase of the STCRSP will eventually be developed. Based on the project documents, it seems that the long-term, large scale program envisioned by the project was a STCRSP II 41. 43. While there was no assessment of the different models tested under the project, the master plans and related action plans are quite comprehensive, as they (i) identified long-term strategic tree crop sector goals for 2020- 2040; prepared a strategic road map for cocoa, oil palm, and rubber (components, expected production targets, activities, and expected outcomes); and (ii) outlined recommendations for policies and spending to support a holistic tree crop program for raising Liberia’s performance to at least the level of other countries in the sub- region42. The GoL used the master plan to reiterate its request – formally submitted to the Bank in early 2018, for a second phase of the project. 44. PDO Outcome 5: Number of project direct beneficiaries of which 15 percent are female . The project reported 3,850 direct project beneficiaries of which women accounted for 900 (23 percent), an achievement of 124 percent and 153 percent of the 3,100 and 15 percent targets respectively. The project database, however, reports 3,518 total farmers of which 698 (20 percent) are women. While lower than reported, the figures still represent 113 percent and 133 percent of the project targets for total beneficiaries and females. The female share was: 20 percent in access to inputs, and 19 percent in access to matching grants (see Annex A6.5). 45. Intermediate indicators: Based on project completion reports, 6 of the 13 intermediary targets were fully achieved; 5 were partially achieved; and 2 were not achieved (see Annex 1). Three of the intermediary activities that were most crucial to the PDO objectives – establishing out-grower schemes, rehabilitating roads, and facilitating out-grower schemes with access to long-term credit, were partially achieved. The partial achievement of some is questionable, such as 16 of 20 out-grower scheme established. The ICR mission was unable to identify any functioning out-grower scheme. The one out-grower scheme that had access to long term credit – the Todee Cooperative, did not have an out-grower relationship with the anchor company selected by the project – MARCO. The volume of cocoa bulked by project supported FOs – 850MT against a target of 1,00MT, was one of the stronger performing activities. Justification of Overall Efficacy Rating 46. On balance, for the post-restructuring phase, given that the area, beneficiary and yield targets were significantly met, that inputs and finance were provided to the beneficiaries, that access to technologies improved, and that the project-developed master plans were comprehensive and used by the government to request support for a second phase of the project, the Efficacy is rated as Substantial. 41 The Aide Memoire for the Mid-Term Review outlined as one of the revised timelines, “Starting PY4, the preparation of a large-scale smallholder tree crop development program (STCRSP II) subject to a two-year extension of the closing date, the preparation of the large scale program is tentatively programmed for 2017;” p.13 The follow up restructuring paper noted that, “the bulk of funding for preparing the large-scale project should come from a project preparation advance under STCRSP 2,” Restructuring Paper, 2016, p.8 42 Tree Crops Master Plan: Road Map to Liberia’s Prosperity; SOCODEVI; 28 November 2019 Page 19 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) C. EFFICIENCY Overall Rating: Substantial Assessment of Efficiency and Rating 47. Economic and Financial Analysis: To assess the efficiency of the project interventions economic and financial analyses were conducted at project completion following a with-and-without simulation of the project net benefit stream, like the methodology used at appraisal. The re-estimated internal rate of return (IRR) and economic rate of return (ERR) results are slightly lower than the appraisal-stage estimates. The stream of economic benefits from project-funded activities over twenty-five years generate a net present value (NPV) of US$26.7 million, an economic internal rate of return (EIRR) of 29.2 percent, and the benefit cost ratio (BCR) of 2.99 (at a 10 percent discount rate). The financial indicators are equally good, as shown in 48. . The analysis for the farm models and from the individual farmer’s perspective has also demonstrated similar positive results (see Annex 4). 49. Sensitivity Analysis: The sensitivity analysis shows that the project returns are highly robust and indicates a solid resilience to increases in costs and reduction in benefits. The sensitivity analysis demonstrates that under the scenarios of 20 percent or more cost increase (including counting the entire project cost by adding components 2 and 3), the corresponding EIRRs remain high with substantial positive NPVs. Similar figures are also observed in the scenarios of decrease in benefits. Indeed, it will take switching values for a rise in costs of over 50 percent and a simultaneous reduction in benefits of over 55 percent for the entire project not to be economically viable (see details in Annex 4). Table 2: Present value of Economic and Financial Flows Economic and Financial Present Value of Flows Performance coefficients Financial analysis Economic Analysis ICR estimates Appraisal ICR estimates Benefit cost ratio 2.8 - 2.99 Rates of Return 27.1% 33% 29.2% Net Present Value US$24.7 million US$28 million US$26.7 million 50. Administrative efficiency: It took about 11.3 months, after approval on 5 June 2012, before the project became effective, due to the time it took for parliamentary ratification and in meeting other effectiveness conditions. These initial delays were exacerbated by the Ebola virus outbreak, which created a huge humanitarian catastrophe in the country. 51. Efficiency rating. Although the project experienced operational and implementation challenges associated with the fragile context in which it operated, the ex-post economic and financial analysis shows the project to be Page 20 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) economically robust and to generate substantial returns to the investment both at the project and individual farmer levels. Efficiency is therefore rated Substantial. D. JUSTIFICATION OF OVERALL OUTCOME RATING Overall Rating: Moderately Satisfactory 52. While the project significantly achieved its targeted indicators resulting in benefits to farmers, overall the project had some shortcomings. For some of the objectives reported as having been achieved, the level of impact on the farmers is inconclusive, specifically the objective related to access to finance and technologies. Due consideration needs to be given to the FCV context and factors outside the control of the project, namely the Ebola outbreak that paralyzed project activities for almost two years and the compressed timeframe for achieving the bulk of the project deliverables. Given the project’s PDO is Highly Relevant, Efficiency is Substantial, and given the Modest and Substantial Efficacy ratings in the pre- and post-restructuring periods respectively, the overall outcome pre- restructuring rating is Moderately Unsatisfactory and the overall outcome post-restructuring is rated as Satisfactory. Weighting the pre- and post- restructuring performance, the final outcome rating is therefore Moderately Satisfactory. (see Table 3 for details on the split rating). Table 3: Application of Split Rating Pre-Restructuring Post-Restructuring Relevance of PDO High Efficacy (PDO) Modest Substantial Efficiency Substantial 1 Outcome Ratings Moderately Unsatisfactory Satisfactory 2 Numerical Value of the Outcome Ratings* 3 5 3 Disbursement 7.5 6.2 4 Share of Disbursement 0.55 (or 55%) 0.45 (or 45%) 5 Weighted Value of the Outcome Rating (Row 2 x Row 4) 1.65 2.25 6 Final Outcome Rating Moderately Satisfactory 1.65+2.25=3.9 rounding it to 4.0 *Note: Highly Unsatisfactory (1); Unsatisfactory (2); Moderately Unsatisfactory (3); Moderately Satisfactory (4); Satisfactory (5); Highly Satisfactory (6) Page 21 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Gender 53. The project was not gender tagged. Although no explicit or systematic focus was given to gender mainstreaming in the design or implementation of the project, it targeted a certain share of women in the total number of project beneficiaries (which it largely exceeded). The project database provides information related to women’s characteristics (age, family size, farm size) and to their share in access to inputs, credit and matching grants 43. Institutional Strengthening 54. Improving the capacities of key institutions was a strategic approach of the project to help achieve its development objectives. Within this context, the project sought to strengthen farmers’ cooperatives to provide improved services to members – particularly in marketing; and relevant government institutions to improve service delivery to the tree crops sector – particularly policy and extension services support. 55. Planned training of cooperatives were incomplete due to reduced budget . The project reported that 16 cooperatives comprising 113 farmer organizations under the cocoa sub-component received training; 13 farmer- based organizations under the rubber sub-component were formed and received training; and 3 farmer-based organizations under the oil-palm sub-component were formed and received training. SOCODEVI, 44 VOSIEDA45, and the government funded Cooperative Development Agency (CDA) provided trainings. SOCODEVI reported that it provided trainings to 14 out of 16 cocoa FOs and cooperatives, accounting for 1,380 farmers, and was able to complete three out of eight modules planned: member-client relations, governance, and economic issues. CDA reported that it covered the counties not covered by SOCODEVI and provided certificates to FOs that were enabled to become cooperatives. Both institutions indicated that trainings were not completed because of the lack of project funds. This was corroborated by cooperative members, some of whom indicated that they received a one- time intensive training. Many cooperative members have requested additional training, including in the modules that were not completed. 56. Despite those shortcomings, there was widespread appreciation among the beneficiaries for the training received and they all requested to have more, particularly in business planning and marketing . All cooperatives trained have had democratic elections and created constitutions and by-laws to govern their activities, with an average increase in membership of 70 percent46. Additionally, it was reported that due to the trainings, five cooperatives are now trying to form a union of cooperatives and are requesting CDA’s assistance to obtain from the Liberia Agriculture Commodities Regulatory Agency (LACRA) a certificate to allow them to export cocoa. While some farmer-based organizations, having formed themselves into a rubber cooperative, used their pooled resources to develop a nursery and were able to supply farmers with rubber stumps to replace those supplied from the project that did not survive planting47. 43 Annex, A6.5 44 Société de coopération pour le développement international 45 A local NGO 46 Annual Progress Report, July 1, 2016-June 30, 2017, STCRSP, Ministry of Agriculture 47 CDA, as reported during ICR Mission Page 22 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) 57. While the project did not achieve its objective of developing a capacity building plan for the Ministry of Agriculture and related agencies, it provided financial support to the CDA for recruiting a legal expert to assist with their update of the draft Cooperative Societies Act. The CDA advised the project to provide this support instead of its proposed plan to develop a “national policy and strategy for FOs and cooperative development” 48. The 2010 Act has been updated and was submitted to Parliament in 2018 for ratification. Five out of the targeted 70 persons were trained. Support was given to financing the final year of PhD studies for five Liberians whose earlier years were financed by the WAAPP. 58. The project supported the creation of the Liberia Agriculture Commodities Regulatory Agency by assisting with feasibility studies and the elaboration of, among other documents, the LACRA Bill . The LACRA Act was signed by the President in 201649. The project also supported the Environmental Protection Agency to establish an Environmental and Social Management Team to help monitor the project. Assistance under the project was provided to the Ministry of Agriculture, its departments and county offices to increase their operational support to the project and their overall operational capacity. Mobilizing Private Sector Financing 59. The project aimed to mobilize private sector finance through the planned out-grower schemes with concessionaires and large estates and the long-term credit scheme. The concessionaires and the large estate that participated in the project provided technical assistance and management services to farmers, but the project was unable to get buy-in from these private sector stakeholders to invest in the schemes. The project was able to facilitate a long-term credit scheme for one cooperative. The PFI indicated interest in replicating the experience with future projects, if risk guarantees were addressed. Poverty Reduction and Shared Prosperity 60. The full impact of the project on poverty reduction and shared prosperity is expected to materialize after new planted and replanted tree crops are harvested, starting at least two years after the closing of the project. If the assumptions of the EFA holds, the project investments will have a positive impact on smallholder tree crop farmers’ livelihoods. For example, for every US$1 invested by smallholders in high input cocoa, a potential return of US$6.73 is expected. Over the life of the project, however, direct beneficiaries generated income from the processing mills installed on the farmer run plantations, from farmer sales to WEINCO under the cocoa rehabilitation intervention, and from cooperatives’ purchases of crops from its members. The income thus generated undoubtedly had some impact on poverty reduction and shared prosperity. The lack of quantitative data, however, makes it difficult to ascertain the extent of this impact. 48 Deputy Director, CDA, 2019 49 STCRSP Project Completion Report, MoA Page 23 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME A. KEY FACTORS DURING PREPARATION 61. The project design depended on some critical assumptions, which were not supported during project implementation, namely: (i) That the co-financing from project stakeholders would materialize. The project’s activities were budgeted at US$23.1 million, of which IDA would cover US$15 million. The co-financing (US$8.1 million) from the Government of Liberia, concessionaires, and farmers, which the project team explained was to be mostly in-kind, did not materialize. This limited the scope of the activities to be implemented. For example, credit disbursement to rubber and oil palm growers were expected to be in-kind (planting materials, fertilizers) provided by concessionaires50. The concessionaires did not commit to this. Similarly, the cost of labor was expected to be in- kind provided by farmers. As discussed below, farmers were in dire need of cash to hire labor. Additionally, data was not collected, or at least was not available to the ICR team, on the calculations of in-kind co-financing from project beneficiaries and other stakeholders. (ii) That the priority investment needs of the smallholder tree crop farmers for establishing and maintaining their farms were for planting materials, fertilizers, and chemicals51. The long-term credit was provided to farmers in the form of in-kind contribution. Assessments were undertaken to determine the amount of fertilizers and agro- chemicals that each farmer would need, based on the size of the farm. Orders were placed for the inputs and the participating financial institution delivered payment to the supplier. Grants were also allocated under the project to purchase similar inputs, and small tools, for farmers who did not have access to the long-term credit scheme. Although they acknowledged that access to planting material and fertilizers remained priorities, most of the farmers indicated hiring labor as a first-order priority, followed by mechanized services, and transport to get their products to market. The mid-term impact study also found that all beneficiaries indicated they had problems either mobilizing enough labor from the household or accessing credit to hire labor for their farming activities 52. Therefore, at least two other priority needs – mechanization53 and labor54 services – were not addressed. (iii) That concessionaires and large-scale farm owners would be willing and/or had the capacity to engage in a formal and/or long-term relationship with smallholders. The project targeted two concessionaires (EPO- LIBINCO and SRC) and one large estate (MARCO) to support out-grower schemes. The expectations were for these concessionaires and the large estate to sign management contracts to provide technical assistance to 50 STCRSP PAD; p.12 51 STCRSP PAD; p.86 52 Mid-Term Impact Study, Draft Report; egis; June 2016; p.12 53 The representative for the farmer-managed plantation indicated that due to lack of cash to hire mechanized services, they were only able to prepare 40 ha of the 300 ha land targeted. This group of farmers were not targeted for matching grants or long-term credit. 54 While the project expected farmers to use their own labor, farmers indicated that their labor needs far exceeded what was available within the family. For the cocoa and coffee crop models and oil palm planting/replanting, the project assumed no need for external labor; while it acknowledged the need for hired labor for rubber planting/replanting, it assumed that revenues from the sale of old rubber trees for charcoal would be used to hire labor; STCRSP PAD Page 24 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) farmers in out-grower schemes linked to their plantations. While these implementing partners were not expected to contribute to the project costs, the intent was for them to maintain the technical assistance after the four-year implementation period55. Additionally, the project anticipated that the concessionaires would participate in the long-term credit scheme56. The cooperative with which the out-grower-nucleus farm linkage was supposed to be established with MARCO reported that their farmers did not receive any technical assistance from the company. The Head of MARCO acknowledged that the company was, and still is, experiencing challenges that prohibited them from making commitments beyond what was outlined in their contract with the project. (iv) That the continuation of the long-term credit scheme would be possible at least until the tree crop farms established under the project reached maturity . Credit is required to finance the maintenance of the tree crops farms established. The expectation of the project was that a follow-up large-scale tree crops program would continue, facilitating long-term credit for the beneficiary farmers. In the absence of such a program, the project expected that the Ministry of Finance and Development Planning would commit to provide funding for lending to the PFI under a Subsidiary Financing Agreement between MFDP and the Central Bank of Liberia 57. Neither the large-scale program, nor the Subsidiary Financing Agreement materialized. Additionally, the long-term credit scheme was only made available to one cooperative. Some farmers have reported that, due to the lack of credit to hire labor, their farms have begun to deteriorate again. This has been confirmed by at least one concessionaire, who indicated that due to the lack of financing to supplement their income, some of the farmers have begun to plant food crops on their farms and practice slash and burn. Members from one cooperative indicated that if they do not obtain credit to maintain their farms, there is concern that the quality of their harvest will not meet the standard to ensure profitable market access. (v) That access to quality seeds, fertilizers, agro-chemicals would be readily available in Liberia, particularly in the counties targeted. The project anticipated that the initial grant to cover the costs of pesticides and fertilizers in the cocoa and coffee rehabilitation models would act as strong incentives to national and foreign-based input suppliers to intervene in the cocoa and coffee value chains. It was also expected that input suppliers would extend farmers and their organizations in-kind credit, drawing from their own resources and possibly from draft facilities or commercial credit lines from PFIs58. Sourcing inputs for farmers during the project turned out to be a challenge. Seeds had to be imported from Ivory Coast and Ghana and were very expensive. The distance to transport seeds from these countries to the targeted beneficiaries posed a risk to the seeds’ quality. The project recommended the Government of Liberia sign a bilateral agreement with Ghana 59 and Ivory Coast60 to facilitate access to seeds from these countries’ research institutes. The process was complex and resulted in delays. Eventually cocoa seeds 55 STCRSP PAD; pp.60-61 56 Ibid; p.90 57 Ibid 58 STCRSP PAD; p. 60 59 STCRSP ISR No. 9; 18 June 2016 60 STCRSP ISR No. 10; 23 November 2016. Page 25 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) were procured from Ghana61. Figure 2: Schema of Project Milestones and Key Factors Affecting Implementation Year Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec 2012 Approval 2013 Effective 2014 *EBOLA RS 2015 **EBOLA 2016 RS OCD 2017 Election/Mission Hiatus 2018 NG NG Transition. Board preparation of new project- STAR-P RS Close Source: Project documents, feedback from Bank Task Team Leader and Project Coordinator. Note: RS- Project restructuring; OCD- Original close date- Nov 30; NG-New government elected * Ebola Outbreak; **Declared Ebola free B. KEY FACTORS DURING IMPLEMENTATION The project took just over 11 months to become effective on 13 May 2013. Less than a year later, an Ebola pandemic broke out and the project activities ground to a halt for almost two years. During the outbreak, all projects in the country were halted and focus was on responding to the epidemic. Emergency funds were provided for increasing rice production to address critical food shortages. Even though the country was declared Ebola free in May 2015, it took a couple of months for project activities to resume. Other key factors included the presidential electoral activities, during which – towards the end of 2017 – missions to the country were suspended (see 62. Figure 2). 63. Apart from the above-mentioned factors, most of which were outside the control of the project team, a key issue that affected project implementation was the limited number and capacity of staff and high staff turnover at the project implementing unit (PIU). Weak capacities were further stretched with the decision, at the mid-term review, to use the PIU to manage the agriculture component of the Bank-financed Liberia Forestry Support Project 62. IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME A. QUALITY OF MONITORING AND EVALUATION (M&E) M&E Design 64. The M&E framework for the STCRSP was expected to serve as a tool for better management and decision-support and learning and accountability throughout the project period. The M&E framework consisted of (i) a baseline study, (ii) a web-based online management information system (MIS), (iii) regular progress monitoring studies, (iv) thematic studies, (v) a mid-term review, and (v) an impact evaluation. In particular, the web-based M&E system was developed to (a) ensure a high level of transparency and accountability; (b) ensure effective results- based management and to keep the focus on outcomes and impacts; (c) capture and communicate lessons learned to improve performance during project implementation, and allow other similar and related projects to benefit 61 STCRSP TTL, March 2019 62 Mid-Term Aide-Memoire; November 2015; STCRSP ISR No. 10; November 2016 Page 26 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) from improved practices; and (d) provide useful lessons to formulate the proposed future larger tree crop development program. 65. The original M&E design had several shortcomings: (i) some outcome indicators were not clearly linked to the PDO statement (access to finance, input, technologies, and markets), and (ii) some of the original indicators were not SMART. For example, “Incremental yearly net cash-flow,” in addition to being difficult to measure, cannot be solely attributable to the project. The mid-term review attempted to correct the design flaws by substantially revising the results framework, but challenges related to interpretation and appropriate level of indicator remained. The MIS was completed at mid-term. The system was also linked to a Geographic Information System (GIS) . However, there was limited success in training in the use of the systems. The only user of the M&E/MIS was the M&E Officer. M&E Implementation 66. Shortcomings in the M&E design contributed to shortcomings in the method, level, and the type of data collected and analyzed. There were also gaps in data and analysis on key project activities affecting development outcomes, for example the use of the matching grants, particularly regarding the price paid to farmers by the cooperatives for the purchase of their crops and the price obtained by cooperatives in the resale of the crops. The firm that conducted the mid-term impact assessment sourced data directly from cooperatives on yields and prices obtained for crops sold by cocoa and coffee cooperatives. However, there were few project activities at mid-term and after the mid-term assessment data collection on volumes sold and prices obtained for rehabilitated tree crops was not maintained. There were also data gaps on unit costs of some investments – including the feeder roads, and/ or delivering services compared with unit costs incurred by similar projects or standard costs commonly used in Liberia or the sub-region. 67. While the project reported an aggregate number of farmers trained, feedback from farmers and implementing partners during the ICR mission signaled there were significant differences in the type and level of trainings. There is no data on the type of trainings provided, and no detailed data on beneficiaries of those trainings. 68. Limitations in the MIS negatively impacted strategic decision making. The project’s M&E system was equipped with Geographic Information System (GIS) applications to enhance the monitoring and analysis of the overall performance of the project, to inform decision making on any required strategic changes, and to provide data for monthly reports. However, the online system failed to deliver detailed information, including on the specific model smallholder farmers adopted, time of adoption, and attrition of smallholder farmers. The system also did not provide adequate information on parameters to effectively monitor risks. 69. Although the MIS was incomplete, it still provided useful and detailed demographic information on project beneficiaries. However, due to the weaknesses in the M&E design, project management’s adjustments to factors adversely affecting project activities – notably the failure of the critical assumptions to materialize and the Ebola outbreak – were limited. Page 27 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) M&E Utilization 70. Despite its design shortcomings and the challenges encountered in implementation, the MIS generated some materials used to inform project beneficiaries, the Ministry of Agriculture, and the World Bank on the status of project activities. Progress on project activities and results were provided in the form of quarterly and semi-annual reports, a documentary video, maps, fact sheets, snapshots and fliers. Additionally, in the early stages of project implementation, radio jingles were produced to inform stakeholders about the project and criteria for participation. The MIS also generated socio-economic descriptions of STCRSP beneficiaries including gender, age, and information on the size of beneficiaries’ farms, which are geo-coded, and this information was used to inform the project’s implementation completion review. However, the M&E data was primarily on the application of inputs and the achievements of outputs as opposed to providing evidence on the achievement of outcomes. Justification of Overall Rating of Quality of M&E 71. Although the MIS provided some useful data, due to design flaws in the M&E system, challenges in its implementation, and the limited use of results, the overall rating for M&E is modest. B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE Environmental and Social 72. Because the project implementation unit had limited capacity for monitoring safeguard issues, an Environmental and Social Management Team (ESMT) was set up under the project, a safeguard implementation structure that emphasized strong collaboration with the Environmental Protection Agency of Liberia and other relevant agencies such as the Ministry of Agriculture. The PIU was also staffed with an Environmental Specialist. Additionally, an Environmental and Social Management Framework (ESMF) and Resettlement Policy Framework was prepared; the ESMF was used to screen beneficiary farms or subprojects; and site-specific Environmental and Social Management Plans (ESMPs) were prepared in most cases before field implementation. 73. No specific environmental or social issues arose during implementation nor, issues surrounding land acquisition; restriction on land use, or benefit sharing. Subproject screening also ensured that farms with contested ownership were not included in the project. The project remained in compliance with the Bank’s environmental and social policies Financial Management 74. At project closing, the project’s financial management (FM) was rated moderately satisfactory given that: (1) the funds for the project were used for the intended purposes with due regard for economy, efficiency and effectiveness, (2) timely and reliable financial reports for the project were then being produced, and (3) measures were in place to protect the assets of the project. The FM system, including accounting, auditing, and reporting, was adequate and consistent with the Bank’s FM guidelines and requirements. The project would have been rated satisfactory, but for the following factors:  At project end, manual accounting was still being used despite numerous attempts to have the project team Page 28 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) to use an accounting software;  Audited reports were submitted late in two financial years; and  A qualified accountant was only hired in the last year of project implementation. 75. Following the recruitment of a qualified FM specialist in January 2018 and the in-depth review of June 2018, the financial management of the project improved, and the FM rating was upgraded from moderately unsatisfactory to moderately satisfactory. At the end, the overall FM arrangements were in place including staffing, controls, accounting, reporting and auditing. The internal audit of the PIU contributed significantly in maintaining project controls. Procurement 76. At project closing, the project procurement performance had a rating of moderately satisfactory. The procurement function suffered during the Ebola outbreak as most activities slowed down. This situation was later worsened by the MoA’s non-renewal of the procurement specialist’s contract during the final year of implementation. This weakened the procurement capacity of the project because the procurement specialist who took over had less experience in World Bank funded procurements procedures. To address this, the Bank procurement team provided hands on support. In addition, the procurement specialist benefitted from clinics organized by the World Bank. C. BANK PERFORMANCE Quality at Entry 77. The following are key issues with quality at entry: (i) The PDOs were highly relevant but the performance indicators were not well aligned with the PDO; (ii) The project design was overly ambitious, particularly given the FCV status of the country. There were multiple activities, multiple models, and multiple targeted counties. The ambitious scope does not align with the fact that the project was supposed to be a pilot intervention; (iii) Too many critical assumptions were made that were not supported during project implementation. In this regard, insufficient attention, if any at all, was given to outlining risk mitigation measures in the operational risk assessment framework. For example, while the reluctance of financial institutions was identified as a risk, the mitigation measure simply indicated that the project provided several incentives to CBL/PFIs to extend credit to smallholders63, without specifying these incentives. Similarly, “unconfirmed commitment of potentially participating concessionaires” was identified as a high risk yet the proposed risk management was to negotiate agreements between MoA, concessionaires, CBL and PFIs for long-term credit and management contracts between each participating concessionaire and MoA64. Furthermore, it was not apparent from project documentation that the required level of engagement with key stakeholders took place before implementation. 63 STCRSP PAD; p.64 64 Ibid, p.63 Page 29 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) In this regard, it appears that some of the critical assumptions, notably regarding the role that concessionaires and input suppliers would play, were not substantiated. Quality of Supervision 78. The Ebola outbreak had a significant impact on supervision. Given that most of the project results were generated after the mid-term restructuring, the Bank team is credited for enabling most of the project achievements to be completed in a reduced timeframe after the Ebola crisis. The Bank team was also proactive in restructuring the project to mitigate some of the challenges faced. However, the focus was primarily on revising indicators and end- targets without also reviewing the scope and timing of the activities and addressing some other critical gaps: the M&E system remained under-developed; the scale of the proposed activities was maintained; and some of the revised indicators remained problematic. Additionally, it does not appear that steps were taken to put in place alternative measures for the farmers to access credit for maintaining their farms at the end of the project, once it was clear that the expected arrangement between MFDP and CBL was not going to take place. The shortcomings in the M&E were particularly consequential given the pilot nature of the project 65. Justification of Overall Rating of Bank Performance 79. The shortcomings in the quality at entry and the quality of supervision placed the achievements of project objectives and the sustainability of the project’s interventions at risk. The quality at entry is rated as moderately unsatisfactory, and supervision moderately satisfactory. Considering the substantive challenges posed by Liberia’s FCV context and the Ebola outbreak, and the project’s achievements in a short period of time under those challenging circumstances, on balance the overall Bank performance is rated as moderately satisfactory 65 STCRSP PAD; p.57 Page 30 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) D. RISK TO DEVELOPMENT OUTCOME 80. Improving access to finance, inputs, technologies, and markets was an important strategic objective of the project, but since systems were not put in place to ensure continued and increased access or the ability of farmers to independently access these factors, the development outcomes that the project was able to achieve are at risk66. However, a new Bank-financed project – the Smallholder Agriculture Transformation and Agribusiness Revitalization Project (STAR-P, P160945) will serve to mitigate some of this risk, given that STAR- P’s activities also address access to markets, finance, technologies, and inputs for smallholder farmers. Additionally, the focus on oil palm, as one of the targeted value chains under STAR-P, allows opportunities for relevant STCRSP beneficiaries to be further supported. Also, the institutional strengthening of the cooperatives, while not fully achieved as anticipated in the project design, will help farmers to build on some of the project’s achievements. V. LESSONS AND RECOMMENDATIONS 81. The following lessons and recommendations were derived from the STCRSP related to: (i) Project design and country context. Due consideration must be given to the country’s context in deciding the project scope and implementation arrangements. For example, in the case of Liberia, an FCV country with extremely limited capacities, project design should be simple and monitoring indicators should be SMART. In addition, project design should remain flexible and responsive to changes in country context. (ii) Required skill sets for project design and implementation. Ensuring that the staff skill sets on the project team are fully aligned with the project objectives and related activities is critical to successful project implementation. In this case, the lack of a marketing/value chain specialist contributed to the inability to negotiate with the private sector and get formal agreements. Further, given the pilot nature of the project intended to generate lessons learned for a larger-scale tree crop development program, the project was expected to have a strong focus on M&E results67 and therefore required the support of an M&E specialist. (iii) The level and timing of consultations with implementing partners. Engaging in extensive upfront consultations/negotiations to agree on prerequisites with potential partners is required, particularly with the private sector and in an FCV context. Such consultations should allow for due diligence on those partners and, at least, outlines of memorandum of understandings with key partners to be ready at the beginning of the project. (iv) Timing and sequencing of project activities. Good timing/sequencing of project activities is important to ensure outcomes can be achieved and lessons learned. For example, in the case where processing factories and warehouses are to be built, if these are done relatively early in the project, the benefits from access to these facilities could contribute to generating increased income and therefore increased access to finance for 66 The project focused on at least three areas for sustainability after project ended: (i) out-grower schemes between smallholders and concessionaires/large estate; (ii) the long-term credit scheme; and (iii) the long-term, large-scale tree crops program (see eg STCRSP PAD, p. 30, 61) 67 Ibid, pps10, 57 Page 31 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) reinvestment into project activities. Further, given the time it takes for tree crops to mature, such projects, particularly where new planting, replanting is involved, require more than four years, especially in an FCV context, to test results. (v) Preparation of ESMPs. Following the framework approach for environmental and social safeguards, subprojects of similar characteristics in the same region should be combined under one ESMP. With numerous subprojects, developing an ESMP for each sub-project is not feasible. Combining several similar sub-projects under one ESMP worked very well for this project. The approach proved more practical, less time consuming, and very cost-effective. 82. The following recommendation is specific to Liberia: (i) Implementation of Liberia Cooperatives Act. The response from cooperative members to training offered by the project showed that there is an appetite for adopting market-oriented models to cooperative development. Farmers are incentivized once they realize their incomes can increase by using a business approach rather than depending on the government to provide the type of support traditionally expected under public extension services. The Government of Liberia should be encouraged to promote the passage of the revised draft Cooperatives Act and to facilitate awareness raising and implementation of its provisions (ii) . among relevant stakeholders. Page 32 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS A. RESULTS INDICATORS A.1 PDO Indicators Objective/Outcome: Area of smallholder tree crop farms rehabilitated, replanted or planted under the project Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Indicator One : Area of Hectare(Ha) 0.00 11300.00 6056.00 5828.57 smallholder tree crop farms rehabilitated, replanted or 18-Jul-2012 28-Sep-2016 30-Nov-2018 30-Nov-2018 planted under the project, disaggregated New Planting Hectare(Ha) 0.00 2000.00 300.00 630.00 18-Jul-2012 28-Sep-2016 30-Nov-2018 30-Nov-2018 Replanting Hectare(Ha) 0.00 2200.00 1250.00 1004.00 18-Jul-2012 28-Sep-2016 30-Nov-2018 30-Nov-2018 Rehabilitation Hectare(Ha) 0.00 7100.00 4506.00 4194.57 Page 33 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) 18-Jul-2012 28-Sep-2016 30-Nov-2018 30-Nov-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Area of smallholder tree crop Hectare(Ha) 0.00 11300.00 6056.00 5828.57 farms rehabilitated, replanted or planted under 04-Dec-2017 28-Sep-2016 30-Nov-2018 30-Nov-2018 the project, disaggregated Comments (achievements against targets): Objective/Outcome: Yield of major tree crops Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion PDO Indicator 2: Yield of Metric ton 0.12 1.30 1.30 0.54 major tree crops 18-Jul-2012 30-Nov-2018 30-Nov-2018 30-Nov-2018 Yield of High input Metric ton 0.96 1.30 1.30 1.00 Rehabilitated Cocoa 18-Jul-2012 30-Nov-2018 30-Nov-2018 30-Nov-2018 Yield of medium input Metric ton 0.12 0.50 0.50 0.46 rehabilitated cocoa 18-Jul-2012 30-Nov-2018 30-Nov-2018 30-Nov-2018 Page 34 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Yield of Rehabilitated Coffee Metric ton 0.12 0.50 0.50 0.30 18-Jul-2012 30-Nov-2018 30-Nov-2018 30-Nov-2018 Yield of Rehabilitated Oil Metric ton 1.50 5.00 5.00 1.50 Palm 18-Jul-2012 30-Nov-2018 30-Nov-2018 30-Nov-2018 Comments (achievements against targets): Objective/Outcome: Smallholder tree crop farmers with access to finance inputs, market and technologies as a result of Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion PDO Indicator 3: Smallholder Number 0.00 3025.00 3025.00 3750.00 tree crop farmers with access to finance inputs, market and 18-Jul-2012 30-Nov-2018 30-Nov-2018 30-Nov-2018 technologies as a result of project intervention Comments (achievements against targets): Objective/Outcome: Long term large scale tree crop development program formulated and approved by Ministry of Agricultu Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Indicator 4: Long term large Text No Yes Yes No Page 35 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) scale tree crop development 18-Jul-2012 30-Nov-2015 30-Nov-2018 30-Nov-2018 program formulated and approved by Ministry of Agriculture Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Direct project beneficiaries Number 0.00 5000.00 5000.00 3850.00 18-Jul-2012 28-Sep-2016 28-Sep-2018 30-Nov-2018 Female Beneficiaries Number 0.00 1000.00 750.00 900.00 18-Jul-2012 28-Sep-2016 28-Sep-2018 30-Nov-2018 Comments (achievements against targets): A.2 Intermediate Results Indicators Component: Component 1: Smallholder Tree Crops Revitalization Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Component 1 : Intermediate Number 0.00 1.00 1.00 1.00 Results Indicator 1.2 : Farmer's run Oil Palm 18-Jul-2012 01-Jun-2016 30-Nov-2018 30-Nov-2018 Page 36 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) PlantationCooperative strengthened Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Component 1 : Intermediate Number 0.00 20.00 20.00 16.00 Results Indicator 1.3 : FOs strengthened in out 18-Jul-2012 01-Jun-2016 30-Nov-2018 30-Nov-2018 growerschemes Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Component 1 : Intermediate Metric ton 0.00 1000.00 1000.00 850.00 Results Indicator 1.4 : Volume of cocoa bulked by 18-Jul-2012 01-Jun-2016 30-Nov-2018 30-Nov-2018 project supportedFOs Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Component 1: Intermediate Number 0.00 3.00 3.00 1.00 Page 37 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Results indicator 1.5: Out- 18-Jul-2018 30-Nov-2018 30-Nov-2018 30-Nov-2018 grower schemes with long term credit to participating smallholder farmers supported by the project Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Component 1 : Intermediate Number 0.00 43.00 43.00 119.00 Results Indicator 1.1 : Cocoa Farmer Organizations (FOs) 18-Jul-2012 01-Jun-2016 24-Sep-2018 30-Nov-2018 strengthened Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Component 1; Intermediate Kilometers 0.00 145.00 30.00 5.00 Results Indicator 1.6: Farm access roads rehabilitated 18-Jul-2012 01-Jun-2016 28-Sep-2018 30-Nov-2018 Comments (achievements against targets): Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Formally Revised Completion Page 38 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Target Intermediate Result Indicator Text No Yes Yes No 2.1: Capacity building plan for MoA & CDA developed 18-Jul-2012 01-Jun-2016 01-Jun-2017 30-Nov-2018 Comments (achievements against targets): The CDA asked the project to assist with updating the National Cooperative Development Act instead of developing an FO policy and Cooperative Strategy. The project contributed to the update of the Act instead of developing the policy and strategy Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Intermediate Results Number 0.00 70.00 70.00 5.00 Indicator 2.2: Persons trained 18-Jul-2012 30-Nov-2016 30-Nov-2018 30-Nov-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Intermediate Results Number 0.00 5.00 5.00 5.00 Indicator 2.3: County Agricultural Offices equipped 18-Jul-2012 01-Jun-2016 30-Nov-2018 30-Nov-2018 Comments (achievements against targets): Page 39 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Intermediate Results Number 0.00 5.00 5.00 5.00 Indicator 2.4: CDA field offices equipped 18-Jul-2012 01-Jun-2016 30-Nov-2018 30-Nov-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Intermediate Results Text No Yes Yes No Indicator 2.5: National Policy and Strategy for FO and 18-Jul-2012 30-Nov-2015 30-Nov-2018 30-Nov-2018 Cooperative development formulated and validated with all national stakeholders Comments (achievements against targets): The Project was requested to help update the Cooperative Development Act instead of developing a FO policy and strategy. The project contributed to the update Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Intermediate Results Text No Yes Yes Yes Page 40 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Indicator 2.6: Methodology 18-Jul-2012 01-Jun-2016 30-Nov-2018 30-Nov-2018 for community land use rights validation agreed with stakeholders Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Intermediate Results Text No Yes Yes Yes Indicator 2.7: Tree crop master plans formulated in a 18-Jul-2012 28-Nov-2014 30-Nov-2018 30-Nov-2018 participatory process under MoA leadership- Cocoa - Coffee - Oil Palm - Rubber Rubber Text No Yes Yes Yes 18-Jul-2012 28-Nov-2014 30-Nov-2018 30-Nov-2018 Oil Palm Text No Yes Yes Yes 18-Jul-2012 28-Nov-2014 30-Nov-2018 30-Nov-2018 Coffee Text No Yes Yes No 18-Jul-2012 28-Nov-2014 30-Nov-2018 30-Nov-2018 Page 41 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Cocoa Text No Yes Yes Yes 18-Jul-2012 28-Nov-2014 30-Nov-2018 30-Nov-2018 Comments (achievements against targets): Instead of 4 master plans. One master plan was developed that targeted cocoa, oil palm, and rubber Page 42 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) B. KEY OUTPUTS BY COMPONENT Note: The linkage between the objective/outcomes and indicators were not clear. This table captures only some of the intermediate indicators and some of the PDO and intermediate level indicators are linked to objectives/outcomes by inference. Outputs that the ICR were not able to confirm and/or for which there was conflicting data is noted by an asterisk*. Objective/Outcome 1: Increase access to finance/Access to finance increased 1. Smallholder tree crop farmers with access to finance, inputs, markets, and technologies Outcome Indicators 2. Area of tree crops farms rehabilitated, replanted, or planted under the project 3. Yield of major tree crops 1.Long-term credit provided to 3 out-grower schemes 2. Area of tree crops farms rehabilitated, replanted, or planted under Intermediate Results Indicators the project 3. Yield of major tree crops 1. US62,390 in matching Grants provided to 12 cooperatives 2.US$115, 526 provided in in-kind credit to 1 cooperative 3.3,750 farmers had access to finance* 4.The total area of tree crops farms rehabilitated, replanted, or planted was 6,056 ha Key Outputs by Component 5. The following yield per metric ton were achieved for (linked to the achievement of the Objective/Outcome 1) -High input rehabilitated cocoa: 1.00 -Medium input rehabilitated cocoa: 0.47 -Rehabilitated coffee: 0.30* -Rehabilitated oil palm: 1.50 Page 43 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Objective/Outcome 2: Increase access to inputs/Access to inputs increased; Area of tree crops production increased; Yields of tree crops increased 1.Smallholder tree crop farmers with access to finance, inputs, markets, and technologies Outcome Indicators 2. Area of tree crops farms rehabilitated, replanted, or planted under the project 3. Yield of major tree crops Intermediate Results Indicators 1. US$2,196,155 were provided in grants to purchase inputs 2. US$115, 526 provided in in-kind credit to 1 cooperative were allocated for purchase of fertilizers, pesticides 3.The total area of tree crops farms rehabilitated, replanted, or planted was 6,056 ha Key Outputs by Component 4. The following yield per metric ton were achieved for (linked to the achievement of the Objective/Outcome 2) -High input rehabilitated cocoa: 1.00 -Medium input rehabilitated cocoa: 0.47 -Rehabilitated coffee: 0.30* -Rehabilitated oil palm: 1.50 5. 3,750 farmers had access to input* Objective/Outcome 3: Increase access to technologies/Access to technologies increased; Area of tree crops production increased; Yields of tree crops increased 1.Smallholder tree crop farmers with access to finance, inputs, markets, and technologies Outcome Indicators 2. Area of tree crops farms rehabilitated, replanted, or planted under the project 3. Yield of major tree crops Intermediate Results Indicators Page 44 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) 1. The total area of tree crops farms rehabilitated, replanted, or planted was 6,056 ha 2. The following yield per metric ton were achieved for -High input rehabilitated cocoa: 1.00 Key Outputs by Component -Medium input rehabilitated cocoa: 0.47 (linked to the achievement of the Objective/Outcome 3) -Rehabilitated coffee: 0.30 -Rehabilitated oil palm: 1.50 3. 3,750 farmers had access to technologies* Objective/Outcome 4: Increase access to markets/Access to markets increased 1.Smallholder tree crop farmers with access to finance, inputs, Outcome Indicators markets, and technologies 1. Km of road rehabilitated Intermediate Results Indicators 2.Farmer organizations strengthened in out-grower schemes 3. Cocoa farmer organizations strengthened Key Outputs by Component 1. Five kilometers of road were rehabilitated (linked to the achievement of the Objective/Outcome 4) 2. 3,750 farmers had access to markets* Objective/Outcome 5: Develop a long-term development program for the tree crops sector/ long-term development program developed 1. Long-term, large-scale tree crop development program formulated Outcome Indicators and approved by the Ministry of Agriculture 1.Capacity building plan for MoA & CDA developed 2.Persons trained Intermediate Results Indicators 3. Tree Crops master plans formulated in a participatory process under MoA leadership -Cocoa Page 45 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) -Coffee -Oil Palm -Rubber 1. Support was provided to create the Liberia Agriculture Commodities Regulatory Agency (an agency of the MoA) by assisting with feasibility studies and elaboration of, among other documents, the LACRA Bill 2. Five persons were provided financial assistance to complete their Key Outputs by Component final year of PhD program in applied human nutrition, entomology, (linked to the achievement of the Objective/Outcome 2) agricultural economy, and crop science. 3. One master plan was completed, which assessed cocoa, oil palm, and rubber. This was accompanied by an action plan for promoting each of the tree crop Page 46 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION A. TASK TEAM MEMBERS Name Role Preparation Supervision/ICR Abimbola Adubi Task Team Leader(s) Anas Abba Kyari, Komana Rejoice Lubinda Procurement Specialist(s) Saidu Dani Goje Financial Management Specialist Ferdinand Tsri Apronti Team Member Jeanette M. Sutherland Team Member Yeyea Gloria Kehleay Nasser Team Member Vaanii O. Baker Team Member Sachiko Kondo Team Member Zoe Quoi Diggs Duncan Team Member Gloria Malia Mahama Social Specialist Sekou Abou Kamara Environmental Specialist Akhilesh Ranjan Social Specialist B. STAFF TIME AND COST Staff Time and Cost Stage of Project Cycle No. of staff weeks US$ (including travel and consultant costs) Preparation FY09 38.204 103,170.59 FY10 19.885 204,014.02 Page 47 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) FY11 17.125 242,744.60 FY12 18.725 289,130.04 FY13 0 0.00 Total 93.94 839,059.25 Supervision/ICR FY13 8.300 125,544.78 FY14 13.776 88,325.59 FY15 11.560 83,646.21 FY16 20.430 140,642.24 FY17 19.224 95,723.04 FY18 14.518 64,232.27 FY19 33.309 185,815.87 Total 121.12 783,930.00 Page 48 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) ANNEX 3. PROJECT COST BY COMPONENT Amount at Approval Actual at Project Percentage of Approval Components (US$M) Closing (US$M) (%) Component 1: Smallholder 9.467 8.613 91 Tree Crops Revitalization Component 2: Institution Building and Preparation of 3.056 2.078 68 Future Large Scale Tree Crop Development Program Component 3: Project Coordination and 2.477 3.001 121 Management Total 15.00 13.692 91 Page 49 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) ANNEX 4. EFFICIENCY ANALYSIS This annex details the efficiency analysis and covers: (A) Economic and Financial analysis of key project interventions; (B) operational efficiency analysis, which includes a comparison of actual project costs with appraisal estimates; and (C) analysis of administrative and operational aspects of project implementation and management. A. Economic and Financial Analysis68 The ex-post economic and financial analysis follows the same methodology as the ex-ante analysis conducted at the time of appraisal, which was based on a with-and-without simulation of the project net benefit stream. However, the ex-post analysis was conducted using actual project costs and benefits data to date, collected by an independent consulting firm, and forecast benefits and costs based upon the physical data provided by the project and augmented by price projections from the World Bank database on international commodity prices. Economic costs and prices were estimated by transforming financial prices, costs, and benefit streams into economic values through calculating economic import/export parity prices at farm gate, applying conversion factors for each category of costs: eliminating taxes, subsidies and other transfer flows. It also took into consideration incremental costs beyond the project implementation period. Economic benefits considered in the analysis are those derived from increased tree crop production from project supported farmers. Moreover, the net benefit streams were projected for 3,518 direct project beneficiaries of the tree crops revitalization/replanting program and 5,828.6 ha covered under the six farm models. The Analysis the six tree crop models piloted by STCRSP program covering four crops (Cocoa, Coffee, Rubber and Oil palm) in six counties in Liberia. These models include: (1) High input cocoa revitalization; (2) Medium input cocoa revitalization; (3) New planting of cocoa (medium input); (4) Coffee revitalization; (5) Oil palm new and replanting; and (6) Rubber new and replanting. The implementation approach was similar in each of the six counties, supporting smallholder farmers to revitalize (under-brushing, pruning, shade management, etc.), replant or newly plant one of the four tree crops in an area where previous tree crop plots were abandoned or have been under-utilized since the Liberian civil war. The analysis reviews the costs and benefits related to cocoa, coffee, rubber and palm oil across six counties of Liberia, focusing on beneficiary income improvements resulting from project support. These average costs and benefits are then multiplied by the number of beneficiaries under each of the models to understand the impact of the project as a whole. Methodology The analyses rely extensively on data from the September 2018 Intermediate Financial Report, beneficiary costs estimated based on labor estimates in the 2013 STCRSP Project Implementation Manual (PIM) and daily labor costs from the 2013 STCRSP Project Appraisal Document (PAD), actual beneficiary numbers and land areas from the online Monitoring Information System (MIS) - adjusted to reflect field data, and yield and price data from the 2015 baseline surveys and 2017 impact survey. Yield and price variations have been built-in based on price expectations from the World Bank’s Data Portal and yield variations 68The Results of the EFA presented in the section is derived largely from the End of Project Impact Evaluation conducted by LTS International, November 30, 2018. Page 50 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) based on the Project Appraisal Document. All values have been adjusted to 2018 commodity prices. It also combines the information from field interviews and the secondary data collected as part of the End of Project Impact Evaluation. The information collected from the interviews and secondary research was used to generate a set of assumptions about the range of costs and benefits from the project per beneficiary. This was then combined with beneficiary data from the online monitoring information system and cost estimates based on the allocation of Component 1 costs (crop revitalization) against tree crop types (cocoa/coffee, oil palm and rubber). Given the long payback period for tree crops, the analysis follows the approach used for the ex-ante economic and financial analysis undertaken as part of the Project Appraisal Document and models the impacts over a 25-year period. The model generates a cost-benefit ratio and net present value using a 10 percent discount rate. In addition, the analysis has been run from the beneficiary perspective – excluding project costs and considering an EFA for a single beneficiary for each of the tree crop models. For the beneficiary models, a 10-year time horizon is used as poverty and lack of access to credit mean that beneficiaries are likely to have much shorter time preferences. Sensitivity analysis was then undertaken for the project model, considering the impact of reducing the time horizon to ten years, the impact if 20 percent of beneficiaries do not continue with the activities beyond Year 3, as well as cost overruns and reductions in benefits ranging from 25 to 55 percent. Project Costs: The project costs are based on the September 2016 Interim Financial Report, which reports Component 1 costs disaggregated by cocoa/coffee, oil palm and rubber (refer to Table 4.1). This data was used to estimate project cost per beneficiary data for each of the four commodities. The average cost per STCRSP beneficiary was US$2,212, although the cost per beneficiary was lower for cocoa and coffee (US$1,877) compared to oil palm and rubber (US$5,081 and US$3,175 respectively). This may be partially explained by the economies of scale achieved for cocoa, which had significantly more beneficiaries (by around a factor of ten). In addition, the cost of implementing oil palm and rubber were higher than anticipated as the project experienced challenges engaging the private sector and ended up taking on more responsibility for the provision of training and farm inputs than originally anticipated. Table 4.1: Project Costs Total Cocoa Coffee Oil palm Rubber Rehabilitation. Rehabilitation. New Rehabili New New (med. Input) (high input) planting tation. planting planting Project costs (US$) 7,780,463 5,488,693 1,082,185 1,209,585 Beneficiaries (#) 3518 2354 79 406 85 213 381 Cost per beneficiary 2,212 1,877 5081 3,175 (US$) Beneficiary Costs: Beneficiary costs were calculated based on the number of person days required to plant and/or revitalize the smallholder farmers’ tree crops, as well as days required to maintain the tree crops Page 51 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) costed based on an adjusted economic cost of labor of ($2.2674). As well as the cost of fertilizer and chemical inputs, assumed to be insecticide and fungicide (at US$40 each) for all tree crops, plus fertilizer for high input cocoa (at US$65). Given the lack of estimates for oil palm and rubber, their costs are assumed to be similar to coffee, as all three crops require more maintenance than cocoa. Table 4.2: Beneficiary Costs Cocoa Cocoa Cocoa new Coffee Oil palm Rubber rehabilitation. Rehabilitation planting Rehabilitation Med Input. High Input Year 0 192 224 406 146 352 544 Year 1 142 197 131 209 255 395 Year 2 148 199 137 209 255 395 Year 3 148 206 125 209 255 395 Year 4 148 206 91 209 255 395 Year 5 148 213 94 209 255 395 Year 6 148 213 99 209 255 395 Year 7 148 213 108 209 255 395 Year 8 148 213 114 209 255 395 Year 9 148 213 122 209 255 395 Year 10+ 148 213 134 209 255 395 Benefits: The EFA focuses largely on estimated beneficiary income benefits from the revitalized tree crop plantations. However, increased access to credit and improved incomes are likely to lead to additional benefits in the form of improved resilience, food security and access to education. As such, the benefits of the project are under-stated. The income benefits of the project are estimated based on additional tree crop yield as a result of the project, multiplied crop prices. Indirect benefits, such as those from the rehabilitated feeder roads facilitating access to markets for communities around the road corridors, facilitation of businesses for cooperatives, as well as the benefits accruing from newly constructed and rehabilitated infrastructure including warehouses and oil palm processing centers, are not taken into account because they were not been completed in good time for the benefits to flow at the of the ICR. The investments in oil palm processing centers, for instance, were completed during the final year of the project for their full utilization. Table 4.3: Tree Crop Harvesting Cocoa Coffee Oil palm Rubber Rehabilitation, Rehabilitation. New planting Rehabilitation New planting New planting Medium input High Input Year of first 0 0 3 (low) 1 4 5 harvest Peak year 2 5 10 3 12 11-12 Crop type Dried beans Dried cherries Fresh fruit Dry rubber bunches Harvest Pods picked Cherries picked Tapped method Source: STCRSP Assessment Document, 2013 Page 52 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Yields: Tree crop yields vary depending on the age of the tree. The 2013 Project Assessment Document provided estimates of changes in yield over time for the cocoa and coffee models (refer to Table 4.3), including detailed estimates of how yields will vary for four of the STCRSP models (the cocoa and coffee models). This information has been used to model variations in yield over time (adjusted down based on actual yields reported in the 2017 survey see Table 4.4). The PAD did not include yield variations for oil palm and rubber, their yields are therefore modelled as constant once harvesting commences in Years 4 and 5 respectively. To simplify the analysis, the model assumes all new tree crops were planted in the first year of the project. Table 4.4: Project Yields Cocoa Cocoa Cocoa Coffee Oil palm Rubber Rehabilitation. Rehabilitation Replanting Medium input High input Kg/ha Kg/ha Kg/ha Kg/ha Kg/ha Kg/ha Year 0 225 456 Year 1 274 730 101 Year 2 304 791 61 126 Year 3 365 913 122 151 Year 4 365 973 183 151 3,747 Year 5 365 1,095 243 151 3,747 102 Year 6 365 1,095 304 151 3,747 102 Year 7 365 1,095 365 151 3,747 102 Year 8 365 1,095 487 151 3,747 102 Year 9 365 1,095 608 151 3,747 102 Year 10 365 1,095 730 151 3,747 102 Without Project Scenario: In the case of cocoa, STCRSP piloted three models: medium input revitalization, high input revitalization and new plantings. As such, only the additional yield from the rehabilitation activities and modern farming practices are included in the EFA models. The 2015 baseline survey reveals that 59.9 percent of smallholder farmers were already harvesting from the existing trees, albeit without using chemicals or fertilizers. For these farmers, therefore, yields are of the 279 kilograms per hectare cocoa yield reported in 2017 for non-beneficiaries from the Project impact survey. In the case of coffee, farmers had not been maintaining or harvesting from their trees, and in many cases had already converted the coffee orchards to rice paddies, as such the full harvest could be attributed to the result of the project. On the other hand, new plantings are a new tree crop and the full yield is therefore included in the model. This is the case for new cocoa plantings, as well as the replanting and new planting of oil palm and rubber tree crops. Given that both replanting and new planting involve the planting of new trees, productivity and income effects are expected to be the same for both types of beneficiaries and the model therefore treats them as the same. However, given the long timelines between planting and harvesting of oil palm and rubber, these beneficiaries have not yet started harvesting, yield and price data is therefore based on estimates from existing farmers in the project location (2017 Impact Survey). Therefore, without project production could only been considered for old cocoa plantations. Page 53 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Commodity Prices Price data for the EFA is based on World Bank actual and forecast price data for the four tree crops (see Table 4.5). Given that the data is annual to 2025, prices between 2025 and 2030 are based on the average of these two years, while prices from 2030 onwards are assumed to remain constant. For consistency across commodities, prices have been converted to USD per ton. Table 4.5: Tree Crop Price Forecasts69 Nominal Units 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2030 USD Cocoa $/mt 3,100 3,100 2,900 2,000 2,300 2,400 2,400 2,500 2,500 2,600 2,600 2,700 3,000 Coffee $/mt 2,200 1,900 2,000 2,200 1,800 1,800 1,900 1,900 1,900 2,000 2,000 2,000 2,200 (robusta) Palm oil $/mt 821 559 639 649 570 592 615 639 664 690 716 744 900 Rubber $/mt 2,000 1,600 1,600 2,000 1,600 1,600 1,700 1,800 1,800 1,900 1,900 2,000 2,400 Based on the 2017 survey, smallholder reported farm gate prices are lower than the published national rates (refer to Table 4.6). There is a perception among some farmers and cooperative members that middlemen purchasing tree crop harvests may not offer farmers fair prices. Although in 2017 88 percent of smallholder cocoa farmers reported access to information on market prices, this is usually received through middlemen and may therefore not be reliable.70 Nevertheless, farmers reported being better able to negotiate prices since SOCODEVI shared the Cocoa Market Indicative Price list (published annually by the Liberia Produce and Marketing Corporation).71 However, the actual prices smallholder farmers reported are between only 29.2 percent of national prices for coffee and 98.8 percent for oil palm. Table 4.6: Comparison of World Bank and Farmer Reported Tree Crop Prices Cocoa Coffee Oil palm Rubber World Bank price (US$/Mt) 2,000 2,200 649 2,000 Farm Gate price (US$/Mt) 1,607 642 641 1,560 Percent 80.3% 29.2% 98.8% 78.0% Source: WB's Liberia Data Portal (historical and long-term forecasts for prices). Updated 2/11/18. Accessed 14/11/18. Additionally, the value-add activities undertaken by the farmers and cooperatives – such as improvements in drying, fermentation – are expected to result in better quality outputs, which should in turn result in better prices. However, there was no evidence of a price premium resulting from the project nor of better prices received by farmers as a result of selling their produce through cooperatives. Cocoa exporters spoken to as part of this evaluation noted that there is very little grading of cocoa currently and higher quality production may therefore not receive a premium price, this can partly be attributed to the small 69 World Bank’s Data Portal (historical and long-term forecasts for prices). Updated 2/11/18. Accessed 14/11/18 70 Egis (2017). Smallholder Tree Crop Revitalization Support Project – Final Impact Study, p22 71 Egis (2017). Smallholder Tree Crop Revitalization Support Project – Final Impact Study, p54. Page 54 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) volumes of high-grade cocoa therefore being mixed into lower quality batches. It will be important to address this issue if a national program is funded and to continue monitoring prices to see whether, in the future, a premium price can be achieved and supported by farmers or an organic production pilot project which could target niche single origin cocoa market and target the increased farmgate price this brings. For the beneficiary model, World Bank price data was therefore used to estimate the prices the farmers have/will received based on proportion received in 2017. Whereas, the project model uses the full national commodity prices, as it is assumed that the difference in prices will be captured by traders within Liberia. Conversion of Financial prices to Economic: Economic cost and prices were estimated by transforming financial prices, costs, and benefit streams into economic values through calculating economic import/export parity prices at farm gate, applying conversion factors for each category of costs: eliminating taxes, subsidies and other transfer flows. Table 4.7: Conversion factors Cocoa coffee Oil palm Rubber Fertilizer chemicals Conversion factor 1.43 1 1.25 0.68 0.9 0.9 Beneficiary Income Based on the above yield and price estimates, combined with average area cultivated, annual smallholder farmer incomes have been estimated (refer to Table 4.8). This data was then combined with the project and beneficiary cost data to estimate the cost-benefit ratios of the six models from the beneficiary’s perspective, as well as the project’s net present value and internal rate return. From this it can be seen that given current farm gate prices received by beneficiaries, coffee offers beneficiaries a significantly lower annual income than the other crops. Table 4.8: Beneficiary Income (USD) Cocoa - Cocoa - Cocoa - new Coffee Oil palm Rubber revitalization revitalization planting (med)* (high)* USD/HH USD/HH USD/HH USD/HH USD/HH USD/HH Year 3 697.6 1,868.4 316.1 193.1 2,054.9 2,494.6 * Note that oil palm and rubber yields are indicative of those years as they do not start harvesting until years 4 and 5 respectively. Also note that the medium and high input cocoa revitalization incomes presented are full income, whereas income net of non-beneficiary income is included for 59.9% of these beneficiaries in the CBA model. Also note that the project model includes national crop prices, which are higher than the prices received by the beneficiaries in this table. Results of the economic and financial analysis The analysis indicates that the STCRSP tree cropping investment has the potential for a significant positive impact on the livelihoods of smallholder tree crop farmers in Liberia (Table 4.9). If the assumptions used Page 55 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) to develop the analysis hold, the stream of economic benefits from project-funded activities over the period of twenty-five years (at a 10 percent discount rate) generates a Net Present Value (NPV) of US$26.7 million, an Economic Internal Rate of Return (EIRR) of 29.2 percent, and the Benefit Cost Ratio (BCR) of 2.99. Implying, for every $1 invested (by the project and beneficiaries), the tree cropping interventions will result in approximately $3.0 in benefit to smallholder famers. Table 4.9: Present value of Economic and Financial Flows Economic and Financial Performance Present Value of Flows coefficients Financial analysis Economic Analysis ICR estimates Appraisal ICR estimates Benefit cost ratio 2.8 - 2.99 Rates of Return 21.7% 33% 29.2% Net Present Value US$24.7 million US$28 million US$26.7million Sensitivity Analysis: Sensitivity analysis of the results was conducted to assess the impact of various scenarios of changes in project benefits and costs, including reducing the project timeframe to ten-years, adding all project cost (including components two and three) on the prospects of the investments in the tree crops (Table 4.10). In addition, the risk of losing 20 percent of beneficiaries once project support discontinues was considered by modelling a 20 percent reduction in beneficiaries from Year 3. The results of the sensitivity analysis show that the project returns are highly robust and indicates a solid resilience to increases in costs and reduction in benefits. The sensitivity analysis demonstrates that under the scenarios of 20 percent or more cost increase (including counting the entire project cost by adding components 2 and 3), the corresponding EIRRs remain high with substantial positive NPVs. Similar figures are also observed in the scenarios of decrease in Benefits. Indeed, it will take switching values for a rise in costs of over 50 percent (50.2 percent) and a simultaneous reduction in benefits of over 55 percent for the entire project not to be economically viable (Table 4.10). Table 4.10: Project Sensitivity Analysis BCR NPV (US$) EIRR Ten (10)-year timeframe 1.68 7,699,479 23.7% Twenty (20) percent reduction in number of beneficiaries 2.64 20,473,906 24.5% Total cost (including Components 2 and 3) 2.20 21,861,303 15.0% 20% fall in benefits 2.00 18,640,627 19.6% Thirty (30) percent reduction in benefit 2.09 14,632,679 15.3% Forty (40) percent reduction in benefit 1.79 10,624,731 11.2% Fifty (50) percent reduction in benefit 1.49 6,616,783 7.2% Fifty-five (55) percent fall in benefit 1.34 4,612,809 5.1% Benefits falls by 40 and cost rises by 20% 1.43 7,940,140 7.0% Twenty (20) rise in cost 2.48 23,971,931 21.1% Page 56 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Fifty (50) rise in cost 1.99 19,945,044 13.9% Benefits fall 20% and cost rise 20% 1.16 15,956,035 13.9% Similar positive results run the analysis for the crop models, except for the Coffee model (Table 4.11). Investment in coffee was found to have a negative return. With a BCR of only 0.33, suggesting that for every $1 invested in input and labor, farmers are likely to only recoup 30 cents in benefits. This may be partially due to the low yields achieved by STCRSP coffee beneficiaries (only 60.0 percent of the expected yields), as well as the extremely low farm gate prices farmers are attracting (only 29.2 percent of national prices). currently there is low demand for coffee from larger buyers which may also contribute to the presently depressed price. A possible explanation for the low yields is the high labor requirement and lack of finance to support farmers in paying for this labor. In the case of oil palm, beneficiaries achieve a BCR of 3.87 (or $3.37 for every $1 invested). The results are equally good for cocoa and rubber. Table 4.11: Results EFA for Farm Models (25 years, 10% DR, including project and beneficiary cost) Economic Analysis Financial Analysis Crop Model BCR NPV EIRR BCR NPV FIRR Cocoa (rehabilitation/medium input) 2.32 9,791,201 98.9% 1.62 4,614,817 19.0% Cocoa (rehabilitation/high input) 4.23 954,684 n/a* 2.93 575,255 n/a* Cocoa (new planting) 5.28 5,446,177 22.4% 3.69 3,425,439 16.4% Coffee 0.33 -208,147 n/a* 0.19 -253,223 n/a* Oil palm 3.37 3,564,765 16.4% 3.78 4,191,611 18.2% Rubber 3.72 7,107,843 20.4% 5.64 12,115.301 28.6 *no IRR for beneficiaries because project upfront costs excluded, therefore Y0 is positive And from the beneficiary perspective, the economic and financial analysis was also undertaken for each of the six crop models (Table 4.12). Over a ten-year period, with a 10 percent discount rate, the analysis found that, except for coffee, the investment in tree crops is likely to provide a positive return for farmers for all the other five models (the three cocoa models, oil palm and rubber). High input cocoa revitalization was particularly successfully, creating a return of around $6.73 for every $1 invested by the smallholder farmers. Table 4.12: Result of EFA from Beneficiary Perspective (10 years, 10% DR, only beneficiary costs and benefits) Crop Models Economic Analysis Financial Analysis BCR NPV EIRR BCR NPV FIRR Cocoa (rehabilitation/med input) 4.40 3,539 n/a 3.08 2,162 n/a* Cocoa (rehabilitation/high input) 6.73 8,098 n/a 4.60 5,207 n/a* Cocoa (new planting) 4.22 3,444 46.4% 2.95 2,087 36.3% Coffee 0.54 -621 n/a 0.30 -940 n/a* Oil palm 4.82 6,968 47.5% 5.42 8,054 50.6% Rubber 3.50 7,037 39.5% 5.30 12,113 50.7% *no IRR for beneficiaries because project upfront costs excluded, therefore Y0 is positive Page 57 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Limitations of the Analysis: These results are based on a number of assumptions and the quality of some of the data – in particular beneficiaries’ ability to estimate annual yields and prices – is relatively subjective. As such, these results should be interpreted as indicative findings only. Conclusions of economic and financial analysis Based on these assumptions, at a cost per beneficiary of US$2,212, the analysis suggests that, over a twenty- five- year period, with a discount rate of 10 percent for every $1 spent there is a benefit of about $3 created for Liberian smallholder tree crop growers. In addition, the expected economic net present value of productivity increases is US$26.7 million. This indicates that the project is achieving its intended impact of benefiting smallholder farmers. Nevertheless, the EFA results varied by tree crop model, overall the new planting cocoa, high input cocoa revitalization, oil palm and rubber models were the most successful (with BCRs ranging from about $4 to $5 per $1 invested), while coffee results in negative returns. Investment in coffee returns a negative net present value of US$208,147 and a BCR of only 0.33. Based on this analysis, it is not recommended that any future phase of the project should continue to support the coffee model. B. OPERATIONAL/IMPLEMENTATION EFFICIENCY This section assesses operational efficiency of STCRSP implementation. It includes (i) implementation efficiency, including a comparison of actual component costs at the end of the project with estimated component costs at appraisal/revised cost at during project restructuring, (ii) time overrun/underrun; (iii) analysis of operational and administrative cost of project management unit. Implementation efficiency The project was implemented in a very fragile and difficult socio-economic environment. It encountered implementation delays, following a 11.4 months effectiveness lag. The implementation of project activities had barely started on the ground when the devastating Ebola virus disease pandemic hit the country in 2014. Moreover, institutional weaknesses and capacity challenges led to slow project startup and weak implementation performance. This was corrected through project restructuring and the 2-year project closing date extension, which allowed the project to have a significant turn-around. Budget and disbursement The estimated project costs at appraisal was US$23.1 million (Table 4.13), including IDA credit, and in-kind contributions from the Government of Liberia, smallholder farmers (inputs and labor) and participating financial institutions (establishment of rural micro-finance institutions, co-financing cocoa collateral management costs and small working capital loans). However, due to difficulties in measuring in-kind contributions from non-donors, the budget was revised as part of the project restructure to the IDA funding envelop of equivalent US$15million, and then to accommodate the exchange losses between SDR -USD which had eroded the original IDA funding enveloped to US$13.3 million by 2016. Subsequent exchange rate gains on the undisbursed balance helped to raise the project funding to US$13.81 million. Total disbursement at project closing was 99.57 percent of the IDA financing. Page 58 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Table 4.13: Funding Sources Funding source (US$) Appraisal Estimate Revised (2016) Disbursement at end of project IDA Credit $15.0 million $13.3 million $13.81 million* Government of Liberia $1.1 million 0 Smallholders (inputs/labor) $6.2 million 0 Participating financial institutions $0.8 million 0 TOTAL $23.1 million $13.3 million $13.81 million Comparison of Actual and Estimated Component Costs: Available expenditure data at project closing shows that nearly US$8.0million (91.8 percent of revised estimate), US$2.4 million (115.2 percent), and US$3.0million (111.7.1 percent of appraisal estimate) were spent on components 1, 2 and 3 respectively (Table 4.14). The high variance in component 2 was due to relatively higher expenditure on capacity building, while the replacement of the entire project management team resulted in increases in project management cost. Table 4.14: Actual Component Cost vs. Estimated Component Cost as at 12/31/2018 Components Appraisal Revised Actual (US$) Variance Percent of Estimates US$ Million Million Actual US$ million Component 1: Tree Crop 16.90 8.68 8.00 -715,307.10 91.8% Revitalization Component 2: Institutional 3.25 2,.10 2.41 319,249.67 115.2% Building and Preparation of Future Large-Scale Tree Crop Component 3: Project 2.95 3.03 3.38 354,797.32 111.7% Coordination and Management Total project cost 23.10 13.81 13,76 -41,260.11 99.7 Source: PIU expenditure database Time overruns STCRSP experienced serious time overruns in its preparation compared to average, taking 40.1 months from concept to first disbursement compared with the Bank-wide average of 25.3 months. The project preparation (concept to approval) time was 28 months, almost double the Bank-wide average of 15 months. This was more than double the Africa average (13.0 months) and about twice the Liberia average (9.7 months). Its time from effectiveness to first disbursement (1.0 months) is however the lowest in Liberia, is so far. The project experienced one disbursement lag during its implementation period. Table 4.15 provides key project delivery times. Page 59 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Table 4.15: Project processing and implementation times Concept to Approval Approval to Effectiveness to First Concept to First Effectiveness Disbursement Disbursement STCRSP 28.0 11.4 1.0 40.1 Liberia 9.7 7.2 4.3 21.2 Africa 13.0 6.2 2.6 22.4 Bank 15.2 6.0 4.0 25.3 Source: operations monitoring dashboard Staff attrition posed significant challenges to project implementation. Virtually, the entire key project staff, except for the M&E officer, left by the third year of implementation. This resulted in a very steep learning curve for a new team that took over the management of the project for the remaining years of the project life span, including the extension period. Only the project coordinator was still at his post at the time of ICR. On the Bank side, there was an extended period of TTL replacement at the concept stage, resulting in a very long period between project concept and approval. Page 60 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) ANNEX 5. BORROWER, CO-FINANCIER AND OTHER PARTNER/STAKEHOLDER COMMENTS Institution Comments Ministry of The Minister of Agriculture indicated that they appreciated the support provided under Agriculture the project and were pleased with the tree crops targeted by the project, which he believes have strong economic potential for the country. He saw the project as a good start, as it served to instill hope in smallholder farmers that they can earn a livelihood from the sector. He indicated that the Ministry plans to use the Master Plan elaborated by the project to help elaborate relevant policies. At the same time, the Minister expressed concern regarding the sustainability of interventions started by the project. He felt that the farmers could have benefited from some additional support during a longer timeframe to enable them to reach a state of independence. He recalled the Government’s request to the Bank for a second phase and expressed hope that STAR-P could provide some support to some of the beneficiaries, particularly in the oil palm sub-sector. He also recommended that future projects aim to promote larger farm schemes beyond the 1 ha per farmer that the project ended up supporting. He also stressed the need for ensuring the right skills sets are hired for implementing projects, particularly personnel qualified in project management and contract management. He acknowledged that it is the responsibility of the Government of Liberia to recruit project staff but requested that the Bank provide increased oversight to help ensure that qualified staff are hired. Finally, he encouraged greater attention to social inclusion in future projects. Ministry of Finance The following statement was communicated by the Deputy Minister for Economic and Development Management: Planning “Having reviewed the report for the “Smallholder Tree Crop Revitalization Support Project” which had a Project Development Objective: “To increase access to finance, inputs, technologies and markets for smallholder tree crop farmers in Liberia, and to develop a long-term development program for the tree crops sector”; I can confirm that the project as was designed, was meant to do a lot of good for the Agriculture sector in the tree crops sub-sector. However, my review of the report confirmed my already perceived thoughts that funding to the Agriculture sector had not met its intended purpose either due to lack of proper implementation (waste of resources) or poor/limited tracking of data/ results during and after the intervention. Hence, I can clearly say that the project did not meet its objectives and results. Further, it occurs to me that reports presented by the implementing units are faulty and questionable as evidenced by the number of inconsistencies reported in the ICR. In future projects, there must be a stage by stage verification of results (tracking – all stakeholders must be involved) to ensure that financing acquired are used for the intended purpose and results derived from using these funds are really supporting the economic growth of the Country.” Page 61 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Institution Comments IFAD IFAD financed and supervised a Smallholder Tree Crops Revitalization and Support Project in Lofa County, which was running parallel with the World Bank-financed STCRSP and which ended in 2017. IFAD indicated that while the project confirmed that tree crops present large potential for improving living standards for the smallholder producers, an important lesson learned is the need to focus on the entire value chain to unlock the full development potential. IFAD is currently financing the Tree Crop Extension Project in Lofa and Nimba counties, which builds on lessons from their project. EU EU is also financing a tree crops project, focusing on cocoa, and informed that their project benefitted from some of the STCRSP interventions in counties of overlap. Despite their hesitance to become involved in rubber and oil palm, they felt that the World Bank project’s focus on these two tree crops plus cocoa was the right selection. They also commended the project’s focus on smallholders, which remain the backbone of Liberia’s agriculture sector. At the same time, they felt the project “spread itself too thin” focusing on too many activities and too many entry points. They suggested more of the project’s efforts should have focused on developing the value chain itself – in terms of improving logistics, access to markets, and advocate for solutions to gaps, which the project could not address in any substantive way, such as roads and other infrastructure. They recommended that setting up sub-sector technical working groups is key to ensuring coordination among the development partners in promoting tree crops and indicated that the EU was willing to align financing, through a Trust Fund, to ensure greater coordination. Page 62 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) ANNEX 6. TABLES REFERENCED IN THE REPORT A6.1: Revised Project Costs and Financing by Component Component Total Costs (US$’000) IDA Financing (US$”000) Original Revised Original Cost Revised Cost Allocation Allocation 1. Smallholder Tree Crops Revitalization 16,857 14,423 9,467 8,613 2. Institution Building and Preparation of Future 3,251 2,078 3,056 2,078 Large-Scale Tree Crop Development Program 3. Project Coordination and Management 2,951 3,001 2,477 3,001 TOTAL PROJECT COSTS 23,059 19,502 15,000 13,692 Source: Aide-Memoire, STCRSP Mid-Term Review Mission, November 2015 A6.2: Reallocation of Costs across Disbursement Categories Category Category Description Original Allocation Revised Allocation (XDR) (XDR) 1 Civil Works 280,614.00 415,380.50 2 Goods Except Under Cat. 4, 5, & 6 548,474.00 597,193.47 3 Con Ser & Tra except under Cat 4 & 5 5,971,229.00 5,754,702.62 4 Smallholder Sub Grant Under Part 1 1,546,214.00 1,630,141.93 5 Smallholder Sub Credit Under Part 1 189,911.00 80,495.47 6 Incremental Operating Costs 410,292.00 531,706.37 7 PPF Refinancing 753,266.00 690,379.64 8 Unallocated 0.00 0.00 DA-1 Designated Account 0.00 0.00 DA-A Designated Account 0.00 0.00 DA-B Designated Account 0.00 0.00 TOTAL 9,700,000.00 9,700,000.00 Source: Restructuring Paper, October 2018 Page 63 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) A6. 3: Summary of Indicators, End Targets and Actual, Pre and Post Mid-Term Restructuring Objective Indicator Baseline May 2013-May 2016 May 2016-Nov. 2018 Pre-revision Pre-revision Post revision Post revision end target actual end target actual Increased Area of smallholder tree crop farms 0 Total: 11,300 Total: 2318 Total: 6,056 Total: 5,829 access to rehabilitated, replanted or planted under ha ha ha ha inputs, the project, disaggregated by: technologies, (i) Rehabilitation (7100) (i) 0 (i) 7100 (i) 2013.88 (i) 4,506 (i) 4,195 finance (ii) Replanting (2200) (ii) 0 (ii) 2200 (ii) 300 (ii) 1,250 (ii) 1,004 (iii) New planting (2000) (iii) 0 (iii) 2000 (iii) 0 (iii) 300 (iii) 630 Increased Incremental yearly net cash flow in (US$/ha) Dropped Dropped access to project areas Disaggregated by: markets (i) High input rehabilitated cocoa (i) 120 (i) 760 (i) 120 (ii) Medium input rehabilitated cocoa (ii) 120 (ii) 180 (ii) 1829.43 (iii) Rehabilitated coffee (iii) 274 (iii) 200 (iii) 184.45 (iv) Rehabilitated oil palm (farmers’ run (iv) 643 (iv) 500 (iv) 0 plantation) (v) 643 (v) 700 (v) 0 (v) Rehabilitated oil palm (out-growers scheme) Increased Yield of major tree crops, disaggregated Metric NA NA access to by: tons/ha inputs, (i) High input cocoa (i) 0.96 (i) 1.30 (i) 1.00 technologies, (ii) Medium input cocoa (ii) 0.12 (ii) 0.50 (ii) 0.466 finance (iii) Rehabilitated coffee (iii) 0.12 (iii) 0.50 (iii) 0.30 (iv) Rehabilitated oil palm (iv) 1.16 (iv) 5.00 (iv) 1.50 Increased Long term credit delivered to oil palm 0 US$620,000 0 3 out-grower 1 access to and rubber out growers under the project schemes finance (revised to number of out-grower schemes with long-term credit to farmers and moved to intermediate level) Increased Farm access roads rehabilitated (target 0 145 (km) 0 30 5 access to revised; moved to intermediate level) markets Increased No of Smallholder tree crop farmers with --- NA NA 3,025 3,750 access to access to finance, inputs, markets, and finance, technologies inputs, markets, technologies Develop Long-term, large-scale tree crop 0 yes no yes yes long-term development program formulated and development approved by the Ministry of Agriculture program for the tree crops sector Direct project beneficiaries (of which 0 5,000/20% 1392/418 5,000/15% 3,850/750 x% are female) Source: STRCP PAD, 2012; Mid-Term Review Aide Memoire, 2015; Project Completion Report, 2019 Page 64 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) A6.4: Computation of Weighted Average Yield using Paasche Index Formula Rehabilitated Tree Crop Target Yield Achieved Achieved Area Weighing Yt Weighing Ya (Mt/Ha) Yield (%) (Ya/Yt) Achieved(Ha) (Aa*Yt) (Aa*Ya) (Yt) (Mt/Ha) (Ya) (Aa) Cocoa (High Input) 1.3 1.0 76.9 100.0 130.0 100.0 Cocoa (Medium Input) 0.5 0.5 93.2 3,772.2 1,886.1 1,757.8 Coffee 0.5 0.3 60.0 220.4 110.2 66.1 Oil Palm 5.0 1.0 20.0 102.0 510.0 102.0 SUM 4,194.6 2,636.3 2,025.9 1. PI= (sum(Aa x Ya) / sum(Aa x Yt) = 76.8 2. Weighted Average Target Yield (Mt/Ha) = Sum(Aa*Yt)/Sum(Aa) = 0.63 3. Weighted Average Achieved Yield (Mt/Ha) = Sum(Aa*Ya)/Sum(Aa) = 0.48 4. Achievement (%) = (2)/(1)*100 = 76.8 Source: Author’s calculation based on data from STCRSP ISR No. 13; Restructured Paper; and *Government Project Completion Report 2013-2018 Note: The Laspeyre Index (LI), which would require the target area as weight for both target and achieved yields, could not be calculated because of data inconsistencies in the disaggregation of the target area by crop (see Annex A6.6). The PI, however, would still have been preferred to the LI because the area achieved used in the PI was observed and probably more reliable compared to the target area (used in the LI), which was a guess-estimate. However, if the LI could have been calculated, then combining the PI and LI into the Fisher Index (FI) would have been best Page 65 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) A6.5 (a) Project Database Analysis Farmer’s # of Avg Avg Received # of Avg Area Gender Farme Age Family Farms Farm rs Size Tools Chemicals Fertilizers Credit Matching Rehab Size Rehab (ha) (ha) Grant COCOA Female 562 39.66 5.78 562 562 562 333 562 562 1.24 697.9461 Male 2412 40.13 5.82 2412 2412 2412 1319 2412 2412 1.33 3215.8713 TOTAL/ 2974 40.04 5.81 2974 2974 2974 1652 2974 3913.8174 AVERAGE 2974 1.32 Female 18.9 91.1 99.3 18.9 18.9 18.9 20.2 18.9 17.8 Share (%) 18.9 94.4 Female +/- -81.1 -0.9 -0.7 -81.1 -81.1 -81.1 -79.8 -81.1 82.2 average(%) -81.1 -5.6  Female farmers are about 19% of total farmers, in other words, about 81% below parity (gender equality)  Female cocoa farmers' average age and family size are about the same as their male counterparts, with both less than 1% below average  Average farm size of female cocoa farmers is only less than about 6% below average.  All farmers, female and male, is listed as having access to inputs and matching grants; female cocoa farmers access to credit is slightly better (20%) than their total share (19%)  A total of 3,914ha of cocoa farms are rehabilitated, of which 697.95 (17.8%) are for female farmers Note: 1. Area rehabilitated is (total farms rehabilitated x average farm size). There were no missing values on farm size for cocoa farmers 2. No cocoa farmers received credit COFFEE Female 6 36.33 5.50 6 6 6 0 6 6 6 5.43 Male 20 43.10 6.70 20 20 20 0 20 20 20 17.71 TOTAL/ AVERAGE 26 41.54 6.42 26 26 26 0 26 26 26 23.14 Female Share (%) 23.1 87.5 85.6 23.1 23.1 23.1 23.1 23.1 23.1 23.5 Female +/- average(%) -76.9 -12.5 -14.4 -76.9 -76.9 -76.9 -76.9 -76.9 -76.9 -76.5  Female farmers were 23% of total farmers, in other words, about 77% below parity  Female coffee farmers' average age and family size were about 13% and 14% respectively below total average  However, noticeably, average farm size of female farmers was about 2% above total average  Consequently, the share of area rehabilitated for female coffee farmers (23.5%) was relatively better than the share of the number of female farmers (23.1%)  All coffee farmers, female and male, received credit, but none received matching grants OIL PALM Female 9 45.11 7.44 9 9 9 0 0 9 1.14 10.24 Page 66 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Male 25 54.96 8.72 25 25 25 0 0 25 1.03 25.72 TOTAL/ AVERAGE 34 52.35 8.38 34 34 34 0 0 34 1.06 35.96 Female Share (%) 26.5 86.2 88.8 26.5 26.5% 26.5 26.5 107.6 28.5 Female +/- average(%) -73.5 -13.8 -11.2 -73.5 -73.5 -73.5 -73.5 7.6 -71.5  Female farmers are 26% of total farmers, in other words, about 74% below parity  Female oil palm farmers' average age and family size are about 14% and 11% respectively below total average  However, average farm size of oil palm female farmers is about 8% above total average  Consequently, the share of area rehabilitated for female coffee farmers (28.5%) is relatively better than the share of the number of farmers (about 26.5%).  No oil palm farmer, female or male, received either credit or matching grant RUBBER Female 121 41.09 6.07 121 121 121 0 50 1.63* 112 Male 363 47.04 7.01 363 363 363 0 147 1.72* 336 TOTAL/ AVERAGE 484 45.55 6.78 484 484 484 0 197 1.69* 448 Female Share (%) 25.0 90.2 89.5 25.0 25.0 25.0% 25.4 96.0* 25.0 Female +/- average(%) -75.0 -9.8 -10.5 -75.0 -75.0 -75.0% -74.6 -4.0* -75.0  Female farmers are 25% of total farmers, in other words, 75% below parity  Female rubber farmers' average age and family size are both about 10% below total average  Average farm size of rubber female farmers is about 4% above the total average, however, missing farm size data have introduced biases in averages Note 1. * Biased because calculated over missing data. It is not possible to calculate total area of farm rehabilitated because of missing data 2. No farmers were listed as having received credit. However, other documents show that the only farmers to have received the long-term credit were 150 rubber farmers from Montserrado country Source: STCRSP PIU database. Downloaded April 17, 2019 Page 67 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) A6.5 (b) Project Database Analysis # of Farmers Cocoa Coffee Oil Palm Rubber Total Received Inputs Female 562 6 9 121 698 Male 2,412 20 25 363 2,820 Total 2,974 26 34 484 3,518 Note. Above target for access to input but less than reported Received Credit Female 333 0 0 0 333 Male 1319 0 0 0 1,319 Total 1,652 0 0 0 1,652* Note. Below target for access to finance. Also, the number is inflated, most likely due to data entry. Only 150 farmers received access to credit Received Grants Female 562 0 0 50 612 Male 2,412 20 0 147 2,579 Total 2,974 26 0 197 3,197* Note. Above target for access to finance but less than reported Farms Rehabilitated Female 2,974 6 112 112 3,204 Male 562 20 336 336 1,254 Total 2,412 26 448 448 3,334 Number Percent Female Farmers 698 19.8 Male Farmers 2,820 80.2 TOTAL 3,518 100.0 Note  Total number of farmers are above target but less than reported  Number of female farmers above target but less than reported Source: STCRSP PIU database. Downloaded April 17, 2019 Page 68 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) A6.6 (a) Note on Data Inconsistency Data challenges: gaps and reliability The sources of information for the assessment include: Bank staff’s reports (aide-memoires, ISRs, restructuring papers, etc.); project documents (particularly its database); the government’s project completion report and impact assessment report, the ICR team’s observations, and pertinent anecdotal evidence (as recommended by the ICR guidelines). a. Data gaps: The ICR highlights several areas of data gaps due to the limited types of data collected. Additionally, even with indicators for which data was collected (such as size of farms, access to inputs), there are several fields in the project database that are noted not available, or empty, making calculation to verify what is reported in the final RF a challenge. b. Data reliability: There are several instances where information in the project database vary from what is reported in the final ISR and the PIU’s project completion report, notably the total number of project beneficiaries (including the number of women); the number of farmers with access to matching grants and credit (see A6.5 (a&b)), the number of farms rehabilitated. Additionally, there are data inconsistencies across project documents. The Table below (A6.6(b)) provides some examples. Finally, it was reported that coffee farmers abandoned their farms during the first year of the project implementation and project interventions in this sub-sector stopped. No data on yields were reported throughout the project, yet a yield increase of 0.3 metric tons/ha against a target of 0.5 was reported at the end of the project A6.6 (b) Table on Data Inconsistency Indicator Original Revised Target as per: Achieved, as Comment Target* per: MTR- AM 4th PCR MTR- ISR PCR AM ISM RP Number of Direct Beneficiaries Cocoa/Coffee 3,250 2,500 3,000 2,500 3,000  *Both the Mid-Term Review Oil Palm 600 75** 400 400 269 (MTR) Aide Memoire (AM) and Rubber 1,300 450 400 350 581 the AM for the 4th Total 5,150 3,025 3,800*** 3,250 3,100 3,850 3,850 Implementation Supervision Mission (ISM) listed the original targets as reflected here. The Results Framework at appraisal and the ISRs listed the total at 5,000 (there was no breakdown per crop in these two documents by crops)  **Does not include all oil palm farmers (MTR-AM, p.17)  ***The RF indicates the total as 4,000  The project database provides a list of 3,518 farmers as total project beneficiaries. Both the PCR and ISR reported the number of female beneficiaries to be 900, the project database listed 698 females. Area Targeted (Ha) for Rehabilitated Tree Crops Page 69 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) Indicator Original Revised Target as per: Achieved, as Comment Target* per: MTR- AM 4th PCR MTR- ISR PCR AM ISM RP Cocoa (high 3,000 1,200 1,200 100 No  Focus is on rehabilitated tree input) data crops as this was the model Cocoa 2,000 2,485 3,000 3,000 3,700 that targeted yields. (medium  The most significant input) discrepancy is the revised Coffee 1,500 400 400 371 targeted area for high input Oil Palm 300 300 200 200 300 cocoa. While the disaggregated (farmer run) data reflects the revised target Oil palm (out- 300 150 200 200 No to be 1,200 in the Aide grower/ data Memoires for both the Mid- concessionaire) Term Review and the 4th ISM, Aggregated 7,100 4,506+ No data 3,900 4,506+ the PIU reports on 100 ha as data reflected the revised target. Although in in the Results earlier progress reports, the Framework PIU reported on 1,200 ha as the revised target. Similarly, the impact evaluation, which cited the PIU’s progress report used the 1,200 ha as the revised target72.  + The disaggregated data does not add up to the aggregated data reflected in the revised Results Framework.  Given the Restructuring Paper is a legal document, the ICR used the aggregated data reported in this document in calculating yield performance and for the total number of direct beneficiaries Note: 1. AM = Aide Memoire 2. ISM = Implementation Supervision Mission 3. MTR = Mid-Term Review 4. PCR = Project Completion Report (completed by the Project Implementation Unit) 5. RP = Restructuring Paper 72 Impact Evaluation Report, p. 9 Page 70 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) A6.7: Overview of Revised PDO Indicators Original Revisions Rationale 1 Area of smallholder tree crop farms Targets revised to Total (6,056 ha), disaggregated To reflect decrease in project rehabilitated, replanted or planted under by: funding the project (11,300 ha), disaggregated by:  Rehabilitation (7100)  Rehabilitation (4,506)  Replanting (2200)  Replanting (1,250)  New planting (2000)  New planting (300) 2 Incremental yearly net cash flow in project Indicator dropped, replaced by indicator Indicator dropped due to areas (US$/ha). Disaggregated by: originally at the intermediate level: Yield of potential attribution challenges  High input rehabilitated cocoa major tree crops (metric tons/ha), disaggregated related to using farm gate price  Medium input rehabilitated cocoa by: in calculation of net cash flow.  Rehabilitated coffee  High input cocoa (1.30) This price is determined by  Rehabilitated oil palm (farmers’ run  Medium input cocoa (0.50) demand-supply dynamics plantation)  Rehabilitated coffee (0.50) outside the project  Rehabilitated oil palm (out-growers  Rehabilitated oil palm (5.00) interventions; The intermediate scheme) indicators on yields were moved because yield was considered as a medium-term outcome and therefore more appropriate at the PDO level 3 Long term credit delivered to oil palm and Indicator revised and moved to intermediate To better reflect project rubber out growers under the project level: interventions (to establish out- (US$’000) Out-grower schemes with long-term credit to grower schemes with credit participating smallholder farmers supported by component) rather than setting the project an absolute target for credit delivered 4 Farm access roads rehabilitated (145km) Target revised and moved to intermediate level Target moved to intermediate 30km level because it is an output 5 Direct project beneficiaries (of which are Target revised female) (5,000/20%) 3,100 (15%) 6 Long-term, large-scale tree crop No change development program formulated and approved by the Ministry of Agriculture (Yes/No) 7 NA New indicator introduced: Alignment with PDO statement Smallholder tree crop farmers with access to finance, input, markets, and technologies (3,025) Source: Aide-Memoire, STCRSP Mid-Term Review Mission, November 2015; Restructuring Paper on a Proposed Project Restructuring of Smallholder Tree Crop Revitalization Support Project, April 6, 2016 Page 71 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) A6.8: Cocoa Cooperatives Trained No. County Organization Number of registered Targeted number of members at the time members participated in training 1 BONG Kwapaigeh Farmers Asso. 132 100 2 Kokoya Farmers Coop 407 100 3 Panta Farmers Coop 332 100 4 NIMBA Kwakerseh Farmers Coop 419 100 5 Sroh Kwado Farmers Coop 106 100 6 Zodoe Farmers Coop 66 100 7 Waperley Farmers Coop 210 100 8 Zoyah Farmers Coop 75 100 9 Gbeh Farmers Coop 56 80 10 Boe Kparblee Farmers Asso. 106 100 11 Gbao Kwadoe Farmers Asso. 107 100 12 Kwa-pateh Farmers Asso. 102 100 13 GRAND. GEDEH Pooh-Soun Farmers Asso. 256 100 14 Niao Farmers Asso. 106 100 TOTAL 2,480 1,380 Source: SOCODEVI- implementing partner Note: Funds were allocated to accommodate up to 100 trainees Page 72 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) ANNEX 7. BIBLIOGRAPHY 1. EGIS, Smallholder Tree Crops Revitalization Support Project, Mid-Term Impact Study. Draft Report, June 2016. 2. IFAD; Smallholder Tree Crops Revitalization Support Project (STCRSP); Project No: 1100001616; Report No:4728-LR; October 22, 2018. 3. LTS International; Final Report, Project Impact Evaluation: Smallholder Tree Crops Revitalization Support Project (STCRSP); November 30, 2018 4. Ministry of Agriculture, Project Coordination Unit; Smallholder Tree Crops Revitalization Support Project (STCRSP), Annual Progress Report, July 1, 2016-June 30, 2017 5. Ministry of Agriculture, Project Coordination Unit, Smallholder Tree Crops Revitalization Support Project (STCRSP), Dube Oil Palm Plantation Business Support Assessment Support; Ref LR-PMU-MOA-34097-CS- CDS; 13 June 13, 2018 6. Ministry of Agriculture, Project Coordination Unit, Smallholder Tree Crops Revitalization Support Project (STCRSP), Performance Assessment of Workable Models Report; Ref LR-PMU-MOA-34097-CS-CDS, August 6, 2018 7. Ministry of Agriculture, Project Coordination Unit, Smallholder Tree Crops Revitalization Support Project (STCRSP), Konobo District Farmers Multi-purpose Cooperative Society Record Keeping System Assessment Report; Ref LR-PMU-MOA-34097-CS-CDS, June 2, 2018 8. Ministry of Agriculture, Project Coordination Unit, Smallholder Tree Crops Revitalization Support Project (STCRSP), Project database 9. Ministry of Agriculture, Project Coordination Unit, Smallholder Tree Crops Revitalization Support Project (STCRSP), Project Completion Report, November 2018 10. SOCODEVI; STCRSP Mission Report: Training on Changes 1&2 (Membership Growth and Good Governance; July 19-August 14, 2017 11. SOCODEVI; STCRSP Mission Report: Training of Trainers (ToT) and FOs Elected Officers and Managers Training; June 17-30, 2017 12. SOCODEVI, Smallholder Tree Crops Revitalization Support Project (STCRSP)Cocoa-Coffee Sub- Component, Lessons Learnt, October 2017 13. World Bank; Smallholder Tree Crops Revitalization Support Project (STCRSP), Project Appraisal Document, Report No: 68524-LR Page 73 of 74 The World Bank LR Smallholder Tree Crop Revitalization Support Project (P113273) 14. World Bank; Smallholder Tree Crops Revitalization Support Project (STCRSP), Mid-Term Review Mission; Aide Memoire, November 2-13, 2015 15. World Bank; Smallholder Tree Crops Revitalization Support Project (STCRSP), Fourth Implementation Support Mission; Aide Memoire, June 2017 16. World Bank; Restructuring Paper on a Proposed Restructuring of Smallholder Tree Crops Revitalization Support Project P113273 Board Approval Date June 5, 2012 to the Government of Liberia; June 4, 2014; Report No:88076-LR 17. World Bank; Restructuring Paper on a Proposed Restructuring of Smallholder Tree Crops Revitalization Support Project P113273 Board Approval Date June 5, 2012 to the Government of Liberia; April 6, 2016; Report No:RES22877 18. World Bank; Restructuring Paper on a Proposed Restructuring of LR Smallholder Tree Crops Revitalization Support Project P113273 Board Approval Date June 5, 2012 to the Ministry of Finance and Development Planning; October 10, 2018; Report No:RES34169 Page 74 of 74