Document of The World Bank Report No: ICR00003335 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-43660) ON A CREDIT IN THE AMOUNT OF SDR 9.8 MILLION (US$ 15 MILLION EQUIVALENT) TO THE PLURINATIONAL STATE OF BOLIVIA FOR A LAND FOR AGRICULTURAL DEVELOPMENT PROJECT March 12, 2015 Agriculture Global Practice Bolivia, Chile, Ecuador, Peru, Venezuela Country Management Unit Latin America and the Caribbean Region CURRENCY EQUIVALENTS (Exchange Rate Effective March 12, 2015) Currency Unit = Boliviano (BOB) 1.00 BOB = US$ 0.14 1.00 SDR = US$ 1.417330 FISCAL YEAR January 1 – December 31 ABBREVIATIONS AND ACRONYMS BDP Banco de Desarrollo Productivo (Productive Development Bank) BP Bank Procedure CODAL Consejo de Desarrollo Agrario Local (Local Rural Development Council) CPS Country Partnership Strategy DCA Development Credit Agreement EIRR Economic Internal Rate of Return EMP Environmental Management Plan ENPV Economic Net Present Value FAO-TCI United Nations Food and Agriculture Organization Cooperation Program FIRR Financial Internal Rate of Return FNVP Financial Net Present Value GOB Government of Bolivia IFI Intermediary Financial Institution INRA Instituto Nacional de la Reforma Agraria (National Land Reform Institute) IP Implementation Plan IRR Internal Rate of Return MAS Movement for Socialism MDRAMA Ministerio de Desarrollo Rural, Agricultura y Medio Ambiente (Ministry of Rural Development, Agriculture and Environment) M&E Monitoring and Evaluation MDRyT Ministerio de Desarrollo Rural y Tierras (Ministry of Rural Development and Lands) MTR Mid Term Review NDP National Development Plan NIT National Tax Identification Number NPV Net Present Value OP Operational Policy PA Producers Association PCU Project Coordinating Unit PDO Project Development Objective PPA Project Preparation Advance TA Technical Assistance TC Tierra Comunitaria (Communal Land) TCO Tierra Comunitario de Origen (Indigenous Peoples’ Territory) Regional Vice President: Jorge Familiar Country Director: Alberto Rodriguez Sector Manager: Laurent Msellati Project Team Leader: Enrique Pantoja ICR Team Leader: Enrique Pantoja BOLIVIA Land for Agricultural Development Project CONTENTS Data Sheet A. Basic Information B. Key Dates C. Ratings Summary D. Sector and Theme Codes E. Bank Staff F. Results Framework Analysis G. Ratings of Project Performance in ISRs H. Restructuring I. Disbursement Graph 1. Project Context, Development Objectives and Design ............................................... 1 2. Key Factors Affecting Implementation and Outcomes .............................................. 6 3. Assessment of Outcomes .......................................................................................... 13 4. Assessment of Risk to Development Outcome......................................................... 21 5. Assessment of Bank and Borrower Performance ..................................................... 21 6. Lessons Learned ....................................................................................................... 23 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners .......... 25 Annex 1. Project Costs and Financing .......................................................................... 26 Annex 2. Outputs by Component ................................................................................. 27 Annex 3. Economic and Financial Analysis ................................................................. 37 Annex 4. Bank Lending and Implementation Support/Supervision Processes ............ 44 Annex 5. Beneficiary Survey Results ........................................................................... 46 Annex 6. Stakeholder Workshop Report and Results................................................... 49 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR ..................... 50 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders ....................... 51 Annex 9. List of Supporting Documents ...................................................................... 52 MAP IBRD 41480 A. Basic Information Bolivia Land for Country: Bolivia Project Name: Agricultural Development Project. Project ID: P087925 L/C/TF Number(s): IDA-43660 ICR Date: 03/12/2015 ICR Type: Core ICR PLURINATIONAL Lending Instrument: SIL Borrower: STATE OF BOLIVIA Original Total XDR 9.80M Disbursed Amount: XDR 9.56M Commitment: Revised Amount: XDR 9.80M Environmental Category: B Implementing Agencies: Ministry of Rural Development and Lands Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 03/08/2004 Effectiveness: 03/13/2009 03/13/2009 Appraisal: 04/11/2005 Restructuring(s): 04/06/2010 Approval: 10/30/2007 Mid-term Review: 10/10/2011 02/06/2012 Closing: 06/29/2012 09/15/2014 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Moderately Satisfactory Risk to Development Outcome: Substantial Bank Performance: Moderately Satisfactory Borrower Performance: Moderately Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Moderately Satisfactory Government: Moderately Satisfactory Implementing Quality of Supervision: Moderately Satisfactory Moderately Satisfactory Agency/Agencies: Overall Bank Overall Borrower Moderately Satisfactory Moderately Satisfactory Performance: Performance: C.3 Quality at Entry and Implementation Performance Indicators QAG Assessments (if Implementation Performance Indicators Rating any) Potential Problem Project at any time: Yes Quality at Entry: None Quality of Problem Project at any time: Yes None Supervision: DO rating before Closing: Mod. Satisfactory D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) General agriculture, fishing and forestry sector 80 85 Sub-national government administration 20 15 Theme Code (as % of total Bank financing) Indigenous peoples 14 50 Land administration and management 29 15 Rural markets 29 35 Social safety nets 28 0 E. Bank Staff Positions At ICR At Approval Vice President: Jorge Familiar Pamela Cox Country Director: Alberto Rodriguez Carlos Felipe Jaramillo Practice Manager: Laurent Msellati Ethel Sennhauser Project Team Leader: Enrique Pantoja Malcolm Childress ICR Team Leader: Enrique Pantoja ICR Primary Author: Fernando Galeana F. Results Framework Analysis Project Development Objectives (from Development Credit Agreement) The PDOs are (a) to establish an alternative decentralized beneficiary-driven land distribution mechanism for organized landless or poor farmers; and (b) to support productive investments thereon to allow increased and more sustainable livelihoods for such beneficiaries.1 1 The wording of the PDO in the PAD is slightly different from the DCA’s version. The PDO in the PAD is “to establish a decentralized beneficiary-driven land distribution mechanism that allows organized landless or poor farmers to acquire suitable agricultural lands and implement investment subprojects which puts them in a sustainable higher-income, livelihood.” The PDO from the DCA was ratified in the first project restructuring. Revised Project Development Objectives (as approved by original approving authority) Not revised. (a) PDO Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised Target approval Completion or Values documents) Target Years Indicator 1: Direct project beneficiaries (new indicator added at restructuring) Value (Quantitative or 0 N/A 11,000 11,488 Qualitative) Date achieved 07-Dec-2007 24-Jan-2013 15-Sep-2014 Comments Target surpassed (104%). Estimate based on original assumption of 10 (Incl. % of households (HHs) per productive association (PA), and an average of 5 people achievement) per HH. Indicator 2: At least 20% increase in beneficiary family income Value (Quantitative or 0 50% 20% 39% Qualitative) Date achieved 07-Dec-2007 07-Dec-2007 24-Jan-2013 15-Sep-2014 Comments Target surpassed (195%). Estimate based on a sample of 83 PAs. (Incl. % of achievement) Local level decision-making and monitoring entities established and functioning Indicator 3: in 5 selected municipalities Value (Quantitative or 0 3 5 5 Qualitative) Date achieved 07-Dec-2007 07-Dec-2007 24-Jan-2013 15-Sep-2014 Comments Target achieved (100%). The number of Local Development Committees (Incl. % of (CODALs) was increased to reflect expansion of the Project area into two achievement) additional municipalities. At least 15% increase in value of productive assets of beneficiary families(new Indicator 4: indicator added at restructuring) 15% for agricultural Value investments and (Quantitative or 0 N/A 15% 10% for livestock Qualitative) investments Date achieved 24-Jan-2013 24-Jan-2013 15-Sep-2014 Target achieved (100%) for agricultural investment and partially achieved Comments (67%) for livestock investments. Estimates based on a sample of 83 PAs. For (Incl. % of agricultural investments, family assets increased by US$122 or a 15 percent achievement) increase compared to baseline. For livestock investments, family assets increased by US$ 156 or 10 percent increase compared to baseline. At least 10% of Productive Associations (PAs) have women in their Boards (new Indicator 5: indicator added at restructuring) Value (Quantitative or 0 N/A 10% 74% Qualitative) Date achieved 24-Jan-2013 24-Jan-2013 15-Sep-2014 Comments Target surpassed (743%). (Incl. % of achievement) Indicator 6: Female beneficiaries (new indicator added at restructuring) Value (Quantitative or 0 N/A 5,610 5,681 Qualitative) Date achieved 24-Jan-2013 24-Jan-2013 15-Sep-2014 Comments Target achieved (101%). (Incl. % of achievement) Monitoring system to assess the net benefits of increased ag. Production Indicator established (dropped and replaced at restructuring by intermediate outcome indicator 8) Value (Quantitative or 0 N/A N/A N/A Qualitative) Date achieved 07-Dec-2007 Comments (Incl. % of achievement) 120 farms selected and acquired by PAs via new mechanism (dropped at Indicator restructuring) Value (Quantitative or 0 N/A N/A N/A Qualitative) Date achieved 07-Dec-2007 Comments (Incl. % of achievement) (b) Intermediate Outcome Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised Target approval Completion or Values documents) Target Years Indicator 1: At least 2,200 beneficiary families select and acquire their farms Value (Quantitative or 0 2,200 2,891 Qualitative) Date achieved 07-Dec-2007 07-Dec-2007 15-Sep-2014 Comments Target surpassed (131%). These families are organized under PAs. (Incl. % of achievement) Indicator 2: At least 190 Productive Associations (PAs) established under the Project Value (Quantitative or 0 120 190 237 Qualitative) Date achieved 07-Dec-2007 07-Dec-2007 24-Jan-2013 15-Sep-2014 Comments Target surpassed (125%). (Incl. % of achievement) Indicator 3: At least 115,000 hectares of land are accessed by Productive Associations (PAs) Value (Quantitative or 0 100,000 115,000 151,579 Qualitative) Date achieved 07-Dec-2007 07-Dec-2007 24-Jan-2013 15-Sep-2014 Comments Target surpassed (132%). (Incl. % of achievement) At least 70% of the productive sub-projects comply satisfactorily with agreed Indicator 4: implementation plans Value (Quantitative or 0% 70% 85% Qualitative) Date achieved 24-Jan-2013 24-Jan-2013 15-Sep-2014 Comments Target surpassed (121%). (Incl. % of achievement) Indicator 5: At least 70% of land accessed by PAs are put under productive use Value (Quantitative or 0% 70% 59% Qualitative) Date achieved 24-Jan-2013 24-Jan-2013 15-Sep-2014 Comments Target partially achieved (84%). Indicator is based on average area put under (Incl. % of productive use by year of project implementation. achievement) At least 80% of PAs are satisfied with the technical assistance provided by the Indicator 6: Project Value (Quantitative or 0% 80% 73% Qualitative) Date achieved 24-Jan-2013 24-Jan-2013 15-Sep-2014 Comments Target partially achieved (91%). Estimate based on a survey of 58 (Incl. % of respondents. achievement) Indicator 7: At least 20% of participants in training sessions are women Value (Quantitative or 0% 20% 38% Qualitative) Date achieved 24-Jan-2013 24-Jan-2013 15-Sep-2014 Comments Target surpassed (195%). (Incl. % of achievement) Project M&E system established and updated periodically, and Mid-term and Indicator 8: final evaluation completed Value (Quantitative or 0 3 3 Qualitative) Date achieved 24-Jan-2013 24-Jan-2013 15-Sep-2014 Comments Target achieved (100%). (Incl. % of achievement) At least US$8.0 million in total value of productive investments (Dropped at Indicator restructuring) Value (Quantitative or 0 N/A N/A N/A Qualitative) Date achieved 07-Dec-2007 Comments (Incl. % of achievement) At least 90% of productive investment subprojects are completed (dropped and Indicator replaced at restructuring by intermediate outcome indicator 4) Value (Quantitative or 0 N/A N/A N/A Qualitative) Date achieved 07-Dec-2007 Comments (Incl. % of achievement) G. Ratings of Project Performance in ISRs Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 12/27/2007 Satisfactory Satisfactory 0.00 2 04/30/2008 Satisfactory Satisfactory 0.00 Moderately Moderately 3 08/13/2008 0.00 Unsatisfactory Unsatisfactory Moderately Moderately 4 12/14/2008 0.00 Unsatisfactory Unsatisfactory 5 04/17/2009 Moderately Satisfactory Moderately Satisfactory 0.00 Moderately Moderately 6 11/17/2009 2.83 Unsatisfactory Unsatisfactory Moderately Moderately 7 02/06/2010 3.09 Unsatisfactory Unsatisfactory 8 05/28/2010 Moderately Satisfactory Moderately Satisfactory 3.60 Moderately Moderately 9 02/13/2011 3.90 Unsatisfactory Unsatisfactory 10 06/29/2011 Moderately Satisfactory Moderately Satisfactory 4.86 11 12/21/2011 Moderately Satisfactory Moderately Satisfactory 5.91 12 06/29/2012 Moderately Satisfactory Moderately Satisfactory 7.46 13 12/14/2012 Moderately Satisfactory Satisfactory 9.41 14 06/22/2013 Moderately Satisfactory Satisfactory 10.73 15 10/21/2013 Moderately Satisfactory Satisfactory 11.57 16 03/29/2014 Moderately Satisfactory Moderately Satisfactory 12.85 17 09/27/2014 Moderately Satisfactory Moderately Satisfactory 14.50 H. Restructuring (if any) ISR Ratings at Amount Board Restructuring Disbursed at Restructuring Reason for Restructuring & Approved Restructuring Date(s) Key Changes Made PDO Change DO IP in USD millions Changes include: (1) modification of Land Credit Facility; (2) inclusion of three 04/06/2010 No MS MS 3.09 additional mechanisms for land access; and (3) a minor reallocation of Credit proceeds. Extension of the closing date 06/14/2012 No MS MS 7.46 from June 29, 2012 to March 31, 2014. Key changes include: (1) reformulation of the Results Matrix; (b) establishment of a new disbursement category to make explicit the expenditures for consultant services, training and operating costs for 01/24/2013 No MS S 9.41 Component 2 (productive investments); and (3) reallocation of Credit proceeds to reflect the evolution of the Project, which resulted in an increased emphasis on the land use rights mechanism and the need of additional field staff for ISR Ratings at Amount Board Restructuring Disbursed at Restructuring Reason for Restructuring & Approved Restructuring Date(s) Key Changes Made PDO Change DO IP in USD millions implementation, technical assistance and for the communication strategy. Extension of the closing date 11/29/2013 No MS S 11.57 from March 31 to June 15, 2014. Extension of the closing date 05/23/2014 No MS MS 12.85 from June 15 to September 15, 2014. I. Disbursement Profile The Bolivia Land for Agricultural Development Project (‘the Project’) was conceived as a pilot effort focused on selected municipalities of the Department of Santa Cruz, while being the last land access project financed by the Bank in Latin America. The Project allowed successful implementation of an innovative mechanism to gain access to land under customary land tenure arrangements that has the potential for countrywide replication. It enabled poor farmers, organized in Productive Associations (PAs), to benefit from the right to land that they already had as members of Indigenous Peoples’ Territories (TCOs) and Communal Lands (TCs) but could not enjoy due to a lack of productive assets. By providing matching grants and technical assistance (TA) to PAs to carry out investment sub-projects, the Project contributed a proven mechanism to consolidate land rights and gain access to agricultural land. Project preparation and implementation occurred within a period of significant political and social transformation for Bolivia. During this crucial historical moment, the Project generated valuable lessons about different options for improving land access. The key lesson was to think beyond the original emphasis on the ‘willing seller-willing buyer’ model to incorporate usufruct rights in TCOs and TCs. As a result, and with increased Government commitment, the Project exceeded most of its key performance indicators. Implementation delays, however, tightened the schedule for the establishment of PAs and execution of their investment sub-projects. Additional TA would have likely improved the sustainability of the new investments. By Project closing, there were no specific arrangements to continue the mechanism in TCOs and TCs. The Project generated other valuable lessons about working with indigenous peoples, monitoring and evaluation (M&E), and piloting land access mechanisms. 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal 1. Context and Sector Background. Lack of access to agricultural land is a long- standing historical problem in Bolivia with negative consequences for rural poverty and development. Although addressed through expropriation and colonization programs since the 1953 Agrarian Reform Act, landlessness continued to increase during the 1960s-1990s. At the time of appraisal, the 1996 agrarian reform law (Law No. 1715) relied on the allocation of public land, either for free or for a fee, and through planned settlements on lands identified during the land title regularization process. These traditional mechanisms were considered bureaucratic, slow, and unable to satisfy the demand for land. Between 1996 and 2005, agrarian reform moved at a slow pace with only 9.3 out of 109 million hectares of available land titled. Meanwhile, the concentration of underutilized private lands was on the rise in the eastern lowlands, particularly in the Department of Santa Cruz. 2. Country Context. During the 2000s, Bolivia experienced a period of political instability and regional tensions that made project preparation, and its initial implementation, challenging. The Project was formally appraised in April 2005 and negotiated the following May. However, political events in the country affected the course of the Project. President Carlos Mesa formally resigned in June 2005 and Eduardo Rodriguez Veltze assumed office as interim president. Supported by a broad coalition of indigenous and peasant movements, Evo Morales’ Movement for Socialism (MAS) won the elections in December. President Morales assumed office in January 2006 and launched 1 a reform that included major modifications to the Constitution, a stronger role of the state in the economy, and the launching of various social programs. During this period, regional tensions between the eastern and western regions heightened. An autonomy referendum for the Department of Santa Cruz was held in May 2008, but it was declared unconstitutional by the National Electoral Court. As per Bolivian law, at the time of appraisal, it was established under the Project that a credit line for land acquisition would be provided through the Productive Development Bank (BDP), a second-tier bank. Under these arrangements, BDP would act as trustee for credit funds that would then be on-lent to farmers’ groups via commercial intermediary financial institutions (IFIs). Bolivian law changed in 2008 making BDP’s role redundant by allowing the provision of credit directly through first-tier commercial IFIs. 3. Government Strategy. The administration of President Morales made agrarian reform one of the central policy issues. Through the newly created Ministry of Rural Development and Lands (MDRyT), the Government of Bolivia (GOB) prepared a new policy framework aimed at improving equity in the distribution and access to agricultural land. A new law (Law No. 3545, November 2006) targeted indigenous and peasant communities as the sole beneficiaries of agrarian reform. The policy framework promoted access to land through the establishment of new settlements in public lands and the consolidation of the rights of TCOs and TCs. The Project supported the implementation of this policy framework by piloting a land access mechanism in TCOs and TCs. This mechanism was initially identified during the implementation of the Project Preparation Advance (PPA). The PPA Agreement was originally signed in April 2005 and amended in May 2008 in order to pilot another three additional mechanisms for land access. 4. Rationale for Bank Involvement. The Bank was uniquely positioned to support Bolivia in launching a participatory, decentralized complementary mechanism for land access and productive investments by poor farmers. During the 1990s, the Bank supported land policy reform in Bolivia with technical assistance and policy dialogue and the implementation of the 1996 agrarian reform law through the Land Administration Project.2 In addition to its work in Bolivia, the Bank had significant experience supporting innovative land access mechanisms in other countries in the region, including Brazil, Mexico, Guatemala, and Honduras. The Project supported the Interim Strategy Note goal of fostering jobs through growth in rural areas. 1.2 Original Project Development Objectives (PDO) and Key Indicators 5. As stated in the Development Credit Agreement (DCA), the Project Development Objectives (PDOs) are (a) to establish an alternative decentralized beneficiary-driven land distribution mechanism for organized landless or poor farmers; and (b) to support 2 The National Land Administration Project (P006197) was approved by the Board in June 1995, became effective in September 1995, and closed on June 2005 after several extensions and additional financing. The total cost of the project was US$ 20.4 million. 2 productive investments thereon to allow increased and more sustainable livelihoods for such beneficiaries.3 6. The original key indicators were: o Establishment of a decentralized, beneficiary-driven land access mechanism; o 50 percent increase in family incomes of participating families; o 120 farms selected and acquired by productive associations (PA) via new mechanisms; o Local level decision-making and monitoring entities established and functioning in 3 selected municipalities; o Monitoring system to assess the net economic benefits of increased agricultural production established. 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification 7. The original PDO was maintained, but the Results Matrix was reformulated in a Level 2 Restructuring in January 2013 based on the assessment of the Mid-Term Review (MTR) and the addition of two new municipalities. These adjustments reflect both an adaptive restructuring to maintain consistency with the Project’s increased focus on TCOs and TCs as well as methodological improvements for measuring outcomes. The indicators for farm acquisition and establishment of an economic monitoring system were dropped. Also, the number of local decision-making and monitoring entities was increased from 3 to 5 and the number of PAs was increased from 120 to 190, which corresponded to the expansion of the Project area to the municipalities of Concepcion and San Ignacio de Velasco. The target of increase in family income was reduced from 50 to 20 percent and a new indicator was added to measure the change on the value of productive assets with a target of at least 15 percent increase. New indicators were added to measure gender- differentiated impacts: the percent of PAs with women in their boards, with a target of at least 10 percent, and number of female beneficiaries (core indicator). A table describing all the changes to the Results Framework is included in Annex 2. 1.4 Main Beneficiaries, 8. The primary target group was 2,200 families of organized landless or poor farmers that would gain access to land through a decentralized mechanism. Located in five municipalities with precarious access to basic services in the Department of Santa Cruz, most of these families were considered to be poor or very poor and approximately 75 percent indigenous. These farmers worked as permanent or seasonal workers in agro- industry, produced in sharecropping modalities, or owned parcels too small to sustain a livelihood. Targeting of beneficiaries also depended on land market conditions that would facilitate access to underutilized agricultural land. 3 The wording of the PDO included in the Project Appraisal Document (PAD) is slightly different from the version included in the DCA. The PDO in the PAD is “to establish a decentralized beneficiary-driven land distribution mechanism that allows organized landless or poor farmers to acquire suitable agricultural lands and implement investment subprojects which puts them in a sustainable higher-income, livelihood.” The PDO from the DCA was ratified in a 2010 Restructuring Paper. 3 1.5 Original Components 9. The Project had three components: 10. Component 1 - Land Acquisition (US$4.86 million of Credit funds at appraisal): This component was to provide loans to PAs for the purchasing of agricultural land. The Project intended to establish a Land Credit Facility managed by BDP and executed by local IFIs. The acquired land would serve as collateral for the loan. Only titled, private land was eligible for purchase. The land acquisition mechanism was not operationalized due to the lack of eligible land at affordable prices, and the Land Credit Facility could not be established. 11. Component 2 - Productive Investments (US$7.86 million of Credit funds at appraisal): This component provided matching grants and TA to PAs for carrying out investment sub-projects. The matching grants covered investments in infrastructure and productive assets. The Project Coordination Unit (PCU) launched a communication campaign to inform potential beneficiaries about the Project, and helped organize them in PAs. The PA had to prepare a preliminary investment plan. The PCU reviewed the proposals to ensure basic compliance with the Project’s guidelines and procedures. The PCU then submitted the proposal to a local level decision-making body, the Council for Local Agrarian Development (CODALs), for approval. The CODALs were established by the Project in each of the participating municipalities and were composed of members from the municipality, traditional land administration authorities, producers’ associations, NGOs, and academia. 12. Component 3 – Project Management, Monitoring, and Evaluation (US$1.75 million of Credit funds at appraisal). A PCU based in the city of Santa Cruz was in charge of coordination, administration, financial management, and M&E. Regional technical units assisted with monitoring and provision of TA. 1.6 Revised Components 13. The initial experience of the PPA showed that land market conditions were not conducive for achieving the PDOs exclusively through the land acquisition mechanism. Accordingly, Component 1 was revised in the Project’s first restructuring in April 2010 to include three additional land access mechanisms: o Leasing agreements (arrendamiento): Under this mechanism, a landowner agreed to temporarily transfer the usufruct for a specific time period in exchange of payment. In practice, PAs signed one-year leases with landowners. The Project covered the cost of lease payments. o Share-cropping agreements (aparceria comunal): Under this mechanism, the landowner granted use of the land to a PA in exchange of a percentage of the equity and a negotiated amount of the production obtained from the land. Share-cropping agreements were also for one year. o Land use rights (derechos de uso de suelo): This mechanism was only available for members of TCOs and TCs. According to the 1996 agrarian reform law, land held under TCOs and TCs is collective property. Traditional authorities (capitanes) determine land use and allocation based on customary practices. PAs received land use 4 rights for carrying out investment sub-projects, which was equivalent to a long-term lease contract. These leases were free and for an indeterminate time period. The condition for holding the lease was the continuation of the investment sub-project. 14. In addition, the 2010 restructuring modified the two-tier design of the Land Credit Facility to allow first-tier IFIs to provide credit directly to PAs. This modification was in response to changes in the country’s financial regulatory framework. 1.7 Other significant changes 15. 2010 Restructuring. As mentioned, the Project had a Level-2 restructuring approved in May 2010. This restructuring included the following changes: (1) modification to the two-tier design of the Land Credit Facility, by eliminating the financial intermediation of the BDP and introducing the direct provision of the credit line through first-tier commercial banks or regulated non-bank IFIs; (2) inclusion of three additional mechanisms for land access; and (3) a minor reallocation of Credit proceeds to reflect the refinancing of the PPA. 16. Expansion of Project area. The GOB proposed to expand the area of coverage to increase Project impact. The PCU prepared social, economic and environmental studies in six municipalities. The analysis recommended expanding the Project to the municipalities of Concepcion and San Ignacio de Velasco in the Chiquitania region. This decision was discussed during the MTR and the Bank gave its no-objection in August 2012. 17. 2013 Restructuring. The Project had a Level 2 restructuring in January 2013 that introduced the following changes: (1) reformulation of the Results Matrix to better reflect and quantify the targets of the Project as discussed in Section 1.2; (b) establishment of a new disbursement category to make explicit the expenditures for consultant services, training and operating costs for Component 2 (productive investments); and (3) reallocation of Credit proceeds to reflect the evolution of the Project, which resulted in an increased emphasis on the land use rights mechanism and the need of additional field staff for implementation, technical assistance and for the communication strategy. 18. Extension of Closing Date. The Project had three extensions to the closing date for a total of 2.5 years of additional implementation. The Project was approved by the World Bank’s Board in October 2007, but it only became effective in March 2009. One factor that delayed effectiveness was the legal change in financial regulations that affected the Project’s two-tier structure for the Land Credit Facility. The DCA had established as one of the conditions of effectiveness the signing of an Administrative Agreement with BDP. Since the GOB could no longer meet this condition, the Bank agreed to a waiver. After the Credit became effective, an amendment to the DCA was required. However, the processing of the amendment took several months due to delays by the Bank and the Borrower. These initial delays affected the implementation schedule. The first extension was approved in May 2012 after the MTR confirmed that the PDO continued to be relevant and achievable. The closing date was extended from June 29, 2012 to March 31, 2014. This extension provided PAs with sufficient time to finalize their investments, complete productive cycles, and receive follow-up, especially in the new municipalities. Two 5 additional three-month extensions were approved in November 2013 and May 2014 bringing the actual closing date to September 15, 2014. The third extension required Regional Vice President Approval since the proposed date was 2.5 years from the original closing date. Overall, these extensions were requested to improve the conditions of sustainability of the PAs and complete the final evaluation. 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 19. Soundness of Background Analysis. Project design was based on a thorough analysis of other Bank-financed ‘willing seller-willing buyer’ projects in Brazil, Honduras, Guatemala, and Mexico. Based on a growing body of knowledge, project design introduced a Land Credit Facility to provide loans for land acquisition accompanied by intensive TA and matching grants. A land market study was prepared in 2004 to collect data on land sales in the Department of Santa Cruz and inform on financial parameters for land acquisition under the Land Credit Facility. In retrospect, this study failed to account for significant factors affecting the supply of agricultural land, such as the legal restriction preventing the transfer of parcels under 50 ha and the stark segmentation of the land market with few parcels available in the 50 to 100 ha range. In addition, a follow-up 2012 analysis suggested that the original study undervalued land prices. A financial and economic analysis informed on the viability of investment sub-projects and loan repayment. Social and environmental studies informed the selection of Project area, characteristics of beneficiaries, social communication, and procedures for the approval and monitoring of investment sub-projects. 20. Redistribution through land markets. The Project was designed to address the inter-related issues of land access and agricultural productivity. When the GOB submitted the original request to the Bank for project preparation in 2002, the aim was to facilitate a ‘willing seller-willing buyer’ model. There was concern that, on its own, traditional land reform would not deliver good quality agricultural land to the poor in a fast and cost- effective manner. The National Land Reform Institute (INRA) had only identified 30,000 ha of public lands suitable for distribution, and much of it was located in remote, high-cost areas. Furthermore, the lack of progress in reverting ill-acquired lands and the virtually stagnant settlement policy (five settlements in 8 years) had led to increased frustration by indigenous and peasant groups. 21. Geographical targeting. In its original request, the GOB requested developing a land access pilot in the Department of Santa Cruz, where resolution to land conflicts was considered to be of the highest priority. This region was experiencing increased land inequality as a result of rapid growth in agro-industrial and livestock production. There were thought to be significant areas of good agricultural land, at an affordable price, that were underutilized as owners preferred to expand the agricultural frontier or held land speculatively. This setting was considered to be an opportunity to benefit indigenous and peasant groups with access to high quality agricultural land. In hindsight, it is questionable how feasible would the land acquisition mechanism be in the agro-industrial areas of the Department. In addition, the selection of the Project area was controversial due to the 6 difficult relationship between the central government and regional groups in the Department of Santa Cruz. 22. Evolving policy environment and pilot nature. The 2006 policy on agrarian reform advocated the free provision of public land to indigenous and peasant communities. However, the analysis carried out during project preparation indicated that this approach would require significant lead time, while not addressing the needs of all landless and peasant groups, particularly the need for land access in locations near existing settlements and public infrastructure as well as the need for an integrated package for productive investments. The Project was designed as a pilot precisely to test the land acquisition mechanism as part of a menu of options. The pilot emphasized the generation of learning and demonstration effects to solve the problem of land access and productive potential. Since only titled land could be acquired, no funds for land regularization activities were allocated under the Project. The pilot ultimately showed that the land acquisition mechanism was not feasible, but this in itself is a valuable lesson. 23. Emphasis on beneficiary-driven, decentralized process. The Project was designed with attention to have a beneficiary-driven, decentralized process that was conducive to building the productive capacity of the beneficiaries. During preparation, INRA’s participation as implementing agency was rejected due to its poor-track record and lack of expertise in investment sub-projects. In hindsight, this decision was adequate because INRA’s institutional mandate does not necessarily cover the distribution of private lands and the promotion of investment sub-projects is outside of its functional competencies. Instead, the Vice-Ministry of Land, as part of the then Ministry of Rural Development, Agriculture, and the Environment (MDRAMA), was made responsible for project implementation. The PCU based in Santa Cruz was in charge of managing the Project with the assistance of CODALs for the evaluation of sub-projects and social monitoring. 24. Risk Assessment. Overall Project Risk was assessed as substantial at appraisal. This rating was appropriate and reflected the complexity of addressing a structural issue as well as the controversial aspects that emerged from a politicized environment. The Project successfully mitigated the opposition posed by proponents of traditional reform by fostering consultation and integrating TCOs and TCs into its land access mechanisms. Risk assessment recognized that the Project would only address one dimension of the land tenure situation in Bolivia, which was the redistribution of land already in the private domain. During the implementation period, the GOB pursued a parallel effort to regularize and redistribute land. By excluding these activities from the Project, there was a greater chance to pilot innovative land access mechanisms and minimize reputational risk. The risks specific to the land acquisition mechanism did not materialize, such as the risk of loan default and favoring the wealthy, since the mechanism was not implemented. One risk identified was that the demand for land could be either too little or too great. Project preparation, however, did not consider the supply of land as a risk. Finally, the risk assessment warned about the inter-regional tensions, although it was difficult to propose mitigation measures. 7 2.2 Implementation 25. Implementation improved during the course of the Project’ life. Between 2005 and 2009, the Project used PPA funding to start piloting the land acquisition mechanism. The PPA for US$ 1.5 million was signed in April 2005 and became effective in June 2005. However, due to the political climate, activities did not start until November 2006. There were further delays caused by the processing of the Supreme Decree to authorize the transfer of funds to beneficiaries. Once the PCU was ready to start operations they realized that they could not find suitable land for purchase. As a result, the PPA Agreement was amended in May 2008 allowing the PCU to pilot the alternative mechanisms of land leasing, sharecropping, and land use rights. The Credit became effective in 2009. The Project had to be immediately restructured to adjust to the new legal changes to the financial regulatory framework and incorporate the other three mechanisms. However, delays in the processing of the first restructuring, and subsequent approval of the required Supreme Decree virtually stalled Project operations during 2010. During this period, there was no approval of new investment sub-projects. Between 2011 and 2014, the Project re- concentrated activities on TCOs and TCs and implementation progress improved significantly. The Project benefitted from a strong political commitment to agrarian reform, the substantial progress achieved on the legal recognition of communal property, and flexibility in adapting to an appropriate model for the provision of TA. Project implementation, however, was negatively affected by the initial lack of government ownership, an increase in land values, and conditions in the financial sector. 26. Several factors contributed to successful implementation: 27. Government commitment to social inclusion and poverty reduction. Overall, the GOB has shown a strong commitment to the goals of eradicating extreme poverty, sharing prosperity along the concept of vivir bien (living well), and empowering indigenous peoples. The 2006 National Development Plan (NDP) captured an ambitious policy agenda to transform the country. Agrarian reform played a central role in the policy scheme for an agrarian revolution. A new law was approved (Law No. 3545) to reinvigorate agrarian reform. This policy environment raised the profile of the Project. 28. Progress made in recognizing land rights in TCOs and TCs. Bolivia has made high- level commitments, through its constitution and international agreements (such as the International Labor Organization Convention 169), to indigenous rights and communal property. The 1996 agrarian reform law defines the procedures for the legal recognition of communal property to indigenous people in the form of TCOs and peasant communities in the form of TCs. Between 1996 and 2005, INRA distributed over 6 million hectares of land to TCOs. This was the most far-reaching of the distribution efforts under the 1996 agrarian reform law. This legal recognition of communal land tenure provided an opportunity to offer a land access option that is fully aligned with customary rights and the rule of law. 29. Adaptation of Technical Assistance to meet the needs of PAs. The first generation of models for investment subprojects was based on complex designs that incorporated multiple production lines. However, this approach required a complex mix of TA and monitoring support, which was not efficient for Project implementation. In 2009, this 8 approach was reoriented to favor specialized models based on investments adapted to local conditions and good potential for commercialization. Also, the original modality for TA provision required beneficiaries to directly select and hire these services, but this approach did not work because of the limited supply of service providers. In response, the PCU helped organize a fee-based firm that operated within its structure to provide the TA. The firm had offices in each of the municipalities and were staffed by technicians specialized in the type of sub-projects in the area and social promoters who spoke indigenous languages. This arrangement worked relative well, but created confusion between the services that the PCU offered for free and those offered by the firm that required payment by the PAs. This modality also increased the PCU’s operational costs. After assessing this situation, it was agreed to increase the technical capacity of the PCU and its regional offices to provide TA in a more integrated way. 30. Several factors – some outside and some within the control of government and implementing agency – adversely affected the Project: 31. Initial lack of government ownership. The decentralized nature of project management affected government ownership. As mentioned, the Project was prepared by a previous administration that leaned more towards market-based policies. In addition, the location of the PCU in the City of Santa Cruz left the Project relatively isolated from the MDRyT and the Vice-Ministry of Lands based in La Paz during the early period of implementation. Following the effectiveness of the legal amendment of 2010, the MDRyT’s ownership of the Project was strengthened resulting in enhanced support for project implementation. 32. Increase in land values. The land acquisition mechanism envisioned in project design was not implemented. Although an ex-ante land market study was conducted during project preparation, some important characteristics were not considered. There were imperfections that inhibited the functioning of land markets: many properties lacked legal title since the land regularization process slowed down; the agrarian legislation prohibited the sale or collaterization of parcels smaller than 50 hectares; large landowners invested in agricultural land for speculation; and some landowners were reluctant to sell through the formal market, seemingly motivated by tax and other privacy issues. The main market for agricultural land was for estates larger than 400 ha, which were too large for the PAs. These conditions contributed to a limited supply of land available for acquisition through the Project. Furthermore, the agricultural commodity boom in 2005-2008 significantly increased land values. A study financed by the Food and Agriculture Organization Cooperation Program (FAO-TCI) in 2012 estimated that since appraisal land prices had increased an average of 300 percent in the three municipalities initially covered by the Project. This study also determined that the prices included in the ex-ante land market study were below the market rate for 2004. This meant that the loan amounts established for land acquisition under the Project were undervalued. Knowledge about the limited supply of eligible land and high land values informed the GOB decision not to continue pursuing the land acquisition mechanism. 9 33. Problems with establishing the Credit Line Facility. In parallel to the problems described above, the Project was unable to establish the revolving credit line for the land acquisition mechanism. Project design included the provision of capital to establish a credit line for land acquisition because private financial institutions in the region were unwilling to participate with their funds due to the perceived high risk and high cost of lending to this sector. Nonetheless, negotiations with BDP for the implementation of the two-tier system were delayed and this system was ultimately modified due to a change in the regulatory framework that eliminated the need of using this second-tier bank for financial intermediation. The credit line could not be established through an IFI either. The tender to contract the administration of the credit line had to be opened twice due to the few offers received. Although the contract was awarded to Banco de la Union S.A., the IFI did not sign it. Low profitability seemed to be the main factor explaining this lack of interest on the part of IFIs. Initially, the Land Credit Facility was supposed to have a capital allocation of nearly US$ 4.6 million but then it was reduced to about US$ 0.9 million when funds were reoriented towards working in TCOs and TCs. The low capital level did not seem to compensate for the IFI’s transaction and administrative costs, which tend to be high for this sector. 34. Financial problems with IFI. In 2011, the PCU selected the Cooperative San Gabriel, a local IFI, to transfer the grant proceedings to PAs. In 2012 and 2013, the IFI had unexpected financial problems and lost liquidity. As a result of this problem, the cooperative withheld the resources needed for sub-project implementation, delaying disbursements to 185 PAs. An extension of the closing date had to be requested to compensate for the delay in the implementation of subprojects and Graduation Plan. This prevented 111 PAs from paying the service providers’ taxes that each contracting entity has to withhold and later on pay to the Tax Agency, according to the local legal framework (15% of the contract’s worth).The MRDyL addressed the situation through various legal measures (including a legal recourse against the cooperative's physical assets). By the closing date, the MRDyL had managed to recover 26 percent of the funds and agreed to use this money to pay the withheld taxes on behalf of the PAs. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization 35. As mentioned, the original Results Framework was modified to reflect the new orientation of the Project and to monitor social inclusiveness and equity. Although a baseline study was not carried out at the beginning of the implementation, every beneficiary had to fill a questionnaire, as part of an affidavit for eligibility, which included key data on household composition, economic activities, income of the members, and land use. Each sub-project had an Implementation Plan (IP) that established targets for the installation of productive assets, a disbursement calendar, and a productivity improvement plan. The PCU used both references to measure the level of achievement of the IP, income changes and the number of hectares of cropped or grazed land. 36. The PCU maintained a specialized M&E unit during the course of implementation. From 2007 to 2010, the process to approve sub-projects was not well established and there was no systematic collection of information about their progress. The situation improved in 2011 when a follow-up system of sub-project milestones was established, including a 10 ranking system that provided a quick snapshot of the physical and financial progress. In 2012, the PCU hired an environmental specialist to support in the monitoring and follow- up of environmental and social aspects of the sub-projects. 37. The PCU prepared a completion report for each sub-project to systematize the experience and monitor key results. The members of the CODAL also participated in the evaluation of sub-projects, which contributed to their capacity to evaluate sub-project proposals and the establishment of a decentralized monitoring system. From 2011 to 2014, FAO-TCI and Bank team provided technical support to the M&E unit by carrying out several field visits and adjusting instruments for data collection. The PCU contracted external evaluations for the MTR and project closing, and FAO-TCI financed a land market study in 2012. 2.4 Safeguard and Fiduciary Compliance 38. Overall safeguard compliance is considered satisfactory. The Project triggered the Environmental Assessment (OP/BP 4.01), Natural Habitats (OP/BP 4.04), Pest Management (OP/BP 4.09), Physical Cultural Resources (OP/BP 4.11), Indigenous Peoples (OD 4.20), and Forests (OP/BP 4.36) safeguard policies. As the majority of beneficiaries were expected to be indigenous peoples, the Project itself was implemented as an Indigenous Peoples Development Plan. The findings and recommendations of a social assessment informed project design. Involuntary Resettlement (OP/BP 4.12) was not triggered as the social assessment did not consider that the Project would affect the land rights of third parties. The PCU was staffed with social and environmental specialists and each of the municipal offices had social promoters who spoke the local indigenous languages. A World Bank social specialist based in La Paz supervised social and gender issues. Supervision confirmed that overall participation and consultation activities were satisfactorily implemented and benefits were consistent with the beneficiaries’ culture and social organization. A grievance redress mechanism was set up in 2012 to provide channels for beneficiaries to present questions or complaints, with a view to strengthen the project’s responsiveness to beneficiaries’ needs. Entry points included telephone lines, complaint boxes located in project offices, and, most importantly, personal assistance in project offices. Entries, most of which concerned questions and requests for technical support, were recorded and analyzed by the Social Specialist responsible for the system. 39. An Environmental Assessment and Environmental Management Plan (EMP) were also prepared. The PCU complied with the criteria and procedures in the EMP for the environmental screening of lands, the environmental management of sub-projects, and the mainstreaming of environmental and sustainable land management practices through technical assistance and training. The first generation of sub-projects experienced delays in obtaining the environmental licenses required by the National Environmental Law. In October 2012, the PCU hired two consultants to assist the PAs in obtaining all the required permits, strengthening the Project’s capacity to comply with the environmental legislation and the Bank’s safeguard policies. The PCU developed environmental management tools, such as screening forms and monitoring reports, in order to ensure the implementation of proper environmental and social practices and systematize the experience with environmental and social management in the sub-projects. All the beneficiaries received 11 training to improve environmental managements such as pesticide use, water conservation, and protection of critical or sensitive areas. 40. Financial Management capacity had significant shortcomings and overall compliance is considered moderately unsatisfactory. Adequate procedures for the implementation of sub-projects were not developed until 2011. Weaknesses in internal control and the lack of an adequate tool to support the timely and comprehensive recording of information at sub-project level prevented an adequate monitoring of sub-projects and timely preparation of financial reports. In addition, the PCU did not manage to document all the expenses incurred by PAs, which caused some of them to be considered as ineligible. There were also some overdue audit reports and delays in resolving observations to audit reports. The cash flow and financial situation of the Cooperative San Gabriel initially put 180 PAs at risk. Weaknesses in internal controls also led to some differences between accounting records and financial reports, all affecting the reliability of financial information. Although FM issues were managed with the agreement of action plans, the PCU had problems attracting and maintaining high-qualified fiduciary staff, which is a general problem in Bolivia’s public sector, and which became more evident in such a complex and challenging project. 41. Procurement capacity improved during the course of implementation and overall compliance is considered satisfactory. Several key processes were put on hold in 2010 due to the uncertainty about the future of the Project, but they were subsequently completed. The PCU had to update the Procurement Manual for PAs to improve operations of sub- projects, analyzing the supply chain for the different needs required in the project activities. In this regard, a protocol for the purchasing of livestock was prepared that included procurement and technical aspects to improve transparency and competitiveness. Also, the principles of transparency and competitiveness were reinforced through a ‘train-the- trainers’ model for the procurement and technical staff who then trained the fiduciary members within the PAs on how to perform each procurement activity. 2.5 Post-completion Operation/Next Phase 42. The Project contributed a tested model for accessing land through the land use right mechanism in TCOs and TCs. The Project was designed as a pilot effort, but a follow up operation was not prepared due to IDA resource constraints and country’s priorities during the final year of the project. Although there were no specific arrangements for continuing the implementation of this mechanism by Project closing, the MDRyT considers the model as a key result that could be replicated in other parts of Santa Cruz and Bolivia. 43. On the other hand, as detailed in Section 4 below, the PCU developed a graduation plan for the PAs to help increase their sustainability beyond the project’s life. By the end of Project, the capacity of the PAs was strengthened. Now, they have the conditions to participate in other rural development programs such as Rural Alliances to continue investing in infrastructure, working capital, and TA. 44. 3. Assessment of Outcomes 12 3.1 Relevance of Objectives, Design and Implementation 45. The PDOs continue to remain highly relevant as corroborated in the 2006 National Development Plan (NDP) and the current WB’s Country Partnership Strategy (CPS) FY2012-2015. The Patriotic Vision 2025 is a more current version of the GOB’s long-term strategy, but the NDP continued to inform its operationalization by the project closing. The NDP included a policy for the ‘Transformation of the Tenure Structure and Access to Land and Forests’ that would pursue equity in the distribution and access to rural land, the consolidation of territorial rights, and the elimination of the large landed estates. The Project adapted to the new policy environment becoming the only intervention during this period to pursue the goal of consolidating land access in TCOs and TCs through productive investment. The Project contributed to the CPS outcome of increasing the income generation capacity for excluded groups and rural producers through increased access to productive land and more predictable access to consumption markets. 46. The pilot nature of project design and its emphasis on a menu of policies was crucial for maintaining relevance in an evolving policy environment and market conditions. The Project redirected efforts to other land access options when the land acquisition mechanism lost its relevance under the new policy framework and land market conditions. The Project’s emphasis on consolidating land rights in the private domain, individual or communal, as an alternative to traditional agrarian reform continued to be relevant. Overall, the quality of the project design was high and contributed to the continued relevance of the PDOs. The risk assessment was candid and proved to be relevant for risk mitigation. The lessons from other ‘willing seller-willing buyer’ land access projects in the region remained highly relevant to the other three land access mechanisms. 47. Project implementation helped to articulate an alternative strategy around the concept of land access to reinforce the goals of agrarian reform. Although the Project faced the risk of early closing in 2010, the expansion of the Project area and a strategic restructuring upheld its relevance and helped the Government and the Bank in agreeing to continue with the project. The decision to cancel the land acquisition mechanism was informed by an independent evaluation prepared for the MTR and the 2012 land market study. The reallocation of credit proceeds contributed to the achievement of PDOs by increasing funds for sub-project grants and technical assistance. 3.2 Achievement of Project Development Objectives 48. The Project partially achieved its two PDOs: Establishing an alternative decentralized beneficiary-driven land distribution mechanism for organized landless or poor farmers (partially achieved) 49. This PDO was achieved through decentralized mechanisms to provide increased access to productive land to beneficiaries organized under PAs (Component A). The Project exceeded the outcome indicators associated with this PDO: a total of 2,890 households (131% of target value) organized in 237 PAs (125% of target value) to access 151,579 ha of land (132% of target value). The PAs were distributed by tenure type as follows: 138 in TCOs, 82 in TCs, and 17 in private lands under the lease contract or 13 sharecrop agreement. TCOs and TCs represented over 99 percent of land approved for productive investments. These achievements were made possible by the establishment and operation of CODALs in five municipalities (100% of target value). 50. The Project successfully piloted alternative mechanisms for land access. The results of the pilot showed that the land acquisition model was not feasible and that the land use rights mechanism had great potential. The leasing contract and sharecropping agreement mechanisms had a minimal impact. Before the Project, there was no formal system in the TCOs or TCs to allocate land to local farmer organizations. In line with the original design, the CODALs functioned as a decentralized decision-making entity and provided social monitoring. The communication strategy ensured that households were informed about this opportunity to access land within their own communities through the use of mass media and, most importantly, community meetings. 51. The land use right mechanism was also effective in reaching the target population. Most beneficiaries in TCOs and TCs already possessed small parcels (1-3 ha as per the affidavit), but according to baseline values they struggled to sustain a livelihood. Without capital, they could not request additional land to expand production. In addition, about 10 percent of beneficiaries were landless prior to Project intervention. This group was mainly composed of single women and young people in their early twenties. All the PAs in TCOs and TCs maintained their land use right at the end of the Project. 52. The other two mechanisms—lease contracts and sharecropping agreements—had a limited impact. Only 17 PAs accessed 826.75 ha of land through these mechanisms. PAs had short-term contracts or agreements that normally had to be renewed on an annual basis. With the exception of one PA, family members provided the land access. Although these mechanisms helped both small and landless farmers gain temporary access to land, the MTR recommended focusing on TCOs and TCs because of the larger potential for impact. 53. In spite of these achievements, the PDO was not fully achieved because the Project was unable to establish a self-sustained mechanism that could be scaled up. The original design included a credit line that would guarantee future funding for land access. Instead, the land use rights mechanism relied on grants and by the time of Project closing it was unclear whether the CODALs would manage to secure additional financing for scaling up. Furthermore, the original design emphasized the importance of establishing a mechanism to redistribute highly productive agricultural land held under individual private ownership. Although the Project helped consolidate land access within TCOs and TCs, it had no impact on this other land tenure sector. For this reason, this PDO is considered to be partially achieved. Supporting productive investments thereon to allow increased and more sustainable livelihoods for such beneficiaries (partially achieved) 54. This PDO was achieved through the provision of grants to PAs for small infrastructure and productive investments on their lands, coupled with technical assistance and capacity building (Component B). The data to assess this PDO came from the PCU’s 14 monitoring system, an external financial and economic analysis, and an independent final evaluation. 55. The Project allocated a total of US$ 9.73 million for sub-projects, 78 percent from Project grants and 22 percent from PAs’ counterpart contributions. Counterpart funding exceeded the estimate of US$ 0.63 million by 347 percent. The sub-projects invested in livestock (70%), agriculture and agroforestry (23%), and other activities such as aquaculture (7%). A total of 27 out of 237 PAs (27%) discontinued their sub-projects, mainly due to organizational problems and loss of interest as a result of project inactivity in 2009-2010. 56. An external financial and economic analysis provided key measurements for evaluation based on a sample of 83 PAs. According to this study, the investment sub- projects increased family income by 39 percent 4 compared to baseline, surpassing the revised target of 20 percent. Family income increased for all sub-projects types and in all municipalities, but there was a wide range of values. The sub-projects in Minero were in the lower end with an 11 percent increase compared to the sub-projects in San Ignacio Velazco and Pailon which reported a 45 and 49 percent increase. In certain areas, alternative crops, such as chia and sesame generated an increase of 50 percent of the producer’s income. Another target was to increase the value of productive assets by at least 15 percent (comprising value of animals and work capital for agriculture). For agricultural investments, family assets increased by US$122 or a 15 percent increase compared to baseline. For livestock investments, family assets increased by US$ 156 or 10 percent compared to baseline. This underperformance is in part explained by the drought events that affected Charagua in 2012-2013 and the slow breeding cycles that delay the economic return of cattle investments in the Chaco region. In a way, it is still early in the productive cycle to measure the impact on livestock assets. 57. Overall, PAs maintained a good organizational and productive capacity. According to the PCU’s monitoring system, 85 percent of PAs complied with their Implementation Plans. This value is above the target of 70 percent. An independent evaluation, however, concluded from a sample of 105 PAs that only 19 percent of them reached an optimal production level (between 80 and 100 percent of targets achieved). This evaluation also found that 52 percent of PAs achieved less than 40 percent of their targets. Another important achievement was that 74 percent of PAs had women in their boards, significantly surpassing the target of 10 percent. 58. As mentioned earlier, PAs accessed 151,579 ha of land, but only 59 percent of this area was under productive use at project closing. This value is below the target of 70 percent. In part, this underperformance is explained by the CODALs’ decision to approve more land than immediately required in anticipation of future increases in livestock production. A positive aspect is that by the end of the Project production was maintained in 99.7 percent of this active land surface. 4 In case of livestock subprojects, the income were calculated taking into consideration the sale of the first animals, which might happen next year or within two years in case of San Ignacio Velasco and Conception. 15 59. The Project had a mixed performance with TA and capacity building activities. Prior to 2011, sub-projects had complex designs with multiple production lines that required intensive TA. This approach was reoriented to develop specialized sub-projects which were well adapted to local conditions and had commercialization potential. In spite of this improvement in TA, the independent evaluation found that only 73 percent of members were satisfied with the TA provided by the Project (rated as good or excellent), a value below the target of 80 percent. Also, the evaluation found that only 67 percent of PAs had completed their training plan. A positive aspect is that 38 percent of members participating in training were women, a value which is above the target of 20 percent. 60. Even as PAs completed their targets and families benefited from higher incomes and an expanded productive asset base, the contribution to sustainable livelihoods is not fully guaranteed. Particularly in Charagua the level of socio-ecological vulnerability is high and the experience of previous projects attests to the difficulty of reaching sustainable outcomes. Sustainability issues are discussed in more detail in Section 4. For this reason, this PDO is considered to be partially achieved. Table 1: Land Approved v. Land under Production in the Project Area Approved Land Land under Production Number Municipality Average Size per Average Size per PA of PAs Size (ha) Size (ha) PA (ha) (ha) Charagua 127 128,001.00 1,015.88 83,010.75 653.63 Concepción 30 3,315.00 110.50 933.50 31.12 Minero 14 562.22 40.16 318.73 24.52 Pailón 19 7,600.73 400.03 503.00 33.53 San Ignacio 47 12,100.39 257.45 3,920.00 83.40 Total 237 151,579.34 364.80 88,685.98 165.24 3.3 Efficiency 61. An ex-post financial and economic analysis was conducted to evaluate the economic efficiency of the Project based on a sample of 83 investment sub-projects. The evaluation measured economic efficiency based on investment subprojects, which represent the bulk of project financing (72.6%). Although technical assistance and training activities also have an economic value, the methodology to quantify the economic value of sub-projects is more straightforward and rigorous. The sampling and evaluation methodology is explained in Annex 3. 62. A direct comparison of ex-ante and ex-post indicators is not possible given that the land acquisition mechanism was not implemented. At appraisal, the ex-ante financial analysis considered indicative models of potential subprojects that acquired land through a loan with a 15-year repayment period at an 8 percent annual interest rate. The aggregate Financial Net Present Value (FNPV) of all investment subprojects was estimated at $17.65 million and the Financial Internal Rate of Return (FIRR) at 16 percent. At completion, the analysis was based on actual performance and a ten-year cash flow projection that started to run at the end of the investment period with a discount rate of 12.67 percent (the official rate in Bolivia). 16 63. The aggregate Financial IRR and NPV of the analyzed sub-projects are 18.3 percent and US$ 1.6 million respectively. These estimates include only the sub-projects’ direct costs. If a proportion of project management and TA costs is included, the FIRR and FNPV are reduced to 14.9 percent and US$ 0.8 million respectively. The municipality with the highest IRR is Pailon (26.9%), followed by San Ignacio de Velasco (25%), Minero (17.6%) and Conception (15.4%). Charagua has the lowest IRR (13.1%). This municipality had more than the half of the investment made during the Project. This lower performance compared to other municipalities is explained by (1) lower incremental returns from investments in cattle (meat and dairy production) due to the longer gestation period of these investments; (2) climatic factors that affected the agricultural sub-projects (corn, beans and sorghum), e.g. the drought faced by Charagua during Project implementation prevented the development of a greater volume of crops in this region and reduced yields; and (3) productive associations seem to have less entrepreneurial skills than their counterparts. 64. The aggregate Economic IRR and NPV of the analyzed sub-projects are 21.76 percent and US$ 2.5 million respectively. The EIRR in each municipality is above the social discount rate, which means that the Project is socially profitable. Minero is the municipality with the highest EIRR (34.4%), but its ENPV is low (US$0.12 million) due to the reduced number of PAs. San Ignacio de Velasco has the highest ENPV (US$ 1.58 million). 65. The cost of the investment and TA package (Component 2) was US$ 1,016 per beneficiary or US$ 4,036 per family. These values are similar the cost per family of investment and TA package in Guatemala (US$ 4,102), Honduras (US$ 3,952), and Brazil (US$ 3,981).5 The context of each country, however, should be taken into consideration. 66. Overall, the Project had a modest impact on efficiency. The financial and economic IRRs are above the discount value, but the gains in NVP are modest. Cost per household are on par with similar projects. 3.4 Justification of Overall Outcome Rating Rating: Moderately Satisfactory 67. As detailed above, the Project remains highly relevant, while its PDOs were partially achieved and most of its performance indicators were met or exceeded, which contributed to improving land access for more than 11,000 individual beneficiaries. The GOB has a tested alternative beneficiary-driven decentralized mechanism to operationalize agrarian reform policy, which also offers the measures required to support livelihoods in the newly accessed land. This mechanism directly reached beneficiaries in indigenous and peasant communities, among the most marginalized population in the lowlands of Bolivia, 5 Guatemala Land Fund Project (P054462), closed in 2005; Honduras Access to Land Pilot (P073035), closed in 2006; Brazil Land Based Poverty Alleviation Project I (P050772), closed in 2008. The calculation assumes that 45 percent of total package was spent on productive investments and TA package; the remaining 55 percent was for land purchase. The estimates are in 2015 dollars. 17 and offered them a solution that is consistent with territorial rights and the rule of law. This alternative is rather cost-effective as beneficiaries did not have to pay for the cost of the land. Overall, this mechanism is consistent with the FAO Voluntary Guidelines on the Responsible Governance of Land Tenure and could become a model to other countries in the region. 68. The Project, however, fell short of establishing a self-sustaining mechanism to continue expanding land access to landless and poor farmers. This aim remains highly relevant as the dynamics of land accumulation continue to persist, particularly in the lowlands of Bolivia. In fact, if the land acquisition mechanism would have been implemented on a timely basis, the first batch beneficiaries would have enjoyed the increase in land prices. Nonetheless, the political, financial, legal, and market conditions were not conducive to the emergence of this approach. The Project took 2.5 additional years for completion. Furthermore, the Project had a moderate impact on efficiency and the sustainability of livelihoods is not guaranteed. PAs had significant gains in organizational and productive capacity as well as gender equity, but they would face challenging agro-ecological and market conditions. This situation was to be expected considering the characteristics of the target population and geographic area. Given these moderate shortcomings, the Project is rated as moderately satisfactory. 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 69. Poverty Impacts and Social Development. The Project reached a sector of the rural population historically excluded from development opportunities. The Project reached members of TCOs and TCs, who legally owned land through communal property, but in practice did not have the capital and technical support to exercise this right. Precisely because of their low asset base, this target group did not qualify for assistance under other rural development projects. The Project raised beneficiaries to a level where they can benefit from programs such as the Bank-financed Rural Alliances Project. In addition, the Project helped reduce poverty through increased income from investment and salaries paid for daily work. The independent evaluation already noted a reduction in the number of members participating in seasonal migration. Indirectly, the Project’s investment on capacity building contributed to the implementation of their Management and Life Plans, which are the planning instruments for TCOs. 70. The Project built upon customary arrangements. The traditional authority (capitan) provided the use rights permits upon request by the PAs. Initially, it was thought that the PAs would pay a fee for using the land. Capitanias, however, accepted to relinquish payments. This decision was taken in order to avoid setting a precedent for fee-based land allocation inside TCOs and TCs. There were already other programs in the area that had not paid this fee. Also, it helped reduce costs to the PAs. During implementation, neither the PCU nor the Bank supervision noticed a complaint from non-beneficiary households about exclusion from the Project. In part, this could be explained from the fact that there were plenty of underutilized lands in the community. 18 71. The establishment of PAs contributed to the strengthening of social capital and empowerment of their members. The creation of PAs built upon the minga, a traditional associative scheme among the Guarani and Chiquitanos. The collective efforts required to agree and implement investment plans strengthened organizational capacity and built leadership skills. A large number of beneficiaries, particularly women and youth, did not have identity documents prior to this intervention. This integration into national records is an important component of democracy building. Through this support, project beneficiaries, 66 percent of them indigenous, have more opportunities to benefit and contribute to the life of their communities and country. 72. Gender Aspects. The Project also improved the conditions for gender equality. Women in rural areas, and particularly among the Guarani, had a limited participation in productive and organizational aspects of their communities. The Project managed to achieve equity in the participation of men and women as beneficiaries. In terms of administrative duties, women held 32 percent of board member positions and held the presidency in 8 percent of PAs (PA boards were composed of a President, a Vice President, a Treasurer, a Secretary, and up to three additional members). This means that women participated in the boards of 176 PAs or 74 percent of the total. Although equality was not reached, this represents an effort to improve equity. These improvements were in part due to the attention that the PCU gave to gender issues. The PCU generated a ‘directory of women’ to monitor their participation in the Project’s various activities. Also, the PCU provided dedicated assistance to women-led PAs to strengthen their leadership role and compensate for any lack of previous experience. (b) Institutional Change/Strengthening 73. Project preparation became a site of consultation and mediation of various positions concerning agrarian reform at a historical moment in Bolivia. Through implementation, the Project fostered a dialogue with the Vice-Ministry of Lands on how to operationalize the country’s agrarian reform policy. In the future, the mechanism piloted in the TCOs and TC could become a model for expanding land access based on usufruct rights in other parts of Bolivia and other countries. The Project also contributed to decentralization and social accountability through the establishment of CODALs. The CODALs have the potential to continue organizing and monitoring development projects in their municipalities. (c) Other Unintended Outcomes and Impacts (positive or negative) 74. Compliance with the National Tax Service Authority (SIN). The PAs faced problem of fines and unpaid taxes due to circumstances outside of their control. When PAs opened their bank accounts in Cooperative San Gabriel, they were registered in the SIN under a business tax filing status that required them to present financial statements. The PAs did not comply with this requirement and incurred fines. The PCU and SIN analyzed the situation and concluded that the filing status had been inappropriate for the type of economic activity conducted by the PAs. The SIN agreed to retroactively change the filing status in order to cancel the fines and then close their National Tax Identification Numbers (NIT). By Project closing date, only three out of 173 PAs had not presented the paperwork to change their filing status. 19 75. Protocol for the purchasing of livestock. The Project supported the preparation of a protocol to strengthen the capacity of PAs when purchasing livestock for their investment sub-projects. The protocol proved to be effective and now cattle ranchers in Chiquitania are using it for their own operations. In this regard, the Project is contributing to increased transparency and competiveness of the rural sector. 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops 76. The independent final evaluation included a beneficiary survey with the Project. The survey had a sample size of 58 beneficiaries from a total of 36 PAs. The PAs were selected using a representative sample. According to respondents, the main sources of information about the Project were municipal projects and programs (32.8%), communal assembly (22.4%), and neighbors (20.7%). After learning about the Project, 52 percent decided to join it during a communal assembly and 48 percent during a PA meeting. In Minero, Pailon, and San Ignacio de Velasco, all investment sub-projects were selected by the members of the PAs. In Concepcion, 78 percent were selected by the PAs and 22 percent by the communal assembly. In Charagua, 48 percent were selected by the communal assembly, 26 percent by the PAs, and 17 percent by another collective mechanism such as a communal authority. 77. Overall, beneficiaries recognized that the Project helped improve access to land, TA, inputs for agriculture and livestock production, and enhanced production. Beneficiaries, however, do not think that the Project has provided access to markets, increased income, or improved livelihoods. Also, the perception about the Project’s satisfaction of needs and expectations did not achieve the end target of 80 percent. Only 71 percent of beneficiaries responded to be highly satisfied or satisfied with the Project in this regard; 26 percent responded to be partially satisfied; and 3 percent as unsatisfied. Beneficiaries have a high level of satisfaction with how their PA functions. 72 percent of respondents rated PA functioning as very good or good, 26 percent as regular, and 2 percent as poor. 4. Assessment of Risk to Development Outcome Rating: Substantial 78. The risk to development outcome is considered substantial given that there are no arrangements to continue financing the mechanism piloted under the Project and that the PAs would face challenges for expanding production. The GOB has a proven mechanism to consolidate access to land in TCOs and TCs. CODALs are composed of local actors performing their activities within established institutional settings, and as such it offers a model that can be easily replicated in other parts of Bolivia. Nonetheless, the GOB would need to allocate funds for the matching grants and TA. 79. PAs will face challenges to maintain production of their investment subprojects. The independent evaluation found that 68 percent of PAs believed that they had the conditions to continue the sub-project, 3 percent believed that they had partial conditions, and 26 percent believed that they had no conditions. The ability of PAs to reach this outcome varies by municipality. The PAs in Pailon and Minero face the better prospects of 20 sustainability. Since they were established in the period 2007-2009, they acquired more experience and received more TA. By the end of the Project, the 14 PAs in Minero were already functioning with relative autonomy. Although PAs in San Ignacio de Velasco started operations as recently as 2012, their agro-forestry investments also have a promising outlook. The highest level of risk is in livestock investments in Charagua because of extreme climatic events. 80. As a mitigation measure, the PCU assisted each PA in the implementation of a Graduation Plan. The Graduation Plan included a set of minimum conditions that the PA must had achieved by the end of project support in order to increase its sustainability. These conditions were related to organizational, administrative, technical, and entrepreneurial competencies. As a result, the capacity of the PAs was strengthened, and by the end of the project they have the conditions to continue with their productive activities and participate in other rural development programs.. 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Moderately Satisfactory 81. This rating is based on the commitment to offer an innovative solution to solve a structural problem in Bolivia. The Project represented a ‘high-risk – high-reward’ undertaking to establish a mechanism to redistribute some of the best agricultural land in the country in favor of poor and landless farmers. Project design reflected good practice related to ‘willing seller-willing buyer’ land access projects in the region and attempted to replicate the successful experience of Brazil in this regard. Although a land market study was prepared, the analysis was not thorough enough to assess the replicability of the Brazilian model in Bolivia. The lending phase spanned over three presidential administrations, and the Bank team had to bridge the gap between evolving positions about the role of land markets in agrarian reform. Ultimately, project design was overly optimistic about the conditions for land acquisition in Santa Cruz, but the Project emphasis on piloting a menu of options provided the flexibility to reorient activities during implementation. (b) Quality of Supervision Rating: Moderately Satisfactory 82. Initial project supervision was marked by disagreements with the Government over the Project’s focus on private land purchase and geographical targeting. The Bank’s internal processing of the first restructuring package delayed implementation. Bank performance improved after the effectiveness of the first amendment. Overall, supervision missions were timely, appropriately resourced, and solution-oriented. Although the MTR was scheduled eight months prior to the closing date, this late date was more aligned with the actual mid-term due to early delays in disbursement. The MTR resulted in effective decisions and agreements (through a Project restructuring), and technical and safeguard issues affecting implementation were adequately handled through a close collaboration 21 with FAO-TCI. Environmental and Social safeguards were well supervised particularly given the level of complexity and risks associated with sensitive areas and indigenous peoples’ land rights. The successful Project restructurings attest to the timeliness and responsiveness of Bank supervision. (c) Justification of Rating for Overall Bank Performance Rating: Moderately Satisfactory 83. Bank performance in ensuring quality at entry is rated moderately satisfactory and quality of supervision is rated moderately satisfactory. Thus, overall Bank performance is rated moderately satisfactory. 5.2 Borrower Performance (a) Government Performance Rating: Moderately Satisfactory 84. Government performance is rated moderately satisfactory mainly due to the lack of government ownership during the initial implementation period. During the course of Project preparation, the policies of the GOB towards agrarian reform changed significantly. Although the GOB eventually embraced the Project as an instrument to advance this policy, the initial period was affected by significant delays. The Government took several months to approve the supreme decrees that were required to initiate activities under the PPA and approve the first amendment to the DCA. These delays put at risk the achievement of the PDO and affected the implementation schedule. (b) Implementing Agency or Agencies Performance Rating: Moderately Satisfactory 85. Implementing agency performance is rated moderately satisfactory reflecting evolving political commitments and the initial shortcomings in project management. Throughout implementation, but particularly after the first restructuring, the MDRyT remained attentive to the actions required to advance the Project and maintained an open communication channel with the PCU and the Bank. The institutional oversight by the Vice-Ministry of Lands, however, had some shortcomings due to the constant rotation of Vice-Ministers during the implementation period. These changes created a discontinuity in processes and introduced new perspectives and demands that affected operations. The PCU had four coordinators and experienced a high rotation of technical staff. Initial project management had significant shortcomings in financial management and M&E. Project management improved in 2011 when the effectiveness of the first amendment allowed hiring a new coordinator and replacing technical staff. The PCU improved its fiduciary, technical, safeguard, and M&E capacities, but continued to experience some shortcomings. (c) Justification of Rating for Overall Borrower Performance Rating: Moderately Satisfactory 22 86. Government performance is rated moderately satisfactory and Implementing Agency performance is rated moderately satisfactory. Thus, overall Borrower performance is rated moderately satisfactory. 6. Lessons Learned 87. A pilot project needs to incorporate a system to test innovation and make results available to policymakers for scaling up. Piloting alternative land access mechanisms for poor and landless peasants carries a high degree of uncertainty. In order to have flexible arrangements, project design should incorporate from the onset a menu of options and establish a system for measuring their effectiveness. This system needs to integrate financial, organizational, and technical data about the investments. In addition, there should be a rigorous impact evaluation based on quasi-experimental methods to measure the effect on income and livelihoods across options and to a control group. Also, there should be benchmarks for scaling up successful interventions in order to avoid a disruption of Project activities. 88. Interventions that rely on market transactions for expanding land access need to conduct a detailed ex-ante evaluation of land markets conditions and monitor land prices throughout implementation. The ex-ante land market study needs to consider structural conditions, such as legal restrictions and market segmentation, to assess the potential supply of eligible land. This should be a multi-sited study that takes into consideration the heterogeneity of local conditions within a macro region. Projects should also monitor prices on a constant basis to identify market opportunities and report on trends. In this regard, the project could link to or help establish a land observatory, which would contribute to improved capacity to govern land deals. 89. In order to encourage the participation of IFIs, capital endowments should be large enough to offer an acceptable profit margin. In general, IFIs are not interested in lending to poor farmers because of high risk and high costs. The capital endowment of a credit facility minimizes the risk, but IFIs are still concerned about the transaction and administrative costs. The capital endowment should be large enough to ensure that the participating IFI would have sufficient profit margin to participate in the scheme. 90. It is possible to implement productive investments in ways that are compatible with customary land tenure. The land use rights mechanism sets a positive precedent for working with indigenous and peasant communities in the regions of Chaco and Chiquitania with potential to expand to other regions. Since land within the TCOs and TCs cannot be sold or used as collateral, it is often excluded from productive investments. Nonetheless, it is possible to implement productive investments in customary land tenure arrangements. The allocation of land rights and investment funds to PAs can be done without causing internal divisions or conflicts with non-beneficiaries. Those who do not participate in PAs, also stand to benefit from improved economic conditions in their communities. 91. Projects looking to expand land access should also consider modalities based on usufruct rights. Land access projects tend to emphasize land acquisition because the 23 transfer of ownership rights represents a permanent change in the agrarian structure and is supposed to incentivize production. The transfer of usufruct rights, however, could also be a meaningful mechanism to expand land access. Modalities based on the transfer of usufruct rights could be implemented in settings where land transfers are restricted, such as communal lands, or land acquisition is too expensive or undesirable. Particularly for newly established PAs, usufruct rights could provide them with a more cost-effective way of increasing their capital and piloting production schemes. 92. Working through decentralized decision-making bodies helps to mainstream consultation and social monitoring and improve productive potential. Local-based decision-making bodies that incorporate regional stakeholders (municipality, traditional authorities, universities, and private sector) can play a crucial role in the evaluation and monitoring of investment activities. Given their composition, they can perform the function of social monitoring or consultation, for example among projects that include indigenous peoples. In addition, these local entities have the potential to improve productivity capacity because they can facilitate the creation of a support network for PAs and become recipients of local knowledge about regional best practices. 93. PAs should specialize on a few production lines that are well-adapted to the local conditions and have commercialization potential. Models based on an integrative approach to food security could result in multiple production lines that require different kinds of TA. When the sub-projects are geographically dispersed and the supply of service providers is limited, providing high-quality TA could become very expensive or even impossible. For this reason, it is recommended that PAs specialize in one or two production lines. These activities should be selected from those that are suitable for commercialization in local markets and agro-ecological condition. 94. Projects working to improve the productive capacity of vulnerable rural populations should facilitate insertion into markets. In general, poor rural households need to complement their income with off-farm employment in order to meet basic subsistence consumption. In regions like Santa Cruz, it is common for household members to migrate for temporary employment in the agricultural industry. These jobs have precarious working conditions and pay low wages. Investment activities that result in sales and earnings can help to substitute for sub-optimal employment and capitalize the households. In order to insert sub-projects into markets, a commercialization strategy needs to be developed and implemented early in the production cycle. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/implementing agencies6 95. The GOB implemented the Project during a period of significant political and institutional transformation for the country. The Project benefited 2,800 families organized in PAs with expanded access to land, TA, and agricultural inputs contributing to growth in 6 Comments received from the MDRyT (the Implementing Agency) via electronic mail on March 7, 2015. Team’s translation based on document in Spanish available in Project Files. 24 agricultural and livestock production, environmental sustainability, gender equity, and improved organizational capacity. 96. An independent evaluation confirmed that 70 percent of the PA reached the expected outputs in their implementation plans, which resulted in an increase in family income and productive assets. Indirectly, the Project contributed to a decrease in temporal migration, strengthened democracy, promoted social inclusion by processing the identity documents of women and youth who did not have them, and prepared beneficiaries to access other programs financed by the World Bank, such as the Productive Alliances Project, which have more stringent demands for organizational and productive capacities. 97. The Project contributed a valuable model for the implementation of similar projects in Bolivia in the near future. The Project generated important lessons about the mechanism of land use rights in TCOs and TCs, technical and social monitoring, social vulnerability, taxation, and the protocol for livestock acquisition. (b) Cofinanciers N/A (c) Other partners and stakeholders N/A 25 Annex 1. Project Costs and Financing (a) Project Cost by Component (in USD Million equivalent) Actual/Latest Appraisal Estimate Percentage of Components Estimate (USD (USD millions) Appraisal millions) Component 1. Land Purchase 4.86 0.18 4% Component 2. Productive 7.86 11.67 148% Investment Component 3. Project Management, Monitoring and 1.75 4.237 242% Evaluation Total Baseline Cost 14.47 16.08 111% Physical Contingencies 0.00 0.00 0.00 Price Contingencies 1.16 0.00 0.00 Total Project Costs 15.63 16.08 103% Total Financing Required 15.63 16.08 103% (b) Financing Appraisal Actual/Latest Type of Estimate Estimate Percentage of Source of Funds Cofinancing (USD (USD Appraisal millions) millions) Producer Associations Counterpart 0.63 2.19 347% International Development Credit 15.00 13.89 93% Association (IDA) 7 This amount includes technical assistance costs that were charged to this component prior to the creation of a new disbursement category in 2013. 26 Annex 2. Outputs by Component The development objectives of the Project were (a) to establish an alternative decentralized beneficiary-driven land distribution mechanism for organized landless or poor farmers, and (b) support productive investments thereon to allow increased and more sustainable livelihoods for such beneficiaries. The first PDO (land distribution mechanism) was to be achieved through acquisition of private land. However, the land acquisition mechanism faced several challenges: (i) the credit line could not be established; (ii) the supply of eligible land was limited; and (iii) land values significantly increased putting prices beyond the Project’s financing parameters. Nonetheless, the Project was able to achieve the aim of this objective through other alternative mechanisms. The Project Appraisal Document (PAD) recognized that a menu of policies was required to improve access to productive land for landless and land-poor rural populations. Given the inability to launch the land acquisition mechanism, other three additional modalities were piloted during the Project Preparation Advance (PPA). When the Credit became effective, a restructuring was prepared to adapt the Project to the new circumstances. The first restructuring changed the two-tier design of the Land Credit Facility and introduced three additional mechanisms for land access: (i) Leasing Agreements (Arrendamiento). Under this mechanism, a land owner agreed to temporarily transfer the use and enjoyment of the land for a specific time period in exchange for payment. Under the project, this was reflected in a written contractual agreement between a PA and the land owner(s). This type of contract is well-recognized in Bolivian land law. For leasing transactions financed by the Project, the land had to be registered in the Real Property Registry in the name of the owner who signed the contract, and the contract had to stipulate the rented area, the time period of the rental agreement and the payment(s) required. (ii) Share-cropping Agreements (Derechos de Uso de Suelo) Through this equity- sharing mechanism, the owner of a land parcel gave use of the land to a PA in exchange for a percentage of its equity and a negotiated amount of the production. This is done through a contract which is well-developed in Bolivian jurisprudence. For this type of contract to be acceptable under the Project, the property had to be registered in the name of the owner who signed the contract or the owner’s legal delegate. Moreover, the contract had to specify the physical area and location of the parcel(s) in the transaction, the time duration of the agreement, the share of equity in the enterprise assumed by the land owner, and the specific rights and obligations of the land owner and the PA. (iii) Land use rights (Aparceria Comunal) This mechanism was used only for PAs whose members are part of a TCOs and TC, which are both collective properties recognized by the 1996 agrarian reform law (Law No. 1715). Specifically, agreements were signed between the PA and the communal authorities in charge of land administration in the TCO or TC stipulating that the association was entitled to the long-term use of a determined land area in exchange 27 for compensation paid to the community. In this sense, the mechanism was equivalent to a long-term lease contract within a communal landholding situation. This access was free and for an indeterminate period of time, but according to customary rules, the land might be required by the authorities in case the PA is not using it properly. The location of the land area to be used, the duration of the use agreement and the detailed arrangements of the compensation plan were spelled out in the agreement. The legal basis for this contract is found in the Constitution and in the 1996 agrarian reform law which provide for communal authorities to determine land use and land allocation within their communal land boundaries for community members according to their own internal regulations and customs. Evolving policy environment. The Project took place during a period of significant political and institutional transformations for the Bolivia. The timeline below records some of the key dates for political and Project-specific events. Year Political Events Project-specific Events 2002 GOB submits request for Project preparation (August) 2003 Gas protests (September) President Gonzalo Sanchez de Lozada resigns and Carlos Mesa assumes office (October) 2004 GOB ratified request for Project Preparation (January) Project Concept Review (March) 2005 Gas protests (March-June) Project Appraisal (April) PPA is declared effective (June) New Decentralization Law is approved (April) Hydrocarbons Law is approved (May) President Carlos Mesa resigns and Eduardo Rodriguez assumes office (June) Evo Morales wins election (December) 2006 Evo Morales assumes office PCU is established (November) (January) Hydrocarbons Nationalization Law is approved (May) 28 National Development Plan is approved (November) Law 3545 for targeting agrarian reform to indigenous peoples and peasants is approved (November) 2007 First PAs are conformed Board Approval (October) 2008 Autonomy referendum is held in PPA Agreement is amended (May) Santa Cruz (May) Three alternative mechanisms are piloted New Constitution is approved under the PPA (December) 2009 President Evo Morales is re- Credit is declared effective (March) elected (December) 2010 First Project Restructuring is approved (April) 2011 First legal amendment is declared effective (March) 2012 Mid-Term Evaluation (February) Second Project Restructuring is approved (June) Project area expands to Concepcion and San Ignacio (August) Cooperative San Gabriel reports liquidity problem 2013 Third Project Restructuring is approved (January) Fourth Project Restructuring is approved (November) 2014 President Evo Morales is Fifth Project Restructuring is approved reelected (December) (May) Project closes (September) Second and Third Restructuring. The Mid-Term Review (MTR) was conducted in February 2012. The MTR confirmed that the Project was still relevant and recommended restructuring to ensure that its PDOs were achieved. The restructuring was divided into two parts. The first part of the restructuring was approved in May 2012 and extended the closing date from June 29, 2012 to March 21, 2014. The second part was approved in January 24, 2013 and reformulated the Results Matrix, established a new disbursement category, 29 reallocated Credit proceeds, and recognized the incorporation of the municipalities of Concepcion and San Ignacio de Velasco into the Project area. The specific changes to the Results Matrix are detailed in Attachment 1 below. Fourth and Fifth Restructuring. A fourth restructuring was approved in November 2013 to extend the closing date to June 15, 2014. A fifth restructuring was approved in May 2014 to extend the closing date to September 15, 2014. The cumulative period of extensions was 2.5 years. Project Operational Manual. The Project Operational Manual (POM) established the selection criteria for beneficiaries and land. The POM was amended in 2012 to reflect the first restructuring. The selection criteria established that all land, private or communal, had to be regularized and titled. For TCOs and TCs, the communal authorities had to provide permits and other community members had to agree with the land allocations. At the time of the amendment in 2011, the rules required payment to the community for the use of land. After the MTR, the communal authorities accepted to drop this requirement. For the most part, the selection criteria were complied with, which minimized the reputation risk for the Project and the Bank. The exception was the PCU’s decision to approve sub-projects for three PAs located in the municipality of Warnes. The members of these PAs came from a TCO in Charagua, but the sub-projects were outside of the Project area and were not approved by the CODALs. These sub-projects were cancelled. The outputs by component are explained below. Component 1: Land Acquisition The aim of this component was the financing of land purchase by associations of low- income rural workers. This financing was to be made in the form of long-term loans to PAs through a rotating fund establishing by the Project. The land acquisition mechanism did not operate during the Project. The land use rights mechanism in TCOs and TCs was the most important instrument to increase land access and reach the target population. PAs benefited from the opportunity of using land within their own territories. For the Project, the land use right mechanism provided conditions for long-term land access and the entire territory benefits from the investments on infrastructure. In contrast, land owners had the discretion to set the terms for the duration of contract or agreement under the leasing and sharecropping mechanisms, and fixed investments remained under land owner control. The Project facilitated access to 151,579 ha of approved land, out of which 132,301 ha (87.3%) were located in TCOs, 18,851 ha (12.2%) in TCs, and 827 ha (0.5%) in private lands. Total land allocation exceeded the end target of 115,000 ha. The average amount of allocated land was 52 ha per beneficiary family and 640 ha per PA. By municipality, approved land was distributed 84.4 percent in Charagua, 8.0 percent in San Ignacio de Velasco, 5 percent in Pailon, 2.2 percent in Concepcion, and 0.4 percent in Minero. Out of the 151,579 ha of approved land, 88,686 ha or 59 percent of total were under agricultural production or livestock grazing. This value was below the end target of 80 30 percent. Percentage of land under production was higher for agricultural than livestock sub- projects. Livestock sub-projects included more than in anticipation of future growth and pasture rotation. By municipality, the percentage of land under production was 64.9 percent in Charagua, 56.7 percent in Minero, 32.4 percent in San Ignacio de Velasco, 28.2 percent in Concepcion, and 6.6 percent in Pailon. Between 2007 and 2009, CODALs approved 25,940 ha with PPA financing. In 2010, the Project did not approve any requests because the legal amendment was under revision. The years with more approvals were 2011 and 2012 with 69,023 and 56,615 respectively. No more requests were approved after 2012 because there would not be sufficient time for PAs to finalize their investments by project closing. The Project helped 2,891 beneficiary families gain access to land. The achieved value is higher than the end target of 2,200. This corresponds to 11,488 direct project beneficiaries.8 The TCOs in Charagua are predominately Guarani and TCO and the TCOs and TCs of San Ignacio de Velasco and Concepcion are predominately Chiquitano (besire). In Pailon and Minero, the most beneficiaries were of Andean origin. The Project helped establish 237 PAs, surpassing the end target of 190. The distribution of PAs by land tenure was: 138 in TCOs, 82 in TCs, and 17 in private lands. By municipality, 127 were in Charagua, 47 in San Ignacio de Velasco, 30 in Concepcion, 19 in Pailon, and 14 in Minero. 27 PAs or 11 percent of total PAs were suspended, 11 were in Charagua, 9 in Minero, 4 in Pailon, and 3 in Concepcion. There were no suspended PAs in San Ignacio de Velasco. The main reasons for suspension were internal administrative problems and lack of interest among members to continue with the sub-project. The number of suspended PAs was disproportionally higher in Minero, some of the first to start operations. In terms of administrative duties, women held 32 percent of board member positions and held the presidency in 8 percent of PAs. Since the Project did not finance any land acquisition, actual cost of this component was only US$ 0.18 million or 4 percent of appraisal estimate. Actual expenditures were used to finance consultant services. Counterpart funds covered payments in leasing contracts. 8 Estimate based on original assumption of 10 households per PA, and an average of 5 people per household. 31 Land Use Rights in TCOs Land Use Rights in TCs Leasing/Sharecropping Municipality PAs Approved In Use PAs Approved In Use PAs Approved In Use (ha) (ha) (ha) (ha) (ha) (ha) Charagua 127 128,001 83,011 0 0 0 0 0 0 Concepción 7 700 177 23 2,615 757 0 0 0 Minero 0 0 0 3 200 16 11 362 303 Pailón 0 0 0 13 7,136 327 6 465 176 San Ignacio 4 3,600 1,200 43 8,500 2,720 0 0 0 Total 138 132,301 84,387 82 18,451 3,820 17 827 479 % of Total 58.2% 87.3% 95.2% 34.6% 12.2% 4.3% 7.2% 0.5% 0.5% Source: SAX Independent Evaluation with data from the PCU. Component 2: Productive Investments This component supported each of the 237 PAs with productive investments through grants and technical assistance. The sub-projects were 54 percent for bovine livestock, 15 percent other livestock, 12 percent for agriculture, 11 percent agro-forestry, and 8 percent other (aquaculture, milk production, poultry, and horticulture). Bovine livestock sub-projects are concentrated in Charagua, Concepcion, and San Ignacio de Velasco. Minero has the most diversity of sub-projects with investments in agriculture, poultry, milk production, and porcine livestock. Pailon has investments in agriculture, milk production, and bovine livestock. Since 2012, the PCU monitored the physical and financial progress of sub-projects’ Investment Plans (IPs) using a scale of 1 to 4. A score of 1 represents less than 25 percent of progress, 2 is from 25 percent to 50 percent progress, 3 is 50 percent to 75 percent progress, and 4 is more than 75 percent progress. When the final evaluation was prepared in July 2014, the sub-projects in Charagua and Pailon had an average rate of 3 and the sub- projects in San Ignacio de Velasco and Concepcion had an average of 4. Some of the delays in implementation were the result of the liquidity problem experienced by the Cooperative San Gabriel. Some PAs experienced delays in accessing funds to implement their sub projects. Sub-projects in Minero were not rated using this scale because investments had already closed in this municipality by 2012. Although San Ignacio de Velasco and Concepcion were the last two municipalities to be included in the Project, their sub-projects had the best operational performance. On average, financial execution lagged by 7 percent from physical execution, which according to the independent final evaluation is an acceptable rate for this type of investments. An external financial and economic analysis provided key measurements for evaluation based on a sample of 83 PAs. According to this study, the investment sub-projects increased family income by 54 percent compared to baseline, surpassing the revised target of 20 percent and the original target of 50 percent. Family income increased for all sub- projects types and in all municipalities, but there was a wide range of values. The sub- projects in Minero were in the lower end with a 16 percent increase compared to the sub- projects in Pailon which reported a 96 percent increase. Another target was to increase the value of productive assets by at least 15 percent. For agricultural investments, family assets increased by US$122 or a 15 percent increase compared to baseline. For livestock investments, family assets increased by US$ 1,435 or a 614 percent compared to baseline. 32 This exorbitant increase was made possible by the acquisition of livestock from grant financing. If the cost of livestock were to be included, families would have had an asset loss of US$ 78. This underperformance is in part explained by the drought events that affected Charagua in 2012-2013 and the slow breeding cycles that delay the economic return of cattle investments in the Chaco region. In a way, it is still early in the productive cycle to measure the impact on livestock assets. The Project trained 3,981 beneficiaries in 2012 and 2013. The number of participants covered 100 percent of beneficiaries in Concepcion, 82 percent in San Ignacio de Velasco, 61 percent in Charagua, and 51 percent in Pailon. However, some beneficiaries participated in more than one training event. The exact number of beneficiaries that participated in at least one training event is not known. Women accounted for 38 percent of members participating in training events. This value was higher than the end target of 20 percent. Charagua had the lowest record of women’s participation in training event with only 33 percent. In Concepcion, Pailon, and San Jose de Velasco women’s participation was around 41 percent. The actual cost of this component was US$ 11.67 million or 148 percent of appraisal estimate. The higher cost represents reallocation of Credit proceedings from Component 1 to investment grants and technical assistance as well as larger contributions from counterpart funds. Execution Progress Rating Municipality APs Average Physical Financial Physical Financial Charagua 128 82% 64% 3 3 3 Pailón 19 70% 67% 3 3 3 San Ignacio de V. 47 96% 91% 4 4 4 Concepción 30 94% 90% 4 4 4 Total 224 85% 78% 3 3 3 Source: SAX Final Evaluation with data from the PCU Component 3: Project Management, Monitoring and Evaluation This component financed the coordination, administration, financial management, and M&E of the Project. The Project was managed by a PCU based in the city of Santa Cruz. The PCU was staffed with technical, fiduciary, and monitoring personnel. Local operational units were established in each municipality to provide oversight to the sub- projects. During implementation, the PCU opened a specialized unit to provide technical assistance (TA) to PAs. The PCU made this decision after considering that there were not enough service providers to assist PAs. Project management faced several challenges. In 2010, many staff members left the PCU as a result of uncertainty about the future of the Project. The PCU was re-staffed in 2011 and project management improved. However, these circumstances affected operations. The Project maintained adequate procurement capacity, but financial management had shortcomings. PAs received the assistance of accountants to help with financial management, but accounts were not always fully reconciled. The Project’s M&E system 33 was significantly improved in 2011 when information about the physical and financial execution of sub-projects was integrated. The M&E system was maintained in Microsoft Excel spreadsheets, which was effective for the preparation of reports and sub-project oversight. Also, this component financed consulting services to assist MDRyT, through the Vice- Ministry of Lands and Rural Development, in the preparation of a new land access policy. The actual cost of this component was US$ 4.23 million or 242 percent of appraisal estimate. One reason for this increase is that during the early implementation period technical assistance costs were charged to this component. This situation was fixed when the Project was restructured to create a new category for technical assistance. However, not all payments were reposted. 34 Attachment 1 – Changes to the Results Matrix Project Outcome Indicators Core D=Dropped Baseline (Actual at the time Cumulative Target Values* C=Continued of restructuring) N= New YR 1 YR2 YR 3 YR4 YR5 R=Revised UoM Value [1] Direct Project beneficiaries N Number 8,100 1,000 3,000 8,500 10,000 11,000 X [1(a)] Female beneficiaries Number 4,130 550 1,600 4,335 5,100 5,610 [2] At least 20% increase in R Percentage 16 -- -- 16 -- 20 beneficiary family income [3] Local level decision-making and R Number 5 3 3 3 5 5 monitoring entities established in participating municipalities [4] At least 15% increase in value of N Percentage -- -- -- -- 12 15 productive assets of beneficiary families [5] At least 10% of Productive N Percentage -- -- -- -- 8 10 Associations (PAs) have women in their Boards Monitoring system to assess the net D Number n/a n/a n/a n/a n/a n/a economic benefits of increased agricultural production established 120 farms selected and acquired by D Number n/a n/a n/a n/a n/a n/a Productive Associations (PAs) via new mechanisms Intermediate Results Intermediate Result (Component one): Land access provided to beneficiaries Revised Intermediate Result (Component one): n/a 1.1 At least 2,200 beneficiary R Number 2,000 200 600 1,700 2,000 2,200 families access land through Productive Associations (PAs) 35 1.1 (a) At least 190 PAs established N Number 180 40 100 150 180 190 1.2 At least 115,000 hectares R Hectare 90,000 25,000 45,000 90,000 100,000 115,000 accessed by PAs Intermediate Result (Component Two): Productive investments provided to and carried out by beneficiaries Revised Intermediate Result (Component Two): n/a 2.1 At least 70% of the productive N Percentage -- -- -- -- 70 70 sub-projects comply satisfactorily with agreed implementation plans 2.2 At least 70% of land accessed by N Percentage -- -- -- 55 60 70 PAs are put under productive use 2.3 At least 80% of PAs are satisfied N Percentage -- -- -- -- 80 80 with the technical assistance provided by the Project 2.4 At least 20% of participants in N Percentage 15 -- -- -- 15 20 training sessions are women At least US$8.0 million in total D Amount n/a n/a n/a n/a n/a n/a value of productive investments (USD) At least 90% of production D Percentage n/a n/a n/a n/a n/a n/a investment subprojects are completed Intermediate Result (Component Three): Project managed and evaluated adequately Revised Intermediate Result (Component Three): n/a 3.1 Project M&E system established N Number 2 1 1 2 2 3 and updated periodically, and Mid- term and final evaluation completed 36 Annex 3. Economic and Financial Analysis Background: Through financing from the World Bank of a credit of SDR 9.80 million (US$ 15 million equivalent) to the Republic of Bolivia, the Ministry of Rural Development and Land (MDRyT, in Spanish abbreviations) implemented the Bolivia Land Agricultural Development Project (PROTIERRAS, in Spanish abbreviations). The Project partially helped to establish a decentralized beneficiary-driven land distribution mechanism, while supporting productive investments that helped to improve the livelihoods of beneficiaries. Through the Project financing, mainly productive goods, works and services in the form of matching grants, Productive Associations (PAs) were empowered. Overall impact of the project: As a result of the productive investments financed by the Project, the main economic impacts have been improved farmer incomes, primarily due to increase in production and/or productivity (yield). The Project has also had social impacts by improved land access to the families of organized landless and poor farmers. Many of these families are of indigenous descent. The benefits of producer organizations empowerment, capacity building and technical assistance are difficult to measure economically, but it is expected that they will translate into increased productivity. Project benefits and outputs: The Project has helped to establish 237 Productive Associations (PA), benefitting approximately 2,900 mainly poor indigenous families and incorporating approximately 151,000 hectares of land for productive uses. The Project strengthened the APs organization and co-financed productive investment sub-projects on matching grants basis, which have helped to improve the livelihoods of beneficiaries. According to findings, beneficiary households' income has increased by over 39 percent and their assets by 15 percent for agricultural investments and by US$ 1,437 for livestock investments (more than six times compared to baseline, mainly due to the acquisition of livestock from grant financing). Project costs and financing: As shown in Annex 1, the actual total Project cost was US$ 16.08 million, 103 percent of the appraisal estimate. The IDA contributed US$ 13.89 million, 92.56 percent of the appraisal estimates, and the producers associations’ contribution was US$ 2.19 million, 347 percent of the appraisal. Activities under Component 2, Grant Investments subprojects, accounted for the majority of resources invested, 72.57 percent of actual total costs, compared to Component 1 (Land credit facility), 1.11 percent, and Component 3 (Project Management), 26.30 percent. Technical assistance, training and other operating cost for investment subprojects under Component 2 represented 6.53 percent of the actual total Project costs. Subproject cycle and eligible investments: The Project supported PAs through subprojects contained Investment Plans. The subproject cycle involved four stages: (1) pre- investment stage (producers were accompanied from their organization to the preparation of their sub-projects); (2) execution stage (including procurement of equipment and of technical assistance); (3) operation stage and (4) monitoring stage (assessing at the closure and measurement of the effects on their production and commercialization levels). Through Component 2, the Project financed 237 demand-driven productive investment sub- 37 projects, included the following: (i) agricultural and livestock infrastructure, 43.04 percent of the total investment; (ii) livestock purchase, 33.97 percent; (iii) equipment and tools, 8.69 percent; (iv) agricultural and livestock inputs, 7.31 percent; and others (environment requirements, administrative and operating costs), 6.98 percent. The productive sub- sectors9 more largely supported by the investment were, as follow: (i) cattle (42%); (ii) agricultural crops and cattle (24%); (iii) alternatives crops (18%); (iv) agricultural crops (7%); (v) cattle double purpose (3.5%); (vi) goats (3.5%); (vii) wool sheep (3%); (viii) milk (1%). Target regions: The Project supported producers associations and its investments subprojects in five selected municipalities of Santa Cruz Department, namely Charagua, Minero, Pailon, Concepcion, and San Ignacio de Velasco (the last two were added to the Project area in the Project restructuration). The largest amount of investment was located in Charagua municipality (53.7%), follow by San Ignacio de Velasco (25%), Concepción (10.8%), Pailon (5.8%) and Minero (4.8%). Ex-Ante Economic and Financial Analysis Ex-ante economic and financial analysis at appraisal: Due to the demand-driven nature of the Project, it was not known ex-ante which specific investment subprojects would be financed, and a precise economic and financial analysis for the Project as a whole could not be made. The cost benefit analysis conducted during Project preparation, as presented in the Project Appraisal Document (PAD), was based on four indicative models of potential subprojects which were to be implemented as follow: (i) sugar cane in Minero; (ii) soy and corn in Minero; (iii) sesame and garbanzo production in Pailon; and (iv) cattle project in the Charagua. These models were based on land acquisition through a rotating fund capitalized by the Bank loan, 15 year land purchase repayment periods with annual interest rates of 8 percent. In the models a one-year grace period was assumed, except for the cattle model which required a three-year grace period. Group size was assumed to be twenty families. Foregone family labor income in other activities was not considered because the assumption was made that the agricultural project will be managed by two-three individuals receiving wages. With plausible assumptions about the number and distribution of these stylized models by municipality, the aggregate Net Present Value (NPV) of all investment subprojects was $17.65 million and the Internal Rate of Return (IRR) of 16 percent. It is not possible to compare the ex-ante with the ex-post evaluation’ results, as recalculation of the NPV and IRR at completion was not possible due to the following reasons during Project implementation: (1) factors such as land market conditions made it impossible to utilize land access mechanisms available under the Project, and therefore the unavailability of the models and assumptions used at appraisal, and (2) productive systems 9 Productive systems: (1) Cattle; (2) agricultural crops and livestock (cattle); (3) alternative crops (chia, sesame and peanuts); (4) agricultural crops; (5) livestock (dual purpose); (6) goats; (7) Sheep; (8) milk; (9) pork; (10) agricultural crops and fish; (11) poultry; (12) agricultural crops and poultry 38 benefited by the investment subprojects were different for those selected at appraisal, and the number and distribution of the actual investments by municipalities finally benefited from the Project. Ex-post Economic and Financial Analysis At project completion, and for the purpose of the ICR, a detailed ex-post economic and financial analysis was prepared by an independent firm. The evaluation estimates the returns derived from the productive investments subprojects, which was the bulk of the resources invested (72.57%) in relation with the total resources devoted to land credit facility (1.11%), project management (26.30%).The benefits of technical assistance and training to producers’ organizations for the implementation of the subprojects were not considered in the analysis because they are more difficult to measure economically, though definitely will translate into increased productivity and production in the medium and long term. Methodology of the Ex-Post Analysis A universe of 206 subprojects, which had completed physical execution and whose productive sub-sectors had benefited at least of 1 percent of the project investment, was defined. The universe of 206 subprojects represented 87 percent of the approved investments. Once the universe was defined, a sample of 83 sub-projects was randomly selected from predefined strata (productive sub-sectors and municipality) to conduct an ex- post economic and financial analysis, and infer the results to the whole universe. The sample of subprojects represents 38 percent of the approved investments. The analysis of ex-post costs and benefits of the 83 producers’ subprojects was based on the Project’s M&E system, which contains primary information on the with and without project situation, performance of the investments and profitability of the PA. To validate the data, field visits were conducted to some subprojects. Financial and Economic parameters were estimated, among others: Net Present Value (NPV), at a 12.67 percent annual discount rate; Internal Rate of Return (IRR); ratio Benefits / Cost (B/C). Results and returns of the investments were analyzed at subproject level, aggregated by municipality, by productive sub-sector level, and at Project as a whole. Within each sub- sector, economic parameters and indicators are estimated for each PA as well as for average PA and beneficiary family. To estimate aggregate returns at sub-sector level, the incremental net income flows for the total number of PA in the analyzed sub-sector is added, and relevant aggregated indicators calculated. To estimate aggregated impacts at Project level, the aggregated net income flows for each sub-sector analyzed is added and aggregated feasibility indicators are estimated for three basic scenarios: (i) considering only the direct investment costs; (ii) including costs of component 1 and 3 proportionate to the investment; and (iii) including all Project costs. The cost-benefit analysis builds on a ten-year projection period after the last investments have been carried out. As such, the reference period is 2007-2024 (with last investments 39 occurring in 2014). As the Project has collected cycle of production and sales from 2008 to 2013, an average of those years is used to estimate the projected fund flows. The investment includes the funds transferred by the Project to the PAs plus their counterpart. To make a comparative analysis, current prices are converted into constant prices using the GDP Deflator (Source: INE, Agriculture, Livestock, Hunting and Fishing). The analysis focused on financial flow, applying market (private) prices, and on economic flow that applies economic (social) or shadow prices based on a Conversation Factor of 0.885 for products, inputs and services (shadow prices calculated by deducting the 13 percent Value Added Tax (VAT) from the unit price (1/1.13=0.88) and 0.47 for non- qualified labor (shadow price established by the National Evaluation System for Public Investment in Bolivia). For sensitivity analysis, reduction of income level to account of climate factors, mainly drought (technical or social/organizational aspects) and were introduced in the different analysis. Aggregate and Average Project Indicators As shown Table 1, the Financial IRR (FIRR) and Financial NPV (FNPV) of the analyzed subprojects are 18.3 percent and US$ 1.6 million respectively. Considering other Project costs (project management proportionate to the investment analyzed), the FIRR and FNPV are 14.9 percent and US$ 0.8 million respectively. Pailón municipality, where the investment has been low compared to other municipalities, presents the highest FIRR compared to the other municipalities of intervention, and reaches 26.9 percent. . The FIRR and FNPV in San Ignacio de Velasco is 25 percent and US$ 1.3 million respectively, and the Charagua municipality presents the lowest FIRR in relation to other municipalities, reached 13.1 percent. The Economic IRR and NPV of the analyzed sub-projects are 21.76 percent and US$ 2.5 million, respectively. Charagua presents the lowest EIRR respect to other municipalities, but the EIRR for all municipalities are above the social discount rate set by the VIPFE at 12.67 percent, which means that the project is socially profitable in all municipalities and therefore public investment in the project is justified. Minero municipality is the municipality that has the highest EIRR, 34.4 percent, although its ENPV is low compared to other municipalities (US$ 0.12 million); and is followed by San Ignacio de Velasco with an EIRR of 28.7% and the highest ENPV, reaching US$ 1.58 million, and Pailon (EIRR 28.7% and ENPV US$ 0.15 million) 40 Table 1: Summary of Aggregated Evaluation Indicators for the Analyzed Sub-projects, by Municipality Financial Analysis Economic Analysis FIRR NPV B/C EIRR NPV B/C (%) (US$ (%) (US$ million) million) Charagua 13.1% 0.054 2.48 16.5% 0.461 3.25 Concepcion 15.4% 0.114 2.74 17.3% 0.192 3.32 Minero 17.6% 0.02 1.66 34.4% 0.121 2.04 Pailon 26.9% 0.132 2.23 28.7% 0.157 3.03 San Ignacio de 25.0% 1.338 2.80 28.7% 1.585 3.29 Velasco TOTAL 18.3% 1.669 2.57 21.76% 2.518 3.22 The analysis of the seven most representative sub-sectors in term of received investments shows the following results: Table 2: Summary of Aggregated Evaluation Indicators for the Analyzed Subprojects, by Productive Subsector Financial Analysis Economic Analysis FIRR NPV B/C EIRR NPV B/C (%) (US$ (%) (US$ million) million) Cattle (Meat) 11.8% (0.10) 2.80 13.0% 0.03 3.33 Agricultural crops and cattle 17.3% 0.19 2.23 23.9% 0.45 2.36 Alternative crops 36.7% 1.49 2.23 44.0% 1.74 2.56 Agricultural crops 12.1% (0.004) 1.27 22.9% 0.08 1.44 Cattle (dual purpose) 13.2% 0.04 2.64 15.6% 0.22 3.96 Goat 23.6% 0.15 4.14 22.8% 0.14 3.72 Sheep 15.4% 0.03 2.82 15.4% 0.03 3.07 In the agriculture sector, the alternative crops subprojects (sesame, chia, peanuts) present the highest return on all subsectors financed by the Project, and have a FIRR of 36.7 percent and an EIRR of 44 percent. However, the agricultural crops subprojects (corn, beans and sorghum) present the lowest FIRR of 12.1 percent and EIRR of 22.9 percent. These subprojects, mostly located in Charagua municipality, show an FIRR lower than the discount rate. This is explained by the unfavorable climatic conditions prevailing in Charagua, drought has prevented the development of a greater amount of crops in this region and has reduced yields. As for PA's livestock sub-projects show potential for positive profitability indicators along the horizon of analysis. However, draft cattle breeding has lower rates compared to other livestock sub-sectors’ returns. This is due to two factors: (i) the investment needed to start these subprojects is high, especially in enabling infrastructure and facilities, and purchase of livestock; and (ii) once the start-up investment is executed, these subprojects require a minimum of two years to start generating revenues; therefore the cash flow is negative during the first 2-4 years. These two factors explain the lower profitability indicators, 41 although this do not imply that subprojects breeding cattle are not positive, as shown by the economic return and potential revenue generated by these PA's and by their families members. Another aspect to be highlighted is that larger livestock herds (more than 70 productive heads) become highly profitable, as economy of scale is generated and justify the initial investment in infrastructure and facilities depending on the size of the herd. Sensitivity analysis to include the effects of drought on the analyzed subprojects has been developed by assuming that: (i) The agricultural PA from Charagua have two good years and two bad years of production, and in bad years they recover only 50 percent of the expected production; (ii) The livestock PA from Charagua have lower calving rates and higher mortality rates compared to municipalities that do not suffer the effects of drought as often. These assumptions are consistent and have been validated with technical staff of the Project and through field visits. The sensitivity of the financial performance of sub-projects in Charagua respect to the effects of drought is high. In a scenario that takes into account the effects of drought, the IRR in Charagua will decrease from 13.1 percent to 9.2 percent, and the NPV would fall from US$ 0.054 to a negative amount. However, the overall Project returns would still be positive, but would fall by 1.1 percent falling from 18.3 percent to 17.2 percent. Conclusions and Recommendations The overall conclusions of the evaluation are as follow:  The results of the ex-post cost-benefit analysis show a satisfactory overall performance of the implementation of the productive investment subprojects analyzed, an aggregated Economic IRR of 21.76 percent (considering only the direct investment cost of analyzed alliances). The Financial IRR of 18.3 percent (considering only direct investment cost of the analyzed alliances) and 14.9 percent (including project management costs proportionate to the investment analyzed), respectively, are above the opportunity cost of the investment, stated around 12.67 percent.  The Project investments were in order of importance (i) agricultural and livestock infrastructure, 43.04 percent of the total investment; (ii) livestock purchase, 33.97 percent; (iii) equipment and tools, 8.69 percent; (iv) agricultural and livestock inputs, 7.31 percent; and others (environment requirements, administrative and operating costs), 6.98 percent.  The productive sub-sectors more largely supported by the investment were: (i) cattle (42%); (ii) agricultural crops and cattle (24%); (iii) alternatives crops (18%); (iv) agricultural crops (7%); (v) cattle double purpose (3.5%); (vi) goats (3.5%); (vii) hair sheep (3%); (viii) milk (1%). 42  More than the half of the investment was located in Charagua and about 70 percent of total investments were directed to subprojects for cattle. However Charagua presents the lowest IRR in relation to other municipalities, reached 13.1 percent. This is explained by the following main reasons: (1) lower incremental returns from investments in cattle (meat and dairy production) due to the longer gestation period of these investments; (2) climatic factors affected the agricultural crops subprojects (corn, beans and sorghum) performance, e.g. the drought faced by Charagua during Project implementation prevented the development of a greater volume of crops in this region and reduced yields; (3) productive associations show to have less entrepreneurial skills than their counterparts in the municipalities of Conception and San Ignacio de Velasco.  Out of the seven most representative sub-sectors, alternative crops stand out as the subsector generating the highest incremental returns at Project level, followed by goat and then by the subprojects with a mix system producing agricultural crops and cattle also demonstrated profitability, homogeneously around 17 percent for the financial rate of return.  Cattle subprojects require high investments in infrastructure, facilities, acquisition of livestock enabling environment, pastures and also need at least a period of two years from investment to start generating income (two years to enter the market). These are the main reasons why the financial evaluation for this item is the lower.  The sensitivity of the financial performance of productive Investments in Charagua respect to the effects of drought is very high. Therefore, provision of water (water sources and distribution mechanism) is essential to guarantee productive development in Charagua. 43 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Responsibility/ Names Title Unit Specialty Lending Malcolm Childress Sr. Land Admin. Specialist LCSAR TTL Jorge A. Muñoz Sr. Land Admin. Specialist LCSAR Land Policy George Ledec Sr. Environmental Specialist LCSEN Environment Alonso Zarzar Sr. Social Development Specialist LCSEO Social David Tuchscheider Sr. Rural Development Specialist LCSAR Rural Development Patricia McKenzie Sr. Financial Specialist LCSFM Financial Mgmt. Lourdes Linares Financial Specialist LCSFM Financial Mgmt. Luis Schwarz Financial Specialist LCSFM Financial Mgmt. Miram Cespedes Procurement Officer LCSPT Procurement Alvaro Larrea Procurement Specialist LCSPT Procurement Credit and Private Michael Goldberg Financial Sector Specialist LCOPR Sector Dev. Solange Alialli Lawyer LEGLA Legal Issues Fabiola Altimari Lawyer LEGLA Legal Issues Rolande Pryce Lawyer LEGLA Legal Issues Operations and Teresa Roncal Operations Analyst LCSAR Planning Leila Diana Sarquis Project Assistant LCSAR Program Support Supervision/ICR Malcolm Childress Sr. Land Admin. Specialist GURDR Former TTL Enrique Pantoja Sr. Land Admin. Specialist GSURR TTL Operations and Teresa M. Roncal Operations Analyst GFADR Planning Maria Ruth Llanos Vda De Social Safeguards Consultant GSURR Navarro and Gender George Campos Ledec Lead Ecologist GENDR Environment Stamatis Kotouzas Junior Professional Associate GSURR Land Policy Marco Antonio Zambrano Consultant GENDR Environment Chavez Francisco Javier Obreque Consultant GFADR Rural Development Arqueros Lara Chinarro Consultant Agriculture and GSURR Livestock Carlos Francisco Siezar Consultant Credit and Private GFADR Sector Dev. Maria Manuela Maria Consultant GSURR Communication Maria Lucy Giraldo Consultant GGODR Procurement Jose Yukio Rasmussen Senior Procurement Specialist GGODR Procurement Kuroiwa 44 Keisgner De Jesus Alfaro Senior Procurement Specialist GGODR Procurement Patricia De la Fuente Hoyes Sr Financial Management Specialist GGODR Financial Mgmt. Lourdes Consuelo Linares Sr Financial Management Specialist GGODR Financial Mgmt. Loza Juan Marcelo Berthin Financial Management Financial GGODR Heredia Specialist Management Fabiola Altimari Montiel Senior Counsel LEGLE Legal Issues Ketty Morales Program Assistant LCSAR Program Support Miriam Cespedes Program Assistant GGODR Program Support Beatriz Elena Franco Program Assistant LCC1A Program Support Patricia Gutierrez Program Assistant LCCBO Program Support b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) USD Thousands Stage of Project Cycle No. of staff weeks (including travel and consultant costs) Lending FY05 45.43 206.75 FY06 15.94 69.87 FY07 18.33 73.25 FY08 12.59 65.75 Total: 92.29 415.62 Supervision/ICR FY08 20.74 105.53 FY09 35.25 146.66 FY10 34.67 164.65 FY11 31.55 145.74 FY12 21.58 106.67 FY13 19.41 110.19 FY14 13.11 102.02 FY15 6.55 52.43 Total: 182.86 933.69 45 Annex 5. Beneficiary Survey Results The independent final evaluation included a beneficiary satisfaction survey with the Project. The survey asked questions about participation, communication, benefits, technical assistance (TA), and functioning and support received by the Productive Association (PA). The survey was conducted as part of the field validation exercise with PAs. The firm selected 36 PAs based on a representative sample that stratified PAs by municipality and type of investment. The firm conducted the satisfaction survey among the PA associates who were present during the field visit. This strategy resulted in 58 surveys, distributed as follow: 23 respondents from Charagua, 9 from Concepcion, 4 from Minero, 7 from Pailon, and 15 from San Ignacio de Velasco. The communication strategy was considered a crucial starting point to ensure the beneficiary-driven nature of the Project. The survey asked respondents about the first time that they learned about the Project. The main sources of information were municipal projects and programs (32.8%), communal assembly (22.4%), and neighbors (20.7%). Umbrella organizations and media were not a significant source of information for beneficiaries. Table 2: Source of Information about the Project % of Responses Municipality Responses Umbrella Communal Local Project Other Media Neighbor Organization Assembly Authority staff Program Charagua 23 0.0% 34.8% 0.0% 0.0% 8.7% 0.0% 56.5% Concepción 9 22.2% 11.1% 0.0% 0.0% 0.0% 11.1% 55.6% Minero 4 0.0% 0.0% 0.0% 0.0% 100.0% 0.0% 0.0% Pailón 7 0.0% 42.9% 0.0% 0.0% 57.1% 0.0% 0.0% San Ignacio 15 13.3% 6.7% 26.7% 6.7% 13.3% 26.7% 6.7% Average 6.9% 22.4% 6.9% 1.7% 20.7% 8.6% 32.8% After learning about the Project, 52 percent decided to join it during a communal assembly and 48 percent during a PA meeting. In Charagua, Concepcion, and San Ignacio the Velasco more than half of beneficiaries decided to participate in the Project during a communal assembly is a reflection of its inclusive approach and adherence to customary practices. Once beneficiaries joined a PA, the next step was the selection of an investment for the sub-project. In Minero, Pailon, and San Ignacio de Velasco, all investment sub-projects were selected by the members of the PAs. In Concepcion, 78 percent were selected by the PAs and 22 percent by the communal assembly. In Charagua, 48 percent were selected by the communal assembly, 26 percent by the PAs, and 17 percent by another collective mechanism such as a communal authority. The main reason for joining the Project was to improve income, followed by a desire to solve the problem of agricultural and livestock production. 46 Table 3: Characteristics of Decision-Making Process Decision to Participate Selection of Sub-project Municipality Responses Communal Communal Other PA PA Assembly Assemly Charagua* 23 52% 48% 26% 48% 17% Concepción 9 78% 22% 78% 22% 0% Minero 4 0% 100% 100% 0% 0% Pailón 7 43% 57% 100% 0% 0% San Ignacio 15 53% 47% 100% 0% 0% Average 51.6% 48.3% 69.7% 23.2% 6.9% *There are two missing answers for selection of sub-projects in Charagua. Overall, beneficiaries recognized that the Project helped improve access to land, TA, inputs for agriculture and livestock production, and enhanced production. Beneficiaries, however, do not think that the Project has provided access to markets, increased income, or improved livelihoods. Since livestock sub-projects have not completed their productive cycle, this response is to be expected. Additional years are required for beneficiaries to perceive these improvements. Table 4: What Improved with the Project? Category % of Respondents Agreed Partially Agreed Disagreed Production 58.6% 34.5% 6.9% Access to land 62.1% 15.5% 22.4% Access to services 69.0% 27.6% 3.4% Access to inputs 65.5% 24.1% 10.3% Access to markets 10.3% 74.1% 15.5% Income 43.1% 51.7% 5.2% Livelihood 22.4% 74.1% 3.4% The majority of beneficiaries rated the Project as very positive or positive. Nonetheless, beneficiaries complained about the problems experienced with the Cooperative San Gabriel and the change of PCU coordinators, which in their opinion affected their operations. Also, the perception about the Project’s satisfaction of needs and expectations did not achieve the end target of 80 percent. Only 71 percent of beneficiaries responded to be highly satisfied or satisfied with the Project in this regard; 26 percent responded to be partially satisfied; and 3 percent as unsatisfied. Charagua had the highest level of dissatisfaction with the Project’s fulfillment of needs. Table 5: Level of Satisfaction with the Project Municipality % of Respondents Very Positive Positive Neutral Negative Charagua 4% 87% 9% 0% Concepcion 0% 100% 0% 0% Minero 50% 60% 0% 0% Pailon 57% 43% 0% 0% San Ignacio 53% 47% 0% 0% 47 Beneficiaries have a high level of satisfaction with how their PA functions. 72 percent of responded rated PA functioning as very good or good, 26 percent as regular, and 2 percent as poor. The main reason for a negative rating was the internal problems experienced by a few PAs. The majority of respondents also considered that their individual performance was satisfactory. Only in Charagua a few respondents rated their performance as poor. Table 6: Perception of the PA's Functioning Municipality % of Respondents Very Good Good Regular Poor Charagua 35% 43% 17% 4% Concepcion 0% 56% 44% 0% Minero 0% 25% 75% 0% Pailon 43% 43% 14% 0% San Ignacio 40% 40% 20% 0% In sum, the survey confirmed the positive perception beneficiaries have about project management and activities as well as the uncertainties about material improvements to their conditions. The survey also confirmed the difficult situation faced by beneficiaries in Charagua. 48 Annex 6. Stakeholder Workshop Report and Results (if any) 49 Annex 7. Summary of Borrower's ICR The Bolivia Land for Agricultural Development Project was conceived as a pilot. Project design had a solid concept, but it had to be modified during implementation, particularly Component 1 (Access to Land). Land values in the project area, particularly in the municipalities of Minero and Pailon, were much higher than estimated by the land market study, mainly due to the development of agro-industry. An amendment to the DCA introduced three alternative land access mechanisms. The right of use mechanism in TCOs and TCs was the most suitable for agricultural investments because it had land tenure security. Context. The Project preparation and implementation spanned over three administrations with different approaches to rural development. Between 2004 and 2005, the GOB had an approach based on market development and export promotions. Since 2006, the GOB emphasized food sovereignty and agrarian reform for the rural poor. The Project was the only government program to create new mechanisms for land access during this period. Achievement of PDO. The Project exceeded most of its development indicators. The main achievements were: (i) validating the mechanism for improving access to productive land in TCOs and TCs; (ii) investing in productive sub-projects; (iii) increasing the value of productive assets; (iv) increasing household income; (v) obtaining an acceptable rate of internal return; (vi) organizing producers into Producers’ Associations (PA); (vii) increasing participation and decision-making power of women and youth; (viii) developing capacities in planning, monitoring, accountability, and social auditing; and (ix) organizing five Local Rural Development Councils (CODALs). In spite of these achievements, the Project was faced with many challenges that prevented the development of its original focus on land acquisition. Overall, the achievement of PDO is considered moderately satisfactory. Sustainability. The need for capacity building and technical assistance was high. The Project supported the PAs in developing their organizational, planning, administrative, technical, and commercialization skills. This training, coupled with their individual and collective know-how, would allow them to reach maturity in their investments. Nonetheless, there are still some risks to sustainability. By the time of project closing, the majority of PAs had just finished the investment stage and started the productive stage. Some PAs only received TA for two years. The discontinuance of training and TA could put at risk the production. Climatic risks are also an important factor affecting sustainability. In turn, the organizational sustainability could also be at risk because beneficiaries have little experience working in associations and are more used to household production models. The challenge ahead for the Ministry of Rural Development and Lands would be to continue providing technical assistance in the areas of production, administration, and access to markets as well as assisting PAs in mitigating climate risks. Bank Performance. During preparation, the Bank team showed flexibility and willingness to introduce changes proposed by the GOB. Project supervision was constant and adequately supported the development of the Project and decision-making based on 50 implementation experience. The performance and supervision of FAO-TCI was important for the development of the M&E system, TA strategy, investment operations, as well as the supervision of the final evaluation. In addition, FAO-TCI financed a new land market study in 2012. The Bank team provided advice based on the direct Project experience and knowledge of other countries in Latin America. Overall, Bank performance was satisfactory. Borrower Performance. The GOB could have provided more support to the Project. There were significant delays in the processing of Supreme Decrees to declare the credit effectiveness and approve its amendments. The MDRyT responded adequately to Project needs and complied with the agreements reach with the PCU and the Bank during supervision missions. The MDRyT was a constant source of support and dynamism to the Project. Supervision by the Vice-Ministry of Lands was not always consistent due to the high rotation of vice-ministers during implementation. The Project was virtually closed in 2010, but project management improved significantly after approval of the first amendment to the DCA. Overall, Borrow performance is considered moderately satisfactory. Lessons Learned. The main lesson learned include: (i) importance of preparing and updating a land market study to provide information about land values; (ii) demonstrating the feasibility of market-oriented investments in TCOs and TCs via productive associations; (iii) preparing a technical protocol for purchasing livestock to replace the rule of thumb method; and (iv) opening institutional spaces in the CODALs for planning the development of indigenous peoples. 51 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders n/a 52 Annex 9. List of Supporting Documents Government Documents - Law 1715 (1996) - Law 3545 (2006) - National Development Plan (2006). - Supreme Decree 28160 (2005). - Supreme Decree 28734 (2006). - Supreme Decree 29215 (2007). - Supreme Decree 794 (2011). - Technical Assistance Background Study (2005). - Financial-Economic Background Study (2005). - Production Model Background Study (2005). - Environmental Study (2004). - Social Assessment (2005). - Land Market Background Study (2005). - Operational Manual (2012). - Ex-post Financial and Economic Analysis of Investment Sub-Projects, Gerenssa (September 2014). - Independent Final Evaluation, Strategy Advisors for Government Reform (SAX) (September 2014). - Implementation Completion Report, Ministry of Rural Development and Lands (October 2014). World Bank Documents - Project Appraisal Document (September 27, 2007). - Country Partnership Strategy Progress Report for the Period FY12-FY15 (January 21, 2014). Food and Agriculture Organization - Land Market Study (2012). 53 MAP 54