Document of The World Bank FOR OFFICIAL USE ONLY Report No: ICR00006116 IMPLEMENTATION, COMPLETION, AND RESULTS REPORT ON AN IDA GRANT No. H8350 IN THE AMOUNT OF SDR 46.3 MILLION (US$70 MILLION EQUIVALENT) AN ADDITIONAL IDA GRANT No. D3480 IN THE AMOUNT OF SDR 36.2 MILLION (US$52 MILLION EQUIVALENT) A MULTI-DONOR TRUST FUND SAHEL ADAPTIVE SOCIAL PROTECTION PROGRAM SASPP TF0A2384 IN THE AMOUNT OF US$10 MILLION SAHEL ADAPTIVE SOCIAL PROTECTION PROGRAM SASPP TF0A7655 IN THE AMOUNT OF US$2.4 MILLION TO THE REPUBLIC OF MALI FOR THE EMERGENCY SAFETY NETS PROJECT (JIGISÉMÉJIRI) March 29, 2024 Social Protection and Jobs Global Practice Western and Central Africa Region Official Use Only CURRENCY EQUIVALENTS (Exchange Rate Effective {March 1, 2024) Currency Unit = FCFA FCFA 606 = US$1 US$1.326963 = SDR 1 FISCAL YEAR July 1 - June 30 Regional Vice President: Ousmane Diagana Country Director: Clara Ana Coutinho De Sousa Regional Director: Trina Haque Practice Manager: Christian Bodewig Task Team Leader(s): Kalilou Sylla ICR Main Contributor: Pantaleo Creti Official Use Only ABBREVIATIONS AND ACRONYMS AF Additional Financing AM Accompanying measures CAS Country Assistance Strategy CERC Contingency Emergency Response Component CMS Case Management System CREDD Economic Recovery and Sustainable Development (Cadre pour la Relance Economique et le Developpement Durable) EHCVM Harmonized Survey of Household Living Conditions (L’Enquête Harmonisée sur les Conditions de Vie des Ménages) PAN/EPS National Action Plan for expansion of Social Protection FCFA Franc of African Financial Community (Franc de la Communauté Financière Africaine) FCV Fragile, Conflict-Affected,and Vulnerable GRM Grievance Redress Mechanism ICR Completion and Results Report IDA International Development Association IGAP Income-Generating Activity Program IRI Intermediate Results Indicator LIPW Labor-Intensive Public Work M&E Monitoring and Evaluation MDTF Multi-Donor Trust Fund MIS Management Information System MEFB Ministry of Economy, Finance, and Budget MIS Management Information System NGO Non-Governmental Organization PAD Project Appraisal Document PAN National Action Plan (Plan d’Action National) PDAZAM Mali Drylands Development Project (Projet pour la Diversification Agricole dans les Zones Arides du Mali) PDO Project Development Objective PGFS Governmental Social Safety Net Program (Projet Gouvernemental de Filets Sociaux) PGTMU Governmental Emergency Cash Transfer Program - Covid 19 (Programme Gouvernemental de Transferts Monétaires d’Urgence - COVID 19) PNP Pilot Nutrition Package PRSP Poverty Reduction Strategy Paper RAMED Health Insurance (Régime d’Assistance Médicale) RF Results Framework RI Results Indicator Official Use Only RSU Unified Social Registry (Registre Social Unifié) SCD Systematic Country Diagnostic TTL Task Team Leader ToC Theory Of Change USR Unified Social Registry UTGFS Safety Nets Technical Management Unit (Unité Technique de Gestion des Filets Sociaux) Official Use Only TABLE OF CONTENTS DATA SHEET ......................................................................................................................... 1 I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES ....................................................... 6 A. CONTEXT AT APPRAISAL 6 B. SIGNIFICANT CHANGES DURING IMPLEMENTATION (IF APPLICABLE) 9 II. OUTCOME ..................................................................................................................... 13 A. RELEVANCE OF PDOs 13 B. ACHIEVEMENT OF PDOs (EFFICACY) 14 C. EFFICIENCY 19 D. JUSTIFICATION OF OVERALL OUTCOME RATING 23 E. OTHER OUTCOMES AND IMPACTS 23 III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME ................................ 24 A. KEY FACTORS DURING PREPARATION 24 B. KEY FACTORS DURING IMPLEMENTATION 25 IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME ... 27 A. QUALITY OF MONITORING AND EVALUATION (M&E) 27 B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE 29 C. BANK PERFORMANCE 32 D. RISK TO DEVELOPMENT OUTCOME 34 V. LESSONS AND RECOMMENDATIONS.............................................................................. 35 ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS .......................................................... 39 ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION ........................ 53 ANNEX 3. PROJECT COST BY COMPONENT .......................................................................... 55 ANNEX 4. EFFICIENCY ANALYSIS .......................................................................................... 56 ANNEX 5. BORROWER, CO-FINANCIER AND OTHER PARTNERS/STAKEHOLDERS’ COMMENTS ........................................................................................................................ 65 ANNEX 6. SUPPORTING DOCUMENTS.................................................................................. 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) DATA SHEET BASIC INFORMATION Product Information Project ID Project Name P127328 Emergency Safety Nets project (Jigiséméjiri) Country Financing Instrument Mali Investment Project Financing Original EA Category Revised EA Category Partial Assessment (B) Partial Assessment (B) Related Projects Relationship Project Approval Product Line Supplement P157892-Safety Nets 01-Sep-2016 Recipient Executed Activities Project (Jigisemejiri) Additional Financing P165064-Safety Nets 05-Jul-2018 IBRD/IDA Project (Jigisemejiri) Organizations Borrower Implementing Agency Programme de Filets Sociaux, Unite Technique de Ministry of Economy and Finance Gestion des filets sociaux au Mali - JigisemeJiri Project Development Objective (PDO) Original PDO The objectives of the Project are to provide targeted cash transfers to the poor and food insecure households and to establishbuilding blocks for a national safety net system for the Recipient. Revised PDO The objectives of the project are to increase access to targeted cash transfers for poor and vulnerable households and build an adaptive national safety net system in Recipient's territory . Page 1 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) FINANCING Original Amount (US$) Revised Amount (US$) Actual Disbursed (US$) World Bank Financing P127328 IDA-H8350 70,000,000 68,799,508 64,004,821 P127328 TF-A2384 10,000,000 7,724,849 7,724,849 P127328 TF-A7655 2,400,000 1,122,269 1,122,269 P127328 IDA-D3480 52,000,000 36,414,929 33,575,403 Total 134,400,000 114,061,555 106,427,342 Non-World Bank Financing 0 0 0 Total 0 0 0 Total Project Cost 134,400,000 114,061,555 106,427,341 KEY DATES Project Approval Effectiveness MTR Review Original Closing Actual Closing P127328 30-Apr-2013 17-Aug-2013 17-Jun-2016 30-Jun-2018 30-Jun-2023 RESTRUCTURING AND/OR ADDITIONAL FINANCING Date(s) Amount Disbursed (US$M) Key Revisions 11-May-2018 58.84 Change in Loan Closing Date(s) 26-Sep-2019 78.03 Change in Loan Closing Date(s) 23-Jun-2020 81.49 Change in Loan Closing Date(s) 28-Jun-2023 106.96 Change in Results Framework Change in Components and Cost Cancellation of Financing Reallocation between Disbursement Categories Page 2 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) KEY RATINGS Outcome Bank Performance M&E Quality Moderately Satisfactory Satisfactory Substantial RATINGS OF PROJECT PERFORMANCE IN ISRs Actual No. Date ISR Archived DO Rating IP Rating Disbursements (US$M) 01 04-Jan-2014 Satisfactory Moderately Satisfactory 0 02 14-Jun-2014 Satisfactory Satisfactory 5.67 03 30-Dec-2014 Satisfactory Moderately Satisfactory 7.31 04 09-Jun-2015 Satisfactory Satisfactory 9.65 05 23-Dec-2015 Satisfactory Satisfactory 15.90 06 25-Jun-2016 Satisfactory Satisfactory 24.05 07 24-Dec-2016 Satisfactory Satisfactory 29.90 08 22-Jun-2017 Satisfactory Satisfactory 43.88 09 22-Mar-2018 Satisfactory Satisfactory 55.88 10 06-Dec-2018 Satisfactory Satisfactory 66.81 11 28-Jun-2019 Satisfactory Moderately Satisfactory 76.24 12 10-Feb-2020 Satisfactory Moderately Satisfactory 78.89 13 15-Sep-2020 Moderately Satisfactory Moderately Satisfactory 82.10 14 09-Apr-2021 Moderately Satisfactory Moderately Satisfactory 86.55 15 17-Dec-2021 Moderately Satisfactory Moderately Unsatisfactory 91.21 16 27-Jun-2022 Moderately Satisfactory Moderately Satisfactory 93.31 17 13-Mar-2023 Moderately Satisfactory Moderately Satisfactory 104.99 18 06-Jul-2023 Moderately Satisfactory Moderately Satisfactory 107.75 Page 3 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) SECTORS AND THEMES Sectors Major Sector/Sector (%) Social Protection 100 Social Protection 100 Themes Major Theme/ Theme (Level 2)/ Theme (Level 3) (%) Private Sector Development 100 Jobs 100 Social Development and Protection 85 Social Protection 85 Social Safety Nets 75 Social protection delivery systems 10 Human Development and Gender 15 Health Systems and Policies 9 Health Service Delivery 5 Adolescent Health 2 Child Health 2 Nutrition and Food Security 6 Nutrition 3 Food Security 3 ADM STAFF Role At Approval At ICR Regional Vice President: Makhtar Diop Ousmane Diagana Country Director: Ousmane Diagana Clara Ana Coutinho De Sousa Director: Ritva S. Reinikka Trina S. Haque Page 4 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) Practice Manager: Stefano Paternostro Christian Bodewig Task Team Leader(s): Philippe Auffret Kalilou Sylla ICR Contributing Author: Pantaleo Creti Page 5 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES A. CONTEXT AT APPRAISAL Context 1. Mali is a fragile, semiarid, landlocked, low-income Sahelian country with a predominantly rural economy and an impoverished and rapidly growing population. Agriculture is the country’s largest economic sector and so it faces severe challenges ranging from climate change to increasing human pressure. Its undiversified economy is vulnerable to commodity fluctuations. In 2012, at the time of project design, the poverty rate was estimated at 43.6 percent and population growth at 2.9 percent. Repeated shocks had impacted the already low human development outcomes. In 2012, Mali’s Human Development Index (0.407) was one of the worst in the world, as it ranked Mali 185th out of 190 countries. 2. In 2012, Mali faced an unprecedented political and security crisis. A military coup overthrew the elected government and prevented the organization of elections while armed separatist and jihadist groups occupied the northern regions of the country. Constitutional order was progressively restored with the formation and international recognition of a Transitional Government of National Unity. In July– August 2013, with the support of a United Nations peacekeeping mission, national elections were held. 3. The protracted crisis, combined with the country’s vulnerability to shocks, deteriorated the living conditions of the population, particularly in the northern part of the country. At project design, more than 25 percent of the Malian population was chronically food insecure (over 3 million people), and around 1.7 million people were permanently at risk of hunger. The country’s chronic food insecurity was aggravated by the high number of people displaced due to the crisis. Between January 2012 and February 2013, about 230,000 people fled to the southern Mali and 150,000 people to neighboring countries (mainly Burkina Faso, Niger, and Mauritania). Already vulnerable and poor, the displaced population moved to areas with even higher levels of food insecurity, leading to potentially negative long-term effects on human capital and productivity. 4. Even before the political and security crisis, food insecurity was already widespread despite good macroeconomic performance. Mali’s gross domestic product had been growing since 2000 at the pace of 5 percent, and agriculture was an important element explaining this growth performance. The pace of poverty reduction had slowed down from an annual average of 1.6 percent in 2001–2006 to 1 percent in 2006–2010, reaching 43 percent (or about 25 percent extreme poor). The poverty figures were further exacerbated by the country’s vulnerability to adverse climatic conditions (including drought) and commodity price fluctuations (primarily of cotton, oil, and gold). Once such shocks were struck, the government was unable to reach the poorest segments of the population to provide protection and support, leaving this population particularly vulnerable to greater poverty and food insecurity. 5. The crisis put tremendous pressure on basic service delivery, thus increasing the demand for social safety net programs. In Mali, the existing safety net programs provided ad hoc emergency assistance instead of systematic support and were managed by various ministries. In 2011, a safety net Page 6 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) assessment1 showed that most interventions were temporary, small in scale, and financed by external ad hoc resources; they were not targeted to the poorest and had limited impact. In the lack of a comprehensive strategy and institutional setup, inter-ministerial coordination mechanisms for social protection and social safety nets were weak. Allocating scarce resources to well-targeted and cost- efficient programs was needed, which required setting up a comprehensive safety net system. 6. The government had taken important steps toward building a safety net system. It had decided that the Ministry of Economy, Finance, and Budget (MEFB) would oversee coordination and management for the safety net system. The government had also put in place a multi-sectorial technical team to start collaborating with the World Bank on the design of a social safety net project which would support setting up the envisioned safety net system. Due to the political and security crisis, the preparation process for this project was sped up. In 2013, it became the Emergency Safety Net Project (Jigiséméjiri). Theory of Change (Results Chain) 7. The Project Appraisal Document (PAD) did not describe a theory of change (ToC). For this Implementation Completion and Results Report (ICR), the ToC has been constructed ex post, retrofitted from the project development objective (PDO), the project’s Results Framework (RF), and its detailed description. Figure 1. Theory of Change Activities Intermediate results PDO-level outcomes Long-term outcomes Poor households receive regular Direct cash transfers to poor cash transfers. households To provide targeted cash Component 1 Beneficiaries attend transfers to poor accompanying measures sessions. and food Accompanying measures insecure Mothers and children receive households Preventive nutrition to preventative nutrition package. • Improve human children <5 and pregnant capital outcomes women. and reduce Households increase intergenerational consumption. transmission of Develop MIS poverty. A management information Setup Unified Registry system is developed. • Build households’ resilience to Component 2 shocks. Info, education, and Building blocks for communication a national safety MIS system generates regular reports of adequate quality. net system Setup M&E procedures established Process and impact evaluation Technical assistance and produce reliable information. studies Critical assumptions 1 World Bank Social Safety Nets review (Report No. 53222-ML, 2011). Page 7 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) • Security and political stability Project management expenses • Qualified staff is recruited to ensure capacity and experience Component 3 UTGFS is functional and ratio is less or equal to 20%. • The project is implemented gradually to deal with complexity covers staff and operations costs. • Technical management unit and procedures in place Internal and external audits • Credible institutional setting to ensure project sustainability and staff training Administrative expenses ratio inferior to 20% Source: Constructed ex post by the ICR team based on the original PAD Project Development Objectives (PDOs) 8. The PDO was “to provide targeted cash transfers to the poor and food insecure households and to establish the building blocks for a national safety net system in Mali.” Building blocks included a social registry of poor and vulnerable households and a management information system (MIS). Key Expected Outcomes and Outcome Indicators 9. Outcome 1. Provide targeted cash transfers to the poor and food insecure households. Outcome 1 PDO indicators, as defined in the PAD, are presented below. • Households with access to cash transfers established by the project (number) • Children benefiting from cash transfers (number) • Direct project beneficiaries (core indicator) • Percentage of female beneficiaries (percent) • Beneficiary households that are below the poverty line (percent) 10. Outcome 2. Establish the building blocks for a national safety net system. The PDO indicator for outcome 2 was: • Registry for potential beneficiaries of safety nets programs (report) 11. The project’s Results Framework also contained nine intermediate results indicators (IRI) : five for outcome 1, three for outcome 2, and one related to project management. Components 12. Component 1: Cash transfer program and accompanying measures (original total cost US$112.80 million; actual total cost at closing US$86.2 million). The first component aimed to design and support the implementation of a cash transfer program, combining regular cash transfers and accompanying measures to strengthen investments in households’ human capital. In addition, a subset of children would benefit from nutritional supplements to counter malnutrition. a. Sub-component 1A. Quarterly direct cash transfers to 62,000 poor households in the five regions of the south, plus the district of Bamako, for a period of 36 months. A gradual geographical expansion to the north was envisioned if the security situation improved. Targeting combined geographical and community-based approaches. Payments were delivered to the head of the household, or a person identified by them. Page 8 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) b. Sub-component 1.B. Accompanying measures (AM) consisted of a 12-session package (program information and community campaigns) to both beneficiary and non-beneficiary households in beneficiary villages for a period of 36 months. AMs were delivered by NGOs with the objective of raising awareness of the importance of nutrition, hygiene, and education, hence promoting human capital. c. Sub-component 1.C. Pilot preventive nutrition package dispensed to children under five years and pregnant women as a complementary intervention to AMs in about 10 percent of the villages selected for the cash transfer program. 13. Component 2: Establishment of a basic safety net system (original total cost US$8.30 million; actual total cost at closing US$5.82 million). This component was designed to support the government’s capacity to establish and coordinate the national social safety net system. The component included (i) the development of a management information system (MIS); (ii) the setup of a Unified Social Registry of potential beneficiaries of safety net programs, populated with about 62,000 beneficiaries of the cash transfers program in the first three years and about 122,000 households by the end of 2017; (iii) the setup of an information, education, and communication campaign (IEC); (iv) the setup of monitoring and evaluation procedures; (v) training programs at central and regional levels; (vi) technical assistance (TA) to design a labor-intensive public works (LIPW) component; and (vii) additional studies on targeting mechanisms, payment mechanisms, and AMs. 14. Component 3: Project management (original total cost US$13.30 million; actual total cost at closing US$10.51 million). The component aimed at ensuring efficient project implementation by the Safety Net Technical Management Unit (Unité Technique de Gestion des Filets Sociaux; UTGFS), in conformity with the Financing Agreement, the project document, and the Project Implementation Manual (PIM). It financed: (i) UTGFS personnel; (ii) equipment and operating costs; (iii) regular internal audits and annual external audits; and (iv) training of UTGFS personnel. B. SIGNIFICANT CHANGES DURING IMPLEMENTATION (IF APPLICABLE) Revised PDOs and Outcome Targets 15. In September 2016, the first additional financing (AF1) broadened the scope of the PDO to strengthen the impact on resilience and expand geographical coverage. AF1 changed the PDO statement formulation as follows: “to increase access to targeted cash transfers for poor and ‘vulnerable’ households and build an ‘adaptive’ national safety net system in the Recipient’s territory.” Revised PDO Indicators 16. The two AFs affected the Results Framework, increasing the number of PDOs and intermediate results indicators, as well as their targets. Additional resources also allowed for the inclusion of new activities and a greater number of project beneficiaries. Changes are summarized in Table 1. Page 9 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) 17. AF1 increased the number of PDO indicators from 6 to 11. It also changed the targets of five out of the six original PDO indicators, increasing them to reflect the expanded PDO scope. A new PDO and new IRI indicators were added to capture the new activities (Labor-Intensive Public Works (LIPW) and Income- Generating Activities (IGA)). Furthermore, the indicator on cash transfer beneficiaries was further disaggregated to include school age children (6–14) and elderly people (+60). This was mainly to reflect the broader focus on vulnerabilities. 18. In July 2018, AF2 decreased the number of PDO indicators from 11 to 10. It also increased the targets of six PDO indicators to reflect extra beneficiaries to be reached with the additional funds. Two PDO indicators, which disaggregated beneficiary numbers by age (6–14 and >60 years) were removed since the focus of AF 2 was more on income-generating activities and labor-intensive public works. Following a recommendation of the mid-term review, a new PDO indicator was added to measure the number of households with children under three receiving counseling services or home visits. Eight new intermediate indicators were added to capture LIPW and IGA intermediate achievements and assess home visits to households with children under five (see annex 1). Table 1. Changes in PDO indicators and targets Target Description of changes PDO indicators AF1 AF2 Original Sep.2016 Jul.2018 Outcome 1. To provide targeted cash transfers to the poor and vulnerable households: Households with access to cash 62,000 75,000 100,000 Target increased to account for LIPW transfers established by the project and IGA (AF1); target increased further (AF2) • Households with access to 65,000 87,000 New indicator (AF1) to count only cash cash transfers established by transfer beneficiaries; target increased the project: direct CT (AF2) • Households with access to 20,000 27,000 New indicator to account for 1D (AF1); cash transfers established by Target increased (AF2) the project: LIPW • Households with access to 10,000 32,500 New indicator to account for 1E (AF1) cash transfers established by Target increased (AF2) the project: IGA Direct project beneficiaries 400,000 450,000 550,000 Target revised to account for LIPW and IGA beneficiaries (AF1); Target increased (AF2) • Percent of female 40% 40% 40% No change beneficiaries Children 0–5 benefiting from CT 180,000 65,000 100,000 Target updated due to wrong inference in the parent project preparation (AF1) Children 6–14 benefiting from CT 100,000 Indicator introduced with AF1 and removed with AF2 Elderly +60 benefiting from CT 25,000 Indicator introduced with AF1 and removed with AF2 Share of CT beneficiary households 75% 75% 75% No change who are below national poverty line Page 10 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) Number of households with 10,000 New indicator to measure home visits or children under three benefiting case management system (CMS) as a from counseling services provided follow-up to information sessions by CMS Outcome 2. To establish building blocks for a national safety net system in Mali. Registry for potential beneficiaries Reports 80,000 80,000 AF1 changed the target from ‘producing of safety nets programs produced regular reports’ to ‘number of potential beneficiaries registered’ Revised Components 19. AF1 introduced two new sub-components (1D: LIPW and 1E: IGA) which triggered a change in the safeguards and environmental assessment category. AF1 was funded by a US$10 million Sahel adaptive social protection program multi-donor trust fund (SASPP-MDTF) and extended the closing date by 18 months to December 31, 2019. 20. The LIPW and IGAP were designed with the following features. The LIPW component provided 60 days of work over two years to populations affected by recurrent shocks. Community works included the rehabilitation of basic services infrastructure and land management. Participation was voluntary and associated with soft conditionality such as attending measures on health, nutrition, and education aimed at promoting human capital development. The IGAP component was targeted at cash beneficiaries with a particular focus on woman-headed households. It consisted of an initial community awareness-raising workshop followed by the creation of savings groups, the development of business plans, and the provision of subsidies and coaching. 21. AF2 added an IDA grant of US$52 million and a US$2.4 million grant through SASPP-MDTF. The geographical scope of the project was expanded, and the delivery of AMs was redesigned to promote women’s participation. A Contingency Emergency Response Component (CERC, component 4) was added. The closing date for IDA funds was extended to December 30, 2022. 22. Three restructurings, which took place in 2019, 2020, and 2023, extended the closing dates of the resources added through AF1 and AF2 to ensure achievement of project results. Table 2. Changes with additional financings and restructurings AFs and restructurings Changes AF 1 (September 2016) Change in the PDO statement: “to increase access to targeted cash transfers for Additional funds: poor and vulnerable households and build an adaptive national safety net system SASPP-MDTF TF-A2384: in the Recipient’s territory.” US$10 million Introduction of two new sub-components: 1.D Labor-Intensive Public Works and 1.E. Income-Generating Activities and subsequent change in safeguards triggers and environmental assessment category. Additional US$10 million from SASPP-MDTF to component 1 to increase number of cash transfer beneficiaries: LIPW and IGAP (see table 3). SASPP-MDTF grant TF-A2384 with closing date on June 30, 2018. Extension of the closing date of the parent project IDA-8350 from June 30, 2018, to December 31, 2019. Revision of the Results Framework indicators (see annex 1). Change in the component costs (see table 3). Page 11 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) AF2 (July, 2018) Increase in the number of direct project beneficiaries (see table 3). Additional funds: Improvement in the AM design by emphasizing women’s participation. IDA-D3480: US$52 million Introduction of component 4: Contingency Emergency Response Component SASPP-MDTF TF-A7655: (CERC). US$2.4 million IDA-D3480 grant with closing date on December 30, 2022. SASPP-MDTF grant TF-A7655 with closing date on September 30, 2019. Revision of the Results Framework (see annex 1). Changes in the components’ costs (see tables 4 and 5). Restructuring 1 (September Changes in closing dates of the SASPP-MDTF: 2019) TF-A2384 from September 30, 2019 to December 31, 2019. TF-A7655 from September 30, 2019 to June 30, 2020. Restructuring 2 (June 2020) Changes in closing date: TF-A7655 extended to June 30, 2021. Restructuring 3 (February Changes in closing date: IDA-D3480 grant extended to June 30, 2023. 2023) Cancellation: US$13.5 million Other Changes Rationale for Changes and Implication on the Original Theory of Change 23. Changes throughout implementation reflect the project’s shift toward a more adaptive safety net intervention and system, strengthening the causal linkages leading to the achievement of long-term outcomes. In the revised PDO, “to provide targeted cash transfers” was replaced with “to increase access to targeted cash transfers” to reflect the increased scope of the project activities offering different cash transfer options, with the inclusion of cash for work and income-generating activities targeted at different contexts, including shock-affected areas, and with different specific objectives. The term ‘vulnerable’ replaced ‘food insecure’ to reflect the broadened focus of the project’s target—expanded to households living in areas affected by climate shocks. 24. The term ‘adaptive’ was included to reflect the focus on strengthening the impact of the parent project on increasing resilience and the expansion of the geographical scope in the country where the volatile fiscal and security situation had contributed to increasing poverty and vulnerability. Mali was characterized by repeated and interconnected covariate shocks, such as natural disasters, political crises, conflicts, and forced displacement. In this context, making social protection more adaptive was considered necessary to build the resilience of poor and vulnerable households, which were the hardest hit by these shocks. The project already had adaptive elements with cash transfers helping poor households to become resilient to shocks and a social registry supporting the rapid identification of the poorest. The AF1 strengthened the adaptive nature of the safety net system by increasing the coverage of the social registry in shock-prone areas and the resilience of poor and vulnerable households. Page 12 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) II. OUTCOME A. RELEVANCE OF PDOs 25. At appraisal, the project was aligned with the World Bank's Country Assistance Strategy (CAS)2 and the government’s Poverty Reduction Strategy Paper (PRSP). CAS’s progress report (2011) suggested a two-stage approach toward social protection, combining short-term assistance for the poorest and most vulnerable to cope with shocks with medium- to long-term support for the development of an efficient safety net system. Mali’s PRSP prioritized the reduction of both chronic poverty and food insecurity levels, as well as improvements of the living conditions of the poorest population. One of the PRSP’s strategic areas was aimed at developing integrated programs that shared good and efficient tools and procedures under an effective safety net system. The project supported this aim by enhancing the government's capacity to design, roll out, and scale up an effective safety net system. At appraisal, the project’s objective and design were relevant to Mali’s development priorities, as set out in the government’s national action plan for the expansion of its social protection (Plan d’Action National pour l’Expansion de la Protection Sociale PAN/EPS). 26. At closing, the project’s PDO remained relevant to the latest Country Partnership Framework (CPF 2016–2019) and the strategic pathways identified in the Systematic Country Diagnostic (SCD) 2020 –2025. The PDO was particularly relevant for the achievement of the CPF (2016–2019) objective to strengthen safety nets and risk management mechanisms for the poor and vulnerable and to develop human capital. It also responded to the need to support the government in its efforts to implement the national social protection strategy and extend the coverage of social protection mechanisms, highlighted in SCD 2020–2025. The PDOs contributed to building up key features of a shock-responsive safety net system centered on human capital development and based on technical innovation. During implementation, changes that included strengthening adaptive social protection mechanisms and vulnerable people’s resilience to shocks reinforced the alignment between the PDO, the CPF (2016 –2019), and the SCD (2020–2025). 27. The PDO remained aligned with the government’s 2019–2023 Strategic Framework for Economic Recovery and Sustainable Development (Cadre pour la Relance Economique et le Developpement Durable, CREDD)3 and the National Social Protection Policy (Politique Nationale de la Protection Sociale, 2015). The CREDD promoted social protection to prevent vulnerability and social exclusion. The project particularly resonates with CREDD’s specific objective of promoting social transfers to vulnerable populations and developing mechanisms to improve the effective and efficient implementation of social protection initiatives benefiting the most vulnerable groups. The project is also aligned with the National Social Protection Policy’s objective to reduce poverty among the poorest groups and to improve their access to basic social services in a sustainable manner by enhancing social transfer programs targeted at poor and vulnerable people, in particular single women with families. 2 Mali Country Assistance Strategy. 3 Cadre stratégique pour la relance économique et le développement durable. Page 13 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) Assessment of Relevance of PDOs and Rating 28. The rating is high. The PDOs have been relevant to the World Bank’s CAS and CPF throughout the project’s life. It was aligned with the government poverty reduction policy and remained aligned with its main economic development and social protection policies. B. ACHIEVEMENT OF PDOs (EFFICACY) Assessment of Achievement of Each Objective/Outcome 29. In line with ICR Guidelines, the project outcome has not been evaluated using a weighted split rating since the formal revision of the PDO did not change the substance of the project but rather expanded its scope beyond what had been initially contemplated. Annex 1 summarizes the project’s progress toward PDO indicators and IRIs in the Results Framework. Outcome 1. To increase access to targeted cash transfers for poor and vulnerable households Rating: Substantial Table 3. PDO indicators, targets, and achievements for outcome 1 PDO indicators Revised target Achievement Comment Households with access to cash transfers established 100,000 105,491 Target surpassed by the project (number) • Households with access to cash transfers established by the project: direct cash 87,000 103,541 Target surpassed transfers (number) • Households with access to cash transfers 27,000 14,600 Partially achieved established by the project: LIPW (number) • Households with access to cash transfers 32,500 29,997 Target almost achieved established by the project: IGA (number) Children 0–5 benefiting from cash transfers (number) 100,000 100,000 Target reached Direct project beneficiaries (households) 550,000 576,907 Target surpassed • Female beneficiaries (percent) 40 47.03% Target surpassed Share of CT beneficiary households below national Target surpassed poverty line4 75 66% (poverty among beneficiaries decreased) Households with children under three benefiting from counseling service, and as provided by CMS 10,000 7,270 Target not achieved (number) 30. The project exceeded the targeted number of beneficiary households receiving cash transfers. A total of 105,491 beneficiary households and 576,907 individuals (against the targeted 100,000 4The indicator was interpreted by the project team as the proportion of beneficiaries that were still below the poverty line after the end of the intervention CCT and SBCC cycle. It was measured for the original project through the impact evaluation. At baseline, the impact evaluation found that 88 percent of the beneficiaries were classified as poor, which also represents 88 percent targeting efficiency. At the end of the original project, the endline impact evaluation found that the percentage of beneficiary households still below the poverty line decreased to 66 percent, below the PDO indicator target of 75 percent. Page 14 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) households and 550,000 individuals) received transfers, either in the form of direct cash transfers, LIPW, or IGA. Direct cash transfers were regularly provided over a period of 36 months to 103,541 households, exceeding the PDO target of 87,000. The project nearly reached the targeted number of IGA beneficiaries but faced challenges to achieve the expected number of LIPW beneficiaries. Implementation challenges were due to the three disbursement suspensions caused by the 2020–2021 political crisis, the insecurity situation in the center and north, and the design of the LIPW measures that relied on too many distinct contractors. 31. The combination of cash transfers and AMs led to significant improvements in households’ food security, savings, and asset base, as well as a slight decrease in poverty. An impact evaluation comparing early and late treatment groups at midline found that the project’s cash transfers and AMs increased the proportion of households classified as food secure or having little to no hunger; improved households’ dietary diversity; and improved diet quality. The project’s activities reduced determinants of poverty and increased resilience by strengthening households’ asset base and savings. Beneficiary households held more assets and savings and invested more heavily in income-generating products compared to the control group.5 The asset base was increased particularly in terms of livestock and transport assets.6 The impact evaluation showed a slight decrease in relative poverty (4.6 percent) and the poverty gap (8 percent) in the treatment group compared to the control (late entry) group at the midline. It also found that 94 percent of beneficiary households had a food consumption score above 35 (compared to 83 percent at baselines)—above the targeted 90 percent (intermediate result indicator). 32. Attendance at AM sessions improved over time and was high at the end of the project. Jigiséméjiri delivered information sessions to promote good nutrition, health, education, and family economy practices to participants and other community members. Information sessions ran for 36 months, accompanying cash transfers. The impact evaluation found that about 70 percent of households attended at least one AM session, against the targeted 75 percent (IRI). The evaluation also suggested that since AMs were provided back-to-back (on the same day) with payments made to the head of the household (usually men), women had lower participation rates than men. During the mid-term review, it was recommended that the delivery of the AM package should be revised to take place on a different day than payments. This change increased women’s participation rate to 89 percent. The new AM package included home visits by trained agents to 10,000 poor and vulnerable households with children, directly targeting women/mothers. A total of 7,270 households with children under three benefited from counseling services provided through home visits compared to the PDO target of 10,000 households. Most of them (81.2 percent) received the three planned home visits, exceeding the targeted 75 percent (IRI). 33. AMs had a positive impact on health and child nutrition knowledge and practices. The impact evaluation found initial positive impacts (not included in the Results Framework) of combining cash transfers and AM on two infant and child nutrition practices: a) timely introduction of semi-solid foods and b) prevalence of a minimally acceptable diet. The evaluation found a small reduction in diarrhea prevalence among children 6–23 months, but no immediate significant impact on children’s nutritional status and growth.7 5 Hidrobo, M., at al 2020b. 6 End-line impact evaluation report IFPRI, 2019. 7 Hidrobo M., et al 2020. Page 15 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) 34. LIPW provided access to short-term employment, and beneficiaries perceived the assets created as sustainable, despite the challenging political and security environment which affected achievement of the number of workers employed and workdays created in the center and northern regions. LIPW were introduced in 2017 and offered 60 workdays with the goal of increasing community and households’ resilience to climate shocks. Short-term employment was offered to 14,600 individuals against the 27,000 targeted (55 percent of target), and 556,780 workdays were created against the 1,650,000 originally planned (35 percent of target) (see chapters III-B and IV-b for more details). A total of 105 of the 540 microprojects planned were finalized, mainly in the southern regions. An evaluation8 showed that works selected by communities were in line with community development plans.9 Support measures were implemented with the aim to promote the sustainability of microprojects, particularly through community awareness-raising on land and property rights, soil degradation, and resilience. A total of 88 percent of beneficiaries deemed microprojects to be sustainable, and 62 percent considered that they could lead to more sustainable jobs. 35. Income-generating activities reached 29,997 beneficiaries against the targeted 32,500. IGA activities were initiated as a pilot with AF1 and then scaled up with AF2. An initial IGA pilot (2017–2018) reached 10,000 cash transfer beneficiaries in 20 municipalities.10 Pilot activities were implemented in a timely manner, and three main recommendations emerged from its assessment. First, to extend the IGA duration to two years to include the initial creation of savings and credit groups around which beneficiaries were organized and coached. Second, to raise individual subsidies from US$200 to US$300,11 thus increasing beneficiaries’ investments in productive activities. Third, to increase women’s participation. AF2 took these recommendations on board and increased the IGA target from 10,000 to 32,500 beneficiaries. The expected scale-up coincided with the political and security crisis in 2020–2021, which delayed the procurement process of the implementing partners. The project was able to catch up and regularly implement IGAs, thanks to program extensions. However, the target was not completely achieved because insecurity persisted in the north, particularly in the Mopti region (Youwarou, Koro, and Bankass districts), where no payment agency was able to meet stricter security requirements. 36. The pilot nutrition package (PNP) intervention was effective in improving children’s nutritional status but reached a reduced number of children compared to the target. The procurement process experienced delays due to the lack of local products with the necessary specification. Activities were implemented by WFP through a contract agreement. The number of children benefiting from the nutritional package was 70 percent of the target (70,963 children against 102,000). The impact evaluation, however, found that PNP was effective in reducing chronic child undernutrition and severe anemia. Child morbidity, breastfeeding, and complementary feeding practices were not affected, nor was child dietary diversity. However, the study also found that PNP led to higher rates of caregiver overweight, suggesting that blanket feeding programs such as the PNP intervention can be effective to 8 Sokona, M., 2022. 9 Works involved water and soil conservation, reforestation, rural roads maintenance, construction, and rehabilitation of basic infrastructures. 10 Women participating in the pilot were more likely to run small food commerce and processing activities, subsistence agriculture, and poultry activities, while men were more likely to invest in livestock. 11 Equivalently, from FCFA 120,000 to FCFA 180,000. Page 16 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) reduce child undernutrition but carries the risk of adding calories to subgroups of the population who do not need them.12, 13 37. The proposed rating for outcome 1 is substantial because the targets of the most relevant PDO indicators were surpassed, including the total number of beneficiaries. Overachieved indicators concentrated most of the project funds and beneficiaries. Targets related to the provision of direct cash transfers and AMs were surpassed and proved impactful. The evaluation found smoothed consumption, improved resilience, and enhanced knowledge on health and child nutrition practices among beneficiaries of cash transfers and AM. The PNP intervention was effective in improving children’s nutritional status but reached less children than those initially targeted. Outcome 2. To build an adaptive national safety net system in the Recipient’s territory Rating: High Table 4. PDO indicators, targets, and achievement for outcome 2 Revised target PDO indicator Achievement Comment (Original target if different) Registry for potential beneficiaries of 80,000 households (reports Target 1,300,00014 safety nets programs produced) surpassed 38. The Unified Social Registry (USR), with well-defined institutional settings and implementation structures, has been institutionalized through a decree. The USR was first anchored to the Ministry of Solidarity and Humanitarian Action and then to the Ministry of Health and Social Development after the merger between the two ministries. In October 2015, the National Conference on Social Protection agreed that the USR would be open to contributory and non-contributory social protection schemes. In 2022, a decree15 defined registration procedures and provisions related to data protection and the security of information systems. The USR is currently administratively and financially dependent on a project management unit, which controls different funds and coordinates partners’ activities.16 39. The USR is functional, and several agencies contribute to its database. The key instruments allowing for the establishment of a national social registry started being developed in 2015. These instruments include a questionnaire for collecting household data, a host server, a web portal, a web application, and a mobile application. The design of the questionnaire was the result of participatory discussions with potential users.17 Jigiséméjiri’s beneficiary database has served as the basis for the USR, and several NGOs have actively participated in its construction.18 12 Hidrobo, M., et al 2020. 13 The PNP intervention was set up as a pilot. Funding was not available to continue the pilot. 14 This PDO indicator refers to the total number of individuals in the social registry established by the project. The final achievement includes not only individuals directly registered by the project but also those registered by the government and other agencies utilizing the social registry set up by the project. 15 Décret N°2022-0276/PT-RM du 09 mai 2022. 16 WFP technical note 2020. 17 Idem. 18 Including the Key Program consortia comprising 27 NGOs, as part of an EU-financed project. Page 17 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) 40. The USR database currently hosts data for 1.3 million19 households against the PDO target of 80,000 households. Besides data on the households registered under Jigiséméjiri, it also hosts socioeconomic data for the government’s emergency cash transfer program in response to COVID -19 and several NGO interventions. Since 2015, synergies with the National Medical Assistance Agency (ANAM20 in French) have been effective, and 220,578 households enrolled in the medical assistance scheme (Regime d’Assistance Medicale; RAMED) have been included in the USR database. Initially used mainly as a beneficiary registry, the RSU had its role strengthened, and it now contains information on potential and actual beneficiaries and is used by programs to select beneficiaries.21 41. Jigiséméjiri contributed to the development of the building blocks of the national safety net system. The project set up the UTGFS as the government technical unit for the delivery of safety net programs. The UTGFS was staffed with dedicated specialists, who developed standard operational procedures. The unit created and tested systems (social registry and digital payments) that contributed to an effective and coordinated scale-up of safety net programs by the government and development partners. 42. The safety net system developed under Jigiséméjiri became the platform for the government and development partners to deliver other safety net programs at national scale. In 2017, the government started its own safety net project (PGFS in French) using Jigiséméjiri’s delivery systems and relying on the UTGFS with a dedicated budget line, as the management unit. PGFS was intended to scale up cash transfers and AM across the country, including to the northern regions. Its initial budget for the period 2017–2022 was of FCFA 12.5 billion (around US$20 million), but the political crisis made its financing irregular. As a result, only FCFA 6.2 billion (around US 9 million) were disbursed. In response to COVID-19, the government launched an emergency social program targeted at poor households (Programme Governmental de Transferts Monetaires d’Urgence- Covid 19- PGTMU). Consisting of one- off cash transfers, the program covered 400,000 beneficiary households across the country, building upon Jigiséméjiri’s targeting, beneficiary awareness-raising, procurement, and payment systems. Launched in 2021, the program was maintained by the government in 2022. However, due to the political crisis, only FCFA 42 billion (around US$80 million) out of the initially budgeted FCFA 100 billion (around US$190 million) were disbursed. Stakeholders refer to the Jigiséméjiri as very popular and comprehensive. 43. Systems put in place by Jigiséméjiri were key to expanding project activities in response to shocks. During the 2018 lean season, Jigiséméjiri provided emergency transfers to 31,309 former beneficiary households, demonstrating potential for shock responsiveness. The UTGFS supported and created synergies with safety net components of other WB-funded projects, such as the Mali Drylands Development Project (P164052) (Project pour la Diversification Agricole dans les zones Arides du Mali PDAZAM). 44. The e-transfers pilot improved beneficiaries’ payment conditions and allowed payments during periods of movement restrictions due to COVID-19 and insecurity, despite a few limitations. In 19 The RSU includes a cumulative total 1.3 million records of which 1.16 are house-level and up to date. 20 Agence Nationale d’Assistance Médicale. 21 For example, in zones covered by the Mali Drylands Development Project (P164052, the RSU has universal coverage and proxy- means tests have been used on the database to select project beneficiaries. Page 18 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) 2021, mobile money cash transfers were piloted in the region of Mopti, reaching 3,000 beneficiary households. A qualitative evaluation22 showed that payments through mobile phones gave recipients greater cash-out flexibility and reduced beneficiaries’ travel time to and waiting time at payment sites. E-transfers, however, suffered drawbacks related to network problems in remote sites that forced beneficiaries to use different SIMs. The proposed rating for outcome 2 is high, because the targets have been overachieved. The creation of the social registry resulted in a database with a higher number of registered households than planned. Furthermore, government and partners’ participation in the development and use of the system exceeded expectations. Justification of Overall Efficacy Rating 45. The rating for the overall efficacy is substantial because the operation almost fully achieved outcome 1 and overachieved outcome 2. The targets of the activities under outcome 1 that concentrated most of the project budget (cash transfers and AMs) were overachieved. Instead, the achievement of LIPW and IGAS targets, upscaled in AF2, were affected by increased insecurity, COVID- 19, and discontinued disbursements between 2020 and 2022. Despite the challenges, the project was able to nearly achieve IGA targets, but LIPW targets remained underachieved. However, outcome 2 targets were overachieved. C. EFFICIENCY Assessment of Efficiency and Rating Rating: Substantial Economic analysis 46. Cash transfer amounts were aligned with low-income countries’ standards. A cross-country study shows that, in low-income countries, benefit values represent on average 10 percent of the national poverty line.23 The initial monthly transfer value (FCFA 10,000) was equivalent to 15 percent of the food poverty line and 10 percent of the national poverty line, based on the 2011 household survey (EHCVM 2011). In 2018, considering currency devaluation and inflation, the transfer value was increased to FCFA 15,000 as part of AF2. This was equivalent to 9.5 percent of the national poverty line at the time (EHCVM 2018–2019) and represented 18 percent of the average expenditures of households falling in the first quintile and 20 percent of those falling in the second quintile. At project design, there was minimal experience with cash transfers in Mali. UNICEF's 'Bourse Maman' project, for instance, provided FCFA 5,000 every 8 months to 750 mothers to improve school attendance. However, the Jigiséméjiri approach differed from other agencies, such as the WFP and NGOs, which initiated cash transfer programs in 2018. These programs typically offered higher values (for instance, FCFA 40,000 per month) for shorter durations, often limited to the three-month lean period. The variance in transfer values reflected distinct objectives: shock response transfers aimed to address immediate food needs during hunger periods, while the objective of Jigiséméjiri was to ensure consumption smoothing and to build resilience over a 36-month period. 22 World Bank 2022. 23 Beegle K. et al 2018. Realizing the Full Potential of Social Safety Nets in Africa. AFD & WB. Page 19 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) 47. The amounts were chosen based upon their ability to reduce the incidence and severity of poverty. Simulations conducted at project design showed that an average monthly transfer of FCFA 10,000 targeting the chronically poor (food poverty line) was expected to reduce the food poverty rate by 4.6 percent, the poverty gap by 5.8 percent, and the severity of poverty by 6.7 percent in the targeted regions, even accounting for targeting errors. These estimates were substantial compared to other developing countries.24 The transfer value remained unchanged during the two subsequent additional funding due to the absence of compelling evidence, particularly in terms of poverty data and food inflation, to warrant a reduction in purchasing power or consumption capacity. 48. Ex post simulations25 showed a tradeoff between giving a higher transfer value to fewer beneficiaries or a lower transfer value (FCFA 10,000) to a larger number of beneficiaries. As a follow- up to the ex ante simulation and to estimate the contribution of the transfer value to poverty reduction compared to other transfers values, an ex post simulation was conducted. According to the simulation results, the poverty headcount would have fallen by 1.52 percent if either 23,397 poor households were targeted with FCFA 15,000 or if 35,096 poor households were targeted with FCFA 10,000. Providing the lower transfer value (FCFA 10,000) to a higher number of poor households would have lifted a slightly higher number of people above the food poverty line, which is lower than the poverty line. Giving FCFA 10,000 to 35,096 poor households would have reduced the food poverty headcount in the southern regions by 3.73 percent compared to 2.98 percent if a transfer of FCFA 15,000 were given to 23,397 food poor households. In choosing the transfer value, the long-term objectives of building resilience must be factored in. Higher transfer values allow the poor to meet their basic needs more effectively, enable them to accumulate assets, and give them the opportunity to invest in income-generating activities or small businesses, hence enhancing their resilience. Lower transfers values, though, should be complemented by other forms of support, such as human capital development and productive inclusion to better address the multifaceted nature of poverty and vulnerability. 49. Grants to support income-generating activities were within the standards in the sector. A global review26 of the value of lump-sum grants for productive investments found different average values according to the type of program. The average grant size was US$222 in social safety net programs and US$416 in labor and jobs (L&J) programs. Jigiséméjiri set the grant value in between both: in the pilot phase, cash subsidies for IGA were set at FCFA 120,000 (US$250)27 and given in two installments. In the second phase (2019-22), based on the evaluation recommendations, the value was increased to FCFA 180,000 (US$300). By increasing the value, the program aimed to steer beneficiaries toward productive outcomes rather than to smooth consumption. 50. Over the three years of the program, Mali’s targeting efficiency is comparable to other countries. Social registry costs are a pertinent way to measure the project contribution to the system’s 24 PAD 2013. 25 Mendes 2023. See report in annex. 26 Andrew et al 2021. 27 Exchange rate 2014. Page 20 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) targeting efficiency. A literature review28 on the costs of large-scale social registries in middle-income countries found that they range between US$1 and US$3 per household or less than 2 percent of the benefits channeled through the system. Costs tend to be higher in the inception phase when setup costs have a considerable weight, especially in settings where interoperable administrative data sets are not available. Compared to the total value of three years of cash transfer benefits, the cost-transfer ratio of targeting through the social registry in Mali is 2.5 percent.29 Moreover, if costs associated with the initial setup of the system are excluded and only costs of household interviews (US$10.25 each) are considered, the cost-transfer ratio falls further to 1.15 percent of the total benefit. 51. The development of the building blocks of a national safety net system produced implementation efficiencies in the sector. The UTGFS was set up with specialized staff. It contributed to the development of standard procedures and tested delivery systems. The UTGFS was the platform for the government and development partners to scale up their interventions efficiently and quickly, reducing costs for developing systems (MIS, GRM, and payments). The social registry containing socioeconomic information of more than one million households and beneficiaries enhanced the sector’s efficiency in terms of targeting, coordination, and complementarity. Implementation efficiency 52. The analysis of implementation efficiency has been divided into two. The original IDA financing and subsequent AF1 were implemented without major delays. However, the implementation of AF2 activities was seriously impacted by compound effects deriving from the political and security crisis and COVID-19 pandemic which persisted from 2020 to 2022. 53. The activities foreseen by the original project and AF1 were implemented according to the initial plans. The preparatory phase enabled a quick setup of the implementation unit and the development of the required manuals. This made it possible to start cash transfers six months after project effectiveness. Mitigation measures effectively addressed the risks envisaged in the project document. Emphasis was given to the creation of the UGTFS, the recruitment of specialized staff, and the procurement of payment agencies based on competences. Most importantly, the gradual implementation approach allowed the project to initially focus on safer districts and scale-up once delivery systems were fully functional and adequate security measures were in place. In 2019, the disbursement of the original IDA was at 93 percent. Procurement slowed down due to the increased number of recruitment processes for LIPW and IGA, but activities were conducted regularly. 54. The implementation of AF2 activities, which included the scale-up of the original activities as well as LIPW and IGA sub-components, was affected by three disbursement interruptions. The political crisis and access limitations associated with the worsened insecurity and COVID-19 movement restrictions caused disbursement discontinuity. These continued impediments to a smooth project implementation protracted for over two years (2020–2022) and particularly affected the scale-up of 28 Grosh et al 2022. 29 US$18.3 SR average cost per family / the total benefit of cash transfers beneficiaries 890,64 per 36 months. Page 21 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) LIPWs and IGAs, whose project design required many procurement processes and continued presence of implementing partners in the field (see more detailed discussion in sections III-B and IV-B). 55. The project was able to execute all disbursed funds, except for the last IDA grant (IDA-D3480, part of AF2) of which only 68 percent was executed as of March 25,2024 (Table 6). In 2020 and 2021, Jigiséméjiri was able to spend only a small portion of the planned budget (28.75 and 35.66 percent, respectively),30 which led to two extensions of the AF2 in 2020 and 2023 to accommodate delays and ensure continuity with the new operation. Table 5. Budgetary changes with additional financings Original AF1 AF2 Components (US$, millions) (US$, millions) (US$, millions) Cumulative 56.5 66.5 112.8 1. Cash Transfer Program & 10 MDTF-SASPP 43.9 IDA (IDA-D3480) Accompanying Measures AF (TF-A2384) 2.4 MDTF-SASPP (TF-A7655) Cumulative 48.9 51.3 79.4 1.A Direct Cash Transfers to 26.4 IDA Poor Households AF 2.4 MDTF-SASPP 1.7 MDTF-SASPP 1.B Accompanying Cumulative 3.9 4.2 9.2 Measures AF 0.3 MDTF-SASPP 5 IDA 1.C Pilot Preventing Cumulative 3.7 3.7 3.7 Nutrition Packages AF - 0 0 1.D Labor-intensive Public Cumulative - 4.6 6.7 Works Program AF - 4.6 MDTF-SASPP 2.1 IDA Cumulative - 2.7 13.8 1.E Income Generation 10.4 IDA Activities AF 2.7 MDTF-SASPP 0.7 MDTF-SASPP 2. Establishment of a Basic Cumulative 6.8 6.8 8.3 Safety Net System AF 0 1.5 IDA Cumulative 6.7 6.7 13.3 3. Project Management AF 0 6.6 IDA Cumulative - - 0 4. CERC AF - 0 Total Cumulative 70.0 80.0 134.4 Table 6. Revision of funds throughout the project’s life Fund source IDA MDTF-SASPP MDTF-SASPP IDA IDA-H8350 TF-A2384 TF-A7655 IDA-D3480 Original 70.00 10.00 2.40 52.00 Revised 68.8031 7.72 1.12 47.93* Cancelled 1.20 2.28 1.28 13.5 Disbursed 64.00 7.72 1.12 32.55* Undisbursed 0.00 0.00 0.00 2.11 30The political instability during the time led to several disbursement suspensions, as OP/BP 7.30 procedures were in place. 31The revised amounts (IDA-H8350; TF-A2384; TF7655) came from the last ISR. *Amount in client connection as of March 25,2024. Page 22 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) 56. Project administrative costs32 were maintained below the target and within international standards and decreased over time. According to a global review on safety nets,33 administrative costs on average represent 17 percent of program spending, but this percentage tends to decrease over time because of initial investments in systems. Jigiséméjiri’s administrative costs started at 36 percent, due to large physical investments in equipment needed to kick off the activities. In 2016, before AF1 was effective, administrative costs had declined to 13 percent, while the number of beneficiaries had grown from about 30,000 to over 375,000 people. Even with the addition of two new sub-components (IGA and LIPW) in 2016, administrative costs continued to decrease, reaching 10 and 11 percent in 2017 and 2019, respectively, for a coverage of 390,000 people. Administrative costs slightly increased in mid-2020 due to disbursement suspensions related to political instability that delayed procurement, reaching around 12 percent in 2021 and 2022. 57. The proposed rating for efficiency is substantial. The project’s economic efficiency and cost effectiveness reflect what is expected in the sector. This is inferred from a simulation of the effects of the transfer value on poverty and the comparison of project parameters with global practice and specific experiences in the Sahelian region. The development of a safety net system has generated significant efficiency in the social protection sector. The project’s implementation efficiency was good for the original project and AF1, but external factors undermined the regular implementation of activities under AF2. D. JUSTIFICATION OF OVERALL OUTCOME RATING 58. The overall outcome rating is satisfactory, based on the high relevance of the objectives, its substantial efficacy, and substantial efficiency. The project exceeded its goals regarding the establishment of an adaptive safety net system, the delivery of targeted cash transfers, AMs, and the pilot nutrition package. It almost achieved the targets that AF1 set for IGA and underachieved the goals that AF1 set for LIPW. E. OTHER OUTCOMES AND IMPACTS Gender 59. The evaluation of the parent project provided solid evidence of the impact of cash transfers and AMs on intra-household relationships. Both women and men reported being happier because of the cash transfers and having increased their knowledge thanks to the AMs.34 Jigiséméjiri also brought about a significant decrease in inter-personal violence in polygamous households, where physical violence decreased by 7 percent, emotional violence by 12 percent, and controlling behaviors by 16 percent. 32 The administrative costs ratio is the ratio of the cumulative administrative costs since the start of the project (defined as salaries of UTGFS employees, travel expenses and other travel allowances, rental and maintenance of equipment, costs of operation, maintenance and repair of vehicles, rental and maintenance of offices, equipment and supplies, and subscriptions), divided by the cumulative disbursements since the start of the project. Salaries and allowances of civil servants are not covered by the project (explanatory note in the aide-memoire of November 2015). 33 Beegle et al 2018. 34 Hidrobo et al 2020 (2). Page 23 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) Women beneficiaries in polygamous households experienced less intimate partner violence.35 However, most women did not report important changes in their responsibilities or self-esteem. This can be partly explained by the fact that cash transfers were targeted at household heads, who were mainly men, but the influence of strong gender social norms that limit women’s participation in family decisions must be considered as well. In Koulikoro, an e-payment pilot funded by SASPP provided evidence of the feasibility of targeting women with mobile payment in the context of Mali. Institutional Strengthening 60. The project strengthened the safety net system by institutionalizing a social registry and setting up a safety net management unit. Throughout its implementation, the project built the government capacity on targeting and cash transfer delivery. This enabled the government to scale up safety net programs in case of shocks (as evidenced by the government’s response to COVID-19). Since institutional strengthening is part of project outcome 2, it is discussed in detail in section II-B. Mobilizing Private Sector Financing Poverty Reduction and Shared Prosperity 61. Cash transfers and AMs have a positive impact on poverty reduction, which is reflected in the improvement of poverty indicators as well as in households’ food security, savings, and asset base (discussed in section II-B). The impact evaluation36 observed a reduction in four measurements37 of poverty over time in both early and late entry groups. In the baseline, 87 percent of households were classified as living below the national poverty line, compared to 82 percent at midline and 63 percent at the final survey. At midline, the proportion of households living below the national poverty line was 4.65 percent lower in the early treatment group than in the control group, and the intensity and depth of poverty were 8 percent and 3.65 percent lower, respectively, than in the control group. At the final survey, households in the late treatment group had caught up so well that there was no difference between the two groups. III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME A. KEY FACTORS DURING PREPARATION 62. The initial project design was realistic and simple, building upon extensive preparatory work and policy dialogue, which favored its readiness for implementation. From 2010–2012, the government of Mali, the World Bank, and other development partners were engaged in an intensive dialogue on safety nets. This dialogue resulted in the preparation of an analytical report on safety nets in Mali as well as a feasibility study on cash transfers. These studies combined with discussions held during a series of 35 Heath R., et al 2018. 36 Hidrobo et al 2019. 37 International poverty (defined at USD 1.90 in 2011 PPP); national poverty, poverty gap and severity. Page 24 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) workshops in the context of the preparation of the government’s action plan for the expansion of social protection, retrofitting the dialogue around the development of a safety net project. 63. To deal with the political crisis and insecurity, the project adopted a gradual implementation approach. Initially, the project covered a reduced number of beneficiaries in a few districts in the safer southern regions. Focus was given to cash transfers complemented with AMs. The number of beneficiaries and geographical coverage were expanded gradually, as the government laid the blocks for the creation of a safety net system based on the national social registry. This approach contributed to the successful implementation of the first project phase. 64. Adequate arrangements were made to overcome the limited institutional capacity and sustainability challenges. The technical management unit responsible for managing and implementing the project was established under the Ministry of Economy and Finance. Aiming at institutionalizing Jigiséméjiri, a prime minister’s decree established a project national steering committee to provide policy guidance and supervise project implementation. Other relevant arrangements included the delivery of an information campaign to promote transparency, the creation of an efficient GRM, the development of procurement and financial management guidance, and the recruitment of an independent payment agency based on its qualification. Eventually, as a result of the significant work completed during project preparation, the first quarterly cash payments were implemented six months after project kick-off. Capacity building and technical assistance were planned. 65. An articulated monitoring plan was designed to ensure that lessons learned from program implementation informed the gradual scale-up of the program and filled any implementation gap. The plan included the recruitment of a monitoring specialist and the development of a M&E manual defining the procedures to assess the Results Framework indicators and the progress toward PDO achievement. It also included an impact evaluation focused on three evaluation elements: implementation process; targeting; and impact on beneficiaries, specifically changes in recipients' behavior, levels of consumption, nutritional status, poverty, and vulnerability. The plan was completed with the mid-term review (MTR), which aimed at conducting a participative assessment of the lessons learned to inform the second phase of the project. Monitoring was accompanied by the setup of an MIS aimed at automating data collection, storage, and analysis. B. KEY FACTORS DURING IMPLEMENTATION Factors subject to the control of government and/or implementing agencies. 66. The roles and responsibilities of the different stakeholders at central and decentralized levels were clear and guaranteed a smooth administrative flow, despite some minor delays. The National Social Protection Steering Committee played its guidance role, holding regular meetings and providing recommendations. Appropriate vertical coordination between central and decentralized structures was established. Project implementation and monitoring relied on decentralized structures of the Ministry of Humanitarian Action, Solidarity, and the Elderly at regional and district levels and on local NGOs. According to the UTGFS coordinator, this brought about positive engagement of decentralized actors, Page 25 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) mayors, traditional leaders, and communities at large. Delayed responses from committees slowed safeguards and procurement processes. 67. Key risks and adequate mitigation measures identified at design had to be revised toward the end of the project (2020–2022) due to political instability and exacerbated insecurity. The increased number of security incidents, which also involved financial institutions, brought Jigiséméjiri to reinforce initial mitigation measures with the introduction of bank guarantees for payment agencies38 and the intensification of security directives. Throughout implementation, the project showed flexibility and agility to respond to an ever-changing environment. For example, the project accepted negotiations for higher payment fees for payment agencies to work in insecure areas, allowing them to take additional security measures to reach those areas. 68. Capacity gaps arose during the last phase (AF2), due to an intensified procurement workload resulting from the larger scope of activities and disbursement suspensions. The procurement specialist showed limited capacity to adapt to the increased number of procurement processes after AF1 and AF2. The procurement workload was aggravated by the fact that several contracts with implementing partners expired and also by the fact that procurements had to be relaunched as a consequence of disbursement suspensions. These problems were compounded by the lack of internal coordination, thus affecting the proposal evaluation process, the management of contracts in progress, and the classification and archiving of documents in STEP. Factors subject to World Bank control 69. The World Bank’s proactive approach in collaborating with other partners and facilitating dialogue contributed significantly to the institutionalization of the RSU. The World Bank actively supported the dialogue with development partners in the RSU technical working group. This led to the consensual development of a harmonized questionnaire, which is being used by partners in their data collection (as described in section II-B), and ultimately to the decree institutionalizing the RSU. Factors outside the control of government and/or implementing agencies 70. After the start of the project, while conflict and insecurity continued to worsen, the political situation in Mali improved, creating the necessary conditions for the government to request and obtain additional funds to expand project activities. There was a need for government institutions and services to return to the north of the country, which was most affected by conflict and insecurity. Long-term engagement was required to address Mali’s overall fragile governance and the deteriorated living conditions of the most vulnerable population. The availability of additional funds (Trust Fund and IDA) represented an opportunity to make the Malian social protection system more adaptive and responsive to shocks while building vulnerable households’ resilience. 71. Insecurity upsurge, COVID-19 restrictions, and the recurring political crisis seriously affected the volume of project achievements. Jigiséméjiri was designed in a post political crisis context that was 38That is, a requirement to have an amount equal to the amount to be transferred to beneficiaries available within the Bank’s systems. Page 26 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) followed by a period of relative stability, with widespread food insecurity across the country. In 2020, a succession of events had detrimental effects on project implementation. Insecurity worsened in several intervention areas. The COVID-19 pandemic forced the government to introduce internal movement restrictions. A new period of sociopolitical crises led the World Bank to repeatedly suspend disbursements to the country portfolio. • From 2020 onwards, insecurity escalated, particularly in the north of the country, limiting access to some project areas. An increased number of security incidents was registered in the period 2020–2022, leading to more stringent mitigation measures. These included an improved draft of the security risk management directives, the request for bank guarantees from payment agencies on the funds advanced by the project, and the training of NGO supervisors on environmental and social safeguards (see section IV-B). • In 2020, the introduction of measures in response to the COVID-19 pandemic caused the suspension or delay of field and supervision activities. LIPW and IGA, AMs, and data collection were halted. Activities involving public gathering, including training events and workshops, were suspended. Supervision missions were discontinued. Data collection for RSU expansion was delayed. • Project delivery was negatively affected by repeated suspensions of disbursements by the World Bank to Mali’s portfolio because of the political crisis in August 2020, May 2021, and January 2022 for a cumulative period of 18 months. IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOMES A. QUALITY OF MONITORING AND EVALUATION (M&E) M&E Design 72. The Results Framework was clear and identified relevant and measurable PDO indicators. The Results Framework of the parent project was built around five PDO indicators. AF1 expanded the PDO scope and introduced new activities (LIPW, IGA), increasing PDO indicators to 10. With AF2, PDO indicators were reduced to nine. Two PDO indicators that measured beneficiaries disaggregated by age (6–14 and >60 years) were introduced in AF1 and removed with AF2. A new indicator was added to measure the outreach of households with children under three through counseling services, as recommended in the mid-term review. 73. Most PDO indicators focused on outcome 1 (access of poor and vulnerable households to safety nets) and outcome 2 (strengthening of the safety net system) counted on only one PDO indicator. Initially, the PDO indicator on outcome 2 was not very ambitious: it measured the social registry’s release of annual reports. With AF1, it was broadened to measure the number of households registered in the RSU database. The imbalance between PDO indicators focusing on outcome 1 versus outcome 2 was also visible in the IRI: outcome 1 had 11 IRI and outcome 2 only one. Ideally, the Results Framework should Page 27 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) also have included an indicator on the usage of the RSU by other programs, thereby capturing the cost- saving potentials and public good nature of this tool. 74. The design of Jigiséméjiri included an appropriate monitoring plan, which informed project design improvements. A dedicated M&E specialist was hired. A monitoring strategy was articulated in the M&E manual, including (i) annual process evaluations between 2014 and 2017; (ii) annual spot checks at village level (including beneficiary surveys and qualitative evaluations); (iii) a full impact evaluation with three rounds of data collection; and (iv) annual independent audits of the system. M&E Implementation 75. Activities foreseen in the monitoring and evaluation plan were regularly implemented. Quarterly monitoring reports, including internal audit reports that fed into the annual report had acceptable quality and were submitted regularly (45 days after the end of each quarter). Jigiséméjiri carried out a mid-term evaluation as well as process evaluations to identify programmatic issues and recommend improvements. The impact evaluation was conducted and included three surveys: a baseline in 2014–2015, a midline in 2016, and an endline in 2018. 76. Monitoring was carried out with strong engagement of local authorities and the tight collaboration of local liaison agents, NGOs, government decentralized structures, and village committees. Field monitoring was conducted by NGOs and the National Directorate of Social Protection and Solidarity Economy via its decentralized structures at regional and district (cercle) levels. These decentralized structures were responsible for producing and submitting regular field monitoring reports. In 20 out of 49 districts, the head of the social development services carried out monitoring activities and sent reports. Local monitoring agents were hired by the project at district level to ensure liaison between field partners and the UTGFS at central level. 77. The M&E team faced challenges in the interpretation of some Results Framework indicators, which were revised with the technical supervision of the Bank team. At the end of 2014, the World Bank mission submitted a note to clarify the interpretation of indicators. Issues with the calculation of the indicators reemerged at the end of the project. The M&E specialist experienced challenges in measuring cumulative indicators, probably due to the phased approach of the different activities. This also highlights limitations in the MIS, as this information should have been produced automatically. In the aide-mémoire, the World Bank team noted the persistent discrepancies in the way the M&E specialist consolidated the information provided by sector specialists in the Results Framework. 78. The MIS was developed and generated regular annual reports but was not fully explored by all project components. The technology specialist developed a simple MIS application for the project. In 2018, with the expansion of Jigiséméjiri, a modular MIS was outsourced. The MIS database is hosted in- house and is regularly fed with data on cash transfers produced by payment agencies. The information is received in Excel format and inserted manually in the system. Modules covering other project components (AMs, IGA, and LIPW) are not fed regularly. NGOs and implementing partners send their reports to the Page 28 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) M&E team and relevant project specialists, but information does not transit into the system. This lack of integration is reflected in the monitoring and annual reports, which contain exhaustive information on cash transfers, but not on the other activities. 79. To improve the accuracy and timeliness of field reports, Jigiséméjiri introduced digital monitoring, which was not extensively used due to insecurity. To tackle issues such as distance, timing, and accuracy, the Bank team recommended the use of the Kobo toolbox platform and digital forms to monitor program components, particularly LIPW, IGA, and home visits. This approach had been successfully adopted for similar projects in other West African countries. Kobo toolbox digital forms were first used in 2021 to carry out the LIPW assessment. In 2022, when LIPW activities restarted, the project considered repeating the same approach. However, at that time the government was using geographical information systems (GIS) to track jihadists and did not consent to the use of GPS tracking tools by the project on some sites. Moreover, in areas occupied by jihadists, circulating with tablets or smart phones represented a personal security risk for NGO staff. M&E Utilization Justification of Overall Rating of Quality of M&E 80. The rating of M&E quality is substantial. While there were moderate shortcomings in M&E design and implementation, the M&E system was generally sufficient to assess the achievement of the objectives. Impact evaluations helped to assess causal effects in the results chain and improve the design. However, there were moderate weaknesses in the specification and measurement of the indicators and lack of balance in the analysis of the different components produced by the MIS. B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE Environmental and Social Safeguards. 81. With the introduction of AF1, Jigiséméjiri received environmental classification B and triggered four safeguard policies: (i) Environmental Assessment (OP/BP 4.01); (ii) Involuntary Resettlement (OP/BP 4.12); (iii) Pest Management (OP 4.09); and (iv) Physical Cultural Resources (OP/BP 4.11). Initially, the project fell under environmental classification C and, therefore, did not trigger any safeguards. In 2016, AF1 introduced two new sub-components (LIPW and IGA), moving the project to environmental classification B. This triggered OP/BP 4.01 and led to the disclosure of the Environmental Safeguards Management Framework. In 2018, AF2 triggered OP/BO 4.12 and OP 4.09. The project opted to adopt the directives set out under the Mali Drylands Development Project (PDAZAM) (P164062). Even though project activities were not expected to have adverse impacts on physical cultural resources, AF2 also triggered OP/BP 4.11, and the environmental and social management plan included procedures to be followed in relevant cases. 82. Environmental and social safeguards were handled by decentralized technical services and NGOs at local level. A collaboration agreement was signed between Jigiséméjiri and the National Directorate of Social Protection and Solidarity Economy to ensure coordination. AF2 determined the need Page 29 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) to appoint two safeguards’ specialists by February 201939 to train key stakeholders and NGO staff on the environmental and social management plan (ESMP) and conduct quarterly supervision missions. This arrangement created a direct line management between central and local levels, aiming to ensure the adoption of relevant directives at local level, while keeping proximity monitoring under the responsibility of decentralized agents and NGOs. Recruitments, however, were not finalized before March 2020 for the Social Safeguards Specialist and before March 2021 for the Environmental Safeguards Specialist. 83. Safeguards activities during the period 2019–2022 were classified as moderately satisfactory, mainly because of delays in the implementation of the action plan. The safeguards screenings of the LIPW and IGA projects were regularly conducted but it took up to six months for the safeguards specialists to screen the microprojects and for the local authorities to approve them. The ESMP was regularly shared with decentralized technical services and NGOs. Quarterly monitoring and progress reports on environmental and social safeguards as well as microprojects screening reports were submitted to the Bank team with some delays. 84. The project put in place an effective incident management plan that helped to identify and adjust mitigation, follow-up, and settlement procedures. Since 2019, 13 incidents have been registered: four related to robbery, five related to the resurgence of jihadism, and four road accidents. Three incidents were classified as severe, eight as moderate, and three as low. By the end of the project, all cases had been settled. On September 18, 2021, an incident involved the theft of FCFA 21.9 million out of project funds from a branch of the NGO Stop Sahel. The NGO reimbursed the funds. As mitigation, payment agencies were requested to present bank guarantees on the funds advanced by the project. One payment agency was not able to satisfy this requirement and, therefore, was unable carry out its activities. To deal with insecurity in the implementation areas, the World Bank team recommended the UTGFS to ensure that NGO supervisors had knowledge of environmental and social safeguards. Finally, the attack on three beneficiaries in Niafunké on April 24, 201940 led to the draft of improved security risk management directives. 85. The UTGFS put in place a very simple grievance redress mechanism (GRM) based on verbal communication at community level, which mainly dealt with targeting and payment issues. GRM reports were produced every trimester. The system received a total of 5,692 grievances from Jigisémégiri, the government’s safety net project, and the World Bank-funded PDAZAM project. In total 4,759 grievances were related to Jigisémégiri and 99.8 percent were resolved. At the end of the project only nine grievances (0.2 percent) related to beneficiary phone numbers used to receive payments have remained unresolved. No dissatisfaction or sensitive grievances were registered. This can be partly attributed to the lack of appropriate and alternative uptake channels that guaranteed discretion and anonymity. Efforts were made to set up a green line, but the mobile operator was not ultimately contracted because of the end of the project was approaching. In April 2021, the GRM manual was modified to integrate specific considerations related to the management of SEA/SH cases. The project enforced third parties’ obligations regarding measures to prevent the spread of COVID-19 (hand washing, 39 Four months after the AF2 effective date (October 2018). 40 Attackers stole the amount corresponding to four payments as well as personal property and their means of transportation. Page 30 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) social distancing) and GBV prevention (including the signature of a code of conduct) in the contracts with NGOs, payment agents, equipment suppliers, etc. 86. The project complied with all triggered safeguards policies. Expected frameworks and plans were put in place, microprojects screened, and incident action plans used to improve mitigation measures. Financial Management 87. Administrative, accounting, and financial procedures were mostly sound and have been regularly updated to reflect additional financings and new activities. Half-yearly financial reports and quarterly financial statements were timely and of acceptable quality. In 2016, the departure of both the finance manager and the accountant left a six-month gap in the financial management team. From 2017 onwards, the financial management system has been partly improved but judged as moderately performing due to delays in the collection of supporting documents relative to payment advances. 88. External audits were conducted, and financial statements were considered regular and faithful, but procurement issues delayed audits foreseen in 2019 and 2020. Audits for 2019 and 2020 were only completed in 2021. Financial statements were certified with qualified opinions, due to missing supporting documents relative to advances for field missions and payment agencies. Prompted by the World Bank team’s recommendations, the financial management team made efforts to improve the recovery of supporting documents, but delays continued to occur. Procurement 89. Jigiséméjiri faced important delays in the procurement process. Factors causing procurement delays included capacity and coordination issues, high number of procurements, slow local administrative processes, and the limited capacity of the procurement specialist to adapt to new procedures. Putting together bid evaluation commissions took longer than foreseen because of a limited number of technical specialists, who were often in the field. After recommendations by the World Bank team, the coordinator intervened and ensured that technical specialists were available for bid evaluations. Contracts with service providers were initially prepared per district, which translated into a high number of contracts with the same specifications. Since 2022, based on the World Bank team’s recommendations, new contracts can cover a whole region when implementing agencies have the necessary capacity (framework contracts). 90. Procurement problems caused delays in the implementation of LIPW and PNP sub-components. The procurement of tools and materials for the LIPW component faced several difficulties, the main one being the lack of proper tools and materials specifications by implementing partners. The procurement specialist was obliged to send lists back with requests for clarification. Contracts with payment agencies and NGOs expired before tools and materials could be delivered due to accumulated delays. The bid for the PNP resulted in financial bids that were too high and products that did not meet specifications. Eventually, Jigiséméjiri established a partnership with the World Food Program (WFP) for the implementation of the sub-component, including the procurement and delivery of the nutrition package. Page 31 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) C. BANK PERFORMANCE Quality at Entry 91. The technical design of the project was based on extensive analytical work and policy dialogue on safety nets carried out by the World Bank with the government and development partners. In the years preceding project preparation, extensive economic sector work and policy dialogue on safety nets were carried out by the government supported by the World Bank and development partners. Project features, such as targeting, transfer value, and frequency, were informed by national experience, simulations, and global standards. Studies were carried out to inform the design of activities of the parent project.41 92. The project was strategically relevant and aligned with the government’s and the World Bank’s strategies. The project was consistent with the World Bank’s Country Assistance Strategy. It fit the directives set by the OP 7.30 assessment mission conducted in June 2012 to gradually resume operations in the south and limit new engagements in critical areas to improve the access of the most vulnerable groups to basic services, safety nets, and economic recovery activities. The project was also aligned with the government's poverty reduction policy strategy, which was focused on long-term interventions to protect the country's population, and with the national action plan for the expansion social protection (PAN/EPS). 93. The technical design of the project was sound and based on a gradual approach. The parent project document clearly identified cause-effect relationships between activities, outputs, and outcomes leading to the achievement of PDOs. Project restructurings improved PDO indicators. Key risks were identified and relevant mitigation measures considered. The gradual approach adopted at project design allowed for readiness and effective response in the first phase while preparing for project expansion and scale-up in the second phase. 94. The project was anchored to the Ministry of Economy and Finance, which favored political acceptance and government financial contributions and reduced fiduciary risks. An appropriate institutional setting was in place from the start of the project, including a management unit with permanent core staff set by the government as the lead agency for operationalizing social safety nets in the country. Furthermore, implementation arrangements were based on appropriate vertical coordination with strong engagement of decentralized actors and communities. Quality of Supervision 95. The World Bank team provided regular supervision, produced candid and good quality reports, and obtained the necessary project extensions. Missions were organized twice a year and systematically supported the UTGFS and effectively supervised project implementation. Aide-mémoires were truthful and of good quality. Reports discussed achievements and issues in the different components and technical areas; they included relevant recommendations and action points, which were followed up. In 2018, the 41An analytical report on safety nets in Mali "Safety Nets Review in Mali 2011” followed by a feasibility study “Cash Transfer programs in Mali: A Feasibility Study" (2012). Page 32 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) World Bank team started to hold weekly meetings with the UTGFS to ensure more regular support and follow-up. UTGFS staff interviewed mentioned that high task team leader (TTL) turnover and the consequent discontinuity in the management approach and supporting style affected team dynamics and collaboration. On the positive side, stakeholders appreciated that the co-TTL and the last TTL were based in Mali, which favored closer follow-up and direct technical support to the UTGFS. Delays in the non- objection to the AWPB negatively reflected on procurement timeliness but further checks and controls were deemed necessary to guarantee the effective monitoring of budgeted activities. The World Bank team pursued and obtained three project extensions to ensure the achievement of the development objectives, which had been delayed mainly due to contextual problems. 96. The World Bank team supported the UGTF to roll out AF2 activities after three subsequent disbursement suspensions due to the political and security crisis and COVID-19 restrictions between 2020–2021. Meetings with the UTGFS became more frequent and were held biweekly to ensure more regular support and follow-up. To overcome challenges and delays, security policies and grievance mechanisms were strengthened and additional guarantees from payment agencies introduced. Process evaluations were commissioned to assess the status of suspended activities and define strategies to resume them. As a result, the number of LIPW worksites was reduced. The team also made efforts to tackle procurement issues through capacity building, tight supervision, revision of ToRs, and relevant documents; still, local procurement capacity remained low. The task team also followed up with the project financial management team to find efficient mechanisms to recover supporting documents in a timely manner. 97. The World Bank team ensured adequate technical support and supervision of fiduciary, safeguards, environmental, and social issues. In view of the increased fiduciary and security risks, the team supported the UTGFS to strengthen security directives and introduced the requirement of a bank guarantee from financial partners. It supervised the implementation of safeguard measures and plans and recommended that NGOs, as part of their contractual requirements, identify safeguards focal points. It ensured technical support to enforce third parties’ obligations regarding COVID-19 prevention and guarantee that SEA/SH measures were included in procedures and partners’ activities. 98. The SASPP-TF contributed to the introduction and scale-up of different project activities. It funded the introduction of IGAP and LIPW with AF1 and the scale-up of CT and the Social Registry. It also supported the World Bank team’s regular supervision and technical support. It funded M&E activities, including the review of the IGAP pilot (10,000 beneficiaries) aimed at informing the second phase and the LIPW field evaluation after their suspension. It also supported the digital payment pilot. 99. The World Bank team could have discussed bolder project changes with the government to face the extremely challenging context that affected the implementation of AF2. The team’s support was directed toward introducing mitigation measures to tackle high insecurity and extending the project to allow time for the planned disbursements. The team did not consider more radical solutions such as the early cancellation of the most challenging sub-component (LIPW) or the reallocation of funds toward Page 33 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) other sub-components. This may have represented a missed opportunity toward the full achievement of PDO targets. Justification of Overall Rating of Bank Performance 100. The rating of Bank performance is satisfactory since there were only minor shortcomings in the design, mainly in relation to the indicators’ specifications and the complexity of the additional activities in an extremely challenging environment. The World Bank ensured quality and continuity of supervision, with adequate mitigation measures taken to deal with unexpected challenges, although more audacious changes might have been considered. D. RISK TO DEVELOPMENT OUTCOME 101. The Malian national safety net system and its social safety net program have been put in place but further institutionalization is required. The project supported the establishment of the building blocks of the adaptive safety net system, and the UGTFS led the operationalization and scale-up of the government’s safety net program. The RSU was institutionalized under the National Directory of Social Protection and Solidarity Economy in 2022 and populated with data of more than 1 million households. However, the main management structures of the safety net program and social registry are fully dependent on the World Bank’s funds and technical support. The Government has established a budget line to support extension and updating of the RSU, however, so far, the fund dedicated to the RSU has served to support the meetings of the technical and steering committees of the RSU. The Government should consider increasing its resources and develop a plan that combines its own resources and those of its partners. 102. The discontinuity of the World Bank’s support could impair the system viability. The government has limited financial and technical capacity to take over the system, which could translate into reduced numbers of poor and vulnerable households accessing safety nets once the project ends. This would have negative consequences on determinants of poverty and human capital development. This risk might be mitigated if the government maintains its commitment to fund the national social safety net program with its own resources, as it did throughout project implementation. The Government has committed, through a letter received by the Bank on September 26, 2017, to continue with its contribution of FCFA 12.5 billion (approx. US$25million) to the project. Since then, just about FCFA 6 billion have been transferred to the project. Limited financial and technical capacity of the Government to take over the system would translate into reducing numbers of poor and vulnerable households accessing safety nets once the project ends. This would have negative consequences on determinants of poverty and human capital. 103. Households’ economic gains promoted by Jigisémégiri can be threatened by persistent conflict and insecurity. Jihadist attacks, ongoing violence, and the overall dire humanitarian situation continue to force people to flee the most affected areas, leaving their assets and livelihoods behind. Recrudescent conflict and insecurity can also make it difficult to plan and implement future interventions in struck areas, which are at a higher risk of exacerbated poverty and vulnerability. Page 34 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) V. LESSONS AND RECOMMENDATIONS LESSON 1: The building blocks of adaptive safety net systems can be effectively implemented and scaled up in FCV settings. 104. Jigisémégiri created the necessary conditions for the creation of the Unified Social Registry (USR), including a conducive political setting, strong political acceptance, coordination and collaboration with development partners, a competent management unit, and adequate technological solutions. Recommendations. To effectively manage different partners’ inputs and needs, the USR would benefit from clearly defined procedures and protocols, as well as coordination and accountability mechanisms. In the long term, the USR management unit should become autonomous from the project and be moved under the responsibility of a national agency with adequate governance structures and a national data sharing framework. The USR platform should aim at enhancing data sharing and interoperability between government agencies and humanitarian agencies to ensure targeting accuracy and timely response to crises. 105. Effective coordination mechanisms between humanitarian and development actors should be established, facilitating joint planning, resource mobilization, and implementation to address immediate humanitarian needs while building resilience and ensuring sustainable social protection systems. 106. Jigisémégiri has proven able to effectively lead the implementation of a safety net program at large scale. In a context of recurring shocks, high poverty, and limited resources, a safety net program combining regular cash transfers with AMs implemented by the government reduced determinants of poverty and increased resilience by strengthening poor households’ asset base and savings. Jigisémégiri reached over 500,000 people, and its systems served as a basis for the government’s Emergency Safety Net Project in response to COVID-19 to reach approximately 400,000 households across the country. Recommendation. The scale-up of the safety net program and their gradual take-over by the national safety net system is recommended, given their effective implementation and proven contribution to medium- and long-term development outcomes. 107. The 2018 drought-induced food crisis and the COVID-19 pandemic gave Jigisémégiri the opportunity to demonstrate its potential to scale-up in response to crises. During the 2018 food crisis, Jigisémégiri targeted former beneficiaries in affected areas. In response to COVID-19, the government used Jigisémégiri’s systems and approach to scale up cash transfers. Recommendation. These experiences could be used to inform the establishment of institutional mechanisms that would allow national safety net programs to quickly scale up and contribute to coordinated emergency responses. Policy and technical dialogue between government, humanitarian, and development actors should take place and lead to the integration of safety nets in contingency plans, coordination, and monitoring mechanisms, with agreed-on early warning triggers and operational procedures and common financing mechanisms. Page 35 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) LESSON 2: Safety net project design must be simple, and implementation must be gradual, particularly in unpredictable FCV contexts. 108. The inherent complexity of LIPW associated with the limited time for planning and the high number of microprojects has hindered their integration into the broader safety nets project. Preparatory LIPW activities, including complex procurement, safeguards, and community planning can take at least two years. The one-year period set to implement each LIPW cycle led to procurement and safeguards issues that produced snowball delays resulting in the finalization of only a small portion of the community works planned. Ultimately, the project has not produced evidence of the potential benefits of LIPWs compared to administratively simpler alternatives in terms of skills development, asset value, and sustainability. 109. Recommendation. Particularly in unpredictable FCV contexts, safety net design should be compact and focus on cash transfers with AMs, which have proven cost-effective and are impactful in terms of providing consumption support to vulnerable groups. For their complexity, LIPWs should be planned as a separate project and foresee adequate time for planning and designing microprojects, including the maintenance of the assets created. The procurement for service delivery can represent a key challenge in a FCV context, particularly for activities requiring a high number of contracts. Jigisémégiri faced issues with the high number of procurement processes for LIPW and IGAP. Such issues were minimized by grouping contracts at regional level. Another challenge was the fixed duration of the contracts with service providers, which did not allow for continuity of project activities following project disruptions due to political and security crises and disbursement suspensions. Framework contracts were successfully introduced to ensure the possibility of extending the services after satisfactory evaluation without recurring to new procurement processes. Recommendation: In FCV contexts, strategies to reduce procurement processes (regional procurements) and ensure continuity of services (framework contracts) in case the project is suspended or disrupted due to external factors should be considered at project design stage. LESSON 3: Safety net project design must be gender-sensitive, particularly in contexts where tradition and religion can hinder their participation. 110. Women’s participation is key to promote intergenerational impact and could be further enhanced by including them as recipients of cash transfers. The fact that heads of household (mostly men) were targeted as cash transfer recipients limited the project’s impact on women’s empowerment and on health and child nutrition knowledge. A pilot in Koulikoro provided evidence of the feasibility of targeting women with mobile payments in Mali. Recommendation. To boost intergenerational impact, safety net programs should seek the direct engagement of women as cash transfer recipients and promote their active participation in SBCC activities/AMs. Page 36 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) 111. In the local traditional and religious context, women’s engagement in LIPW remained a challenge, despite measures taken. Aiming at promoting women’s participation, the project set a 30 percent quota. To take local norms into account, women were assigned activities and allocated to locations different from men. Recommendation. Women should be engaged in the design of LIPW activities to guarantee respect of local culture while promoting their participation. LIPWs design should accommodate women’s time constraints and minimize interference with their household responsibilities. Childcare options should be provided. LESSON 4: Strong monitoring and accountability mechanisms are key, but FCV settings can challenge their implementation. 112. Well-developed MIS modules that allow electronic data collection represent the first step toward an effective decentralized monitoring system. The cash transfer module was regularly fed by service delivery agencies and could make timely and comprehensive information available, but this did not happen with the other modules. Most data were collected manually, which impaired data availability, quality, security, efficiency, and timeliness. Recommendation. An effective monitoring system should make use of web-based automated MIS modules that can be remotely accessed by monitoring actors. An automated project MIS can inform the government on the coverage of the project, favoring coordination and reducing duplication and overlapping. It can also inform government response plans and eventually inform policy decisions. 113. A GRM based on local committees can deal with cases relating to programmatic issues but does not offer diverse and appropriate channels for sensitive grievances. Among the more than 5,000 grievances received, there were no cases of dissatisfaction or sensitive grievances, possibly because of the lack of alternative uptake channels that guaranteed discretion and anonymity. Recommendation. An accountable safety net system should make use of a GRM that meets minimum standards, such as offering diverse and accessible uptake channels (including toll-free line operators, interactive voice response, email, and physical complaint boxes) to ensure privacy and anonymity; defining clear timeframes and specific responsibilities to handle grievances with different levels of sensitiveness; and developing transparent referral and beneficiary communication mechanisms. 114. Jigisémégiri’s e-payment pilot showed that e-transfers can improve payment efficiency for beneficiaries and ensure business continuity in contexts of high insecurity and movement restrictions. Payment with mobile phones gave recipients greater cash-out flexibility and reduced beneficiaries’ travel and waiting time at payment sites. Mobile payments were implemented with relative success in the Mopti region when regular cash payments had been interrupted because of movement restrictions related to COVID-19 and insecurity. Page 37 of 67 Official Use Only The World Bank Emergency Safety Nets Project (Jigiséméjiri) (P127328) Recommendation. The potential of e-payments to improve payment efficiency and ensure business continuity in difficult contexts should be further investigated. In remote areas, strategies to deal with distant payment points, low network coverage and low technology uptake among beneficiaries should be considered. In the medium term, e-payments can propitiate digital inclusion. 115. The experience in Mali makes it clear that monitoring in highly insecure contexts remains a challenge. Remote monitoring (such as the Kobo toolbox) is widely recommended to assess project activities in difficult to access and conflict situations. In Mali, however, electronic devices could not be carried in the field without endangering local staff and the population, and the use of GPS coordinates was forbidden. 116. The learning and evidence generation approach of the project could have been more systematic. The evaluation of cash transfers and AMs provided strong evidence of the impact and informed project design improvements. On the other hand, no evidence was produced on the impact and cost effectiveness of IGA and LIPW compared to other livelihood-support alternatives. This would have been A particularly valuable lesson in FCV context. Recommendation. A systematic approach to learning and generating evidence should be envisaged in the design of complex activities, such as LIPW and IGAP, particularly in FCV contexts. . Page 38 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS A. RESULTS INDICATORS A.1 PDO Indicators Objective/Outcome: Households with access to cash transfers established by the project Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Households with access to cash Number 0.00 62,000.00 100,000.00 105,491.00 transfers established by the project 07-Aug-2013 30-Jun-2018 30-Dec-2022 30-Jun-2023 Households with access to Number 43,613.00 62,000.00 87,000.00 103,541.00 cash transfers established by the project: direct cash 17-Jun-2016 30-Jun-2018 30-Dec-2022 30-Jun-2023 transfers Households with access to Number 0.00 0.00 27,000.00 14,668.00 cash transfers established by the project: LIPWP 17-Jun-2016 30-Jun-2018 30-Dec-2022 30-Jun-2023 Households with access to Number 0.00 0.00 32,000.00 27,997.00 Page 39 of !Syntax Error, ! Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) cash transfers established by 17-Jun-2016 30-Jun-2018 30-Dec-2022 30-Jun-2023 the project: IGAP Comments (achievements against targets): The target was achieved with 105,491 households who received cash transfers against the revised target of 100,000 households ( 105 % achievement) Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Children 0-5 benefiting from Number 0.00 180,000.00 100,000.00 100,000.00 cash transfers 07-Aug-2013 30-Jun-2018 30-Dec-2022 30-Jun-2023 Comments (achievements against targets): 100,000 children 0-5 have benefited from cash transfers (100 percent achievement). The ICR considers that the target of 180,000 children on June 30, 2018 has been revised due to wrong interference during parent project preparation. Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Number of households with Number 0.00 0.00 10,000.00 7,270.00 children under 3 years old benefiting from counseling 30-Apr-2019 30-Jun-2018 30-Dec-2022 30-Jun-2023 services provided by CMS Comments (achievements against targets): Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) The target was partially achieved with 7, 270 households with children below 3 years old benefiting from counseling services provided by CMS against a target of 10,000 (73 percent achievement). Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Direct project beneficiaries Number 0.00 400,000.00 550,000.00 576,907.00 07-Aug-2013 30-Jun-2018 30-Dec-2022 30-Jun-2023 Female beneficiaries Percentage 0.00 40.00 40.00 47.03 Comments (achievements against targets): The target of direct beneficiaries of the project was exceeded; 576,907 direct beneficiaries against a revised target of 550,000 people have directly directly benefited from the project ( 101 percent achievement). Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Beneficiaries households that Percentage 0.00 75.00 66.00 are below the poverty line 07-Aug-2013 30-Jun-2018 30-Jun-2023 Comments (achievements against targets): The indicator measures the impact of the project on households that received assistance . There was no baseline at the beginning because the baseline was only measured after appraisal by the impact evaluation (the percentage of beneficiaries that were poor before the start of the project was 88%), which Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) represent also the targeting efficiency. The final evaluation (original project) measured the % of beneficiaries that were still poor at the end of the original project. The project decreased the percentage of beneficiaries households that were poor from 88% at baseline to 66% at the end of the original project, below the target of 75%. Objective/Outcome: Establishing the building blocks for a national safety net system Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Registry for potential Number 0.00 0.00 80,000.00 1,300,000.00 beneficiaries of safety nets programs 07-Aug-2013 30-Jun-2018 30-Dec-2022 30-Jun-2023 Comments (achievements against targets): The target was exceeded. The RSU has a record of 1.3 million against a revised target of 80,000 households. A.2 Intermediate Results Indicators Component: 1. Cash transfer programs and accompanying measures Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Number of beneficiaries that Percentage 0.00 0.00 80.00 70.12 report they are aware of program IGAP objectives and 17-Jun-2016 30-Jun-2018 30-Dec-2022 30-Jun-2023 entitlements Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) Comments (achievements against targets): The target was not achieved because the last survey by the project was conducted in December, 2019 which coincided with the end of the FA1. The bulk of IGA was implemented in the last year of the project and no survey was conducted to reflect the increase (70.1 percent achievement against 80%). Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Beneficiary households that Number 0.00 122,000.00 110,000.00 1,300,000.00 have been identified and registered 07-Aug-2013 30-Jun-2018 30-Dec-2022 30-Jun-2023 Comments (achievements against targets): The target was exceeded with 1,300,000 beneficiary households that have been identified and registered against 110,000 as a target. Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion % of beneficiaries that report Percentage 0.00 0.00 85.00 73.52 they are aware of LIPWP program objectives and 17-Jun-2016 30-Jun-2018 30-Dec-2022 30-Jun-2023 entitlements Comments (achievements against targets): The target was partially achieved mainly because of the difficulty to find payment agencies which will operate in the North (73.2 percent against 85 percent) Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion % of beneficiaries households Percentage 0.00 0.00 75.00 18.10 reporting that their livelihoods has benefitted from LIPWP 17-Jun-2016 30-Jun-2018 30-Dec-2022 30-Jun-2023 created assets Comments (achievements against targets): The target was partially achieved, 18.10 percent achievement against 75 percent. Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Workdays created Days 0.00 0.00 1,620,000.00 556,780.00 17-Jun-2016 30-Jun-2018 30-Dec-2022 30-Jun-2023 Comments (achievements against targets): The target was partially achieved with 556,780 workdays created against 1,620,000. The interruption of the implementation of the component in the center and north regions was the main reason of the non achievement. Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) % of beneficiaries receiving Percentage 0.00 0.00 80.00 36.00 regular payments within the agreed time frame for LIPWP 17-Jun-2016 30-Jun-2018 30-Dec-2022 30-Jun-2023 activities as specified in the Operations Manual Comments (achievements against targets): Target non achieved ( 36 percent of achievement against 80 percent). No payment agencies available in conflict affected areas caused delay. Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion % of beneficiaries receiving Percentage 0.00 0.00 80.00 100.00 contingency resources within 60 days of identification of 17-Jun-2016 30-Jun-2018 30-Dec-2022 30-Jun-2023 needs Comments (achievements against targets): The component was not activated. In the emergency response of Sept 2018, Actual proportion of households paid by Oct 20 was 100 percent. No change. Most payments were done in the south. Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Children benefitting from Number 0.00 102,000.00 70,963.00 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) Nutritional Packages in 07-Aug-2013 30-Jun-2018 30-Jun-2023 household receiving cash transfers Comments (achievements against targets): The project has provided nutritional packages to 105,000 beneficiaries of which 70,963 were children under 5 years and 34,037 pregnant women. The ICR considers that the target was partially achieved because pregnant women were included on the demand of the communities (70,963 children against a target of 102,000). Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Percentage of households with Percentage 0.00 0.00 75.00 81.00 children under 3 years old benefiting from at least 3 13-Apr-2018 30-Jun-2018 30-Dec-2022 30-Jun-2023 sessions of counseling services provided by CMS a year Comments (achievements against targets): The target was exceeded with 81 percent of achievement against 75 percent. Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Proportion of targeted Percentage 0.00 70.00 21.00 households with increased consumption 07-Aug-2013 30-Jun-2018 30-Jun-2023 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) Comments (achievements against targets): Target not achieved ( 21 percent achievement against 70 percent). Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Beneficiaries complying with Percentage 0.00 75.00 89.60 the requirement of participating in the 07-Aug-2013 30-Jun-2018 30-Jun-2023 accompanying measures Comments (achievements against targets): The target was exceeded with 89.6 percent achievement against 75 percent. Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Proportion of targeted Percentage 0.00 70.00 94.00 households with food consumption score >35 07-Aug-2013 30-Jun-2018 30-Jun-2023 (consumption) Comments (achievements against targets): the target was exceeded (94 percent achievement against 70 percent). Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) Component: 2. Establishment of a basic safety net system Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion MIS conceived Yes/No No Yes Yes 30-Sep-2013 30-Jun-2018 30-Jun-2023 Comments (achievements against targets): the target achieved (100 percent of achievement) Component: 3. Project management Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Project management expenses Percentage 0.00 20.00 9.31 ratio 07-Aug-2013 30-Jun-2018 30-Jun-2023 Comments (achievements against targets): The target was exceeded ( 9.31 percent against 20 percent) Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) Administrative expenses Percentage 16.20 20.00 6.31 inferior to 20% 31-Mar-2014 30-Jun-2018 30-Jun-2023 Comments (achievements against targets): the target was exceeded (6.31 percent against 20 percent). Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) B. KEY OUTPUTS BY COMPONENT Objective/Outcome 1: To increase access to targeted cash transfers for poor and vulnerable households 1. 103,541 households had access to direct cash transfers established by the project. 2. 14,600 households had access to cash transfers established by the project for working on LIPW sites. 3. 29,997 households had access to grants to implement income- generating activities. 4. 100,000 children 0–5 years benefited from cash transfers. Outcome indicators 5. 576,907 people have directly benefited from cash transfers 6. 47.03 percent of direct beneficiaries are female. 7. 66 percent of beneficiaries were below national poverty line against 88 percent of beneficiaries at the beginning of the cash transfer program. 8. 7,270 households with children under 3 have benefited from counseling services provided by CMS. 1. 70 percent of beneficiaries reported they are aware of IGAP program objectives and entitlements. 2. 73 percent of beneficiaries that report they are aware of LIPWP program objectives and entitlements. 3. 18 percent of beneficiary households reporting that their livelihoods have benefited from LIPWP-created assets. Intermediate results indicators 4. 556,780 workdays created. 5. 36 percent of beneficiaries received regular payments within the agreed time frame for LIPWP activities as specified in the Operations Manual. 6. % of beneficiaries receiving contingency resources within 60 days of identification of needs (the component was not activated). The Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) project has provided one-time emergency cash transfers to over 30,000 households. 7. 70,963 children benefited from nutritional packages in households receiving cash transfers. 8. 81 percent of households with children under three benefiting from at least three sessions of counseling services provided by CMS each year. 9. 21 percent of targeted households have increased consumption. 10. 89.60 percent of beneficiaries met the requirement of participating in the AMs. 11. 94 percent of targeted households with food consumption score greater than 35 (consumption). 1. 105,491 households had access to cash transfers provided by the project. Key outputs by component 2. 271,146 women have directly benefited from cash transfers. (linked to the achievement of the objective/outcome 1) 3. 94 percent of beneficiaries’ food consumption score is higher than 34. 4. 32,000 people have one-time cash transfers to cope with shocks. Objective/Outcome 2: Build an adaptive national safety net system in the Recipient’s territory Outcome indicators 1. 1.3 million households registered in the social registry. 1. MIS designed. 2. Establishment of a basic social safety net system. 3. M&E procedures set up. Information, education, and Intermediate results indicators communication for the RSU set up. 5. Training at central and regional level: civil servants, municipalities, members of communal committee village committees. 1. 733 communal committees have been had awareness-raising Key outputs by component sessions and were trained in targeting. (linked to the achievement of the Objective/Outcome 2) 2. Civil servants in 146 cercles had awareness-raising sessions. Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) 3. 300 communal committees and 15,000 villages committees have been trained in monitoring the project activities including cash transfers, LIPW, IGA, AMs. 4. Cash transfer delivery system developed (MIS, M&E, Social Registry, GRM, payment mechanism). Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION A. TASK TEAM MEMBERS Name Role Preparation Philippe Auffret Task Team Leader(s) Mahamadou Bambo Sissoko Procurement Specialist(s) Adjalou Celestin Niamien Financial Management Specialist Demba Balde Social Specialist Maman-Sani Issa Social Specialist Supervision/ICR Kalilou Sylla Task Team Leader(s) Mamadou Sangare, Boubacar Diallo Procurement Specialist(s) Ndri Marina Assoumou Financial Management Specialist Tahirou Kalam Financial Management Specialist Mohamed Youba Sokona Environmental Specialist Sinse Bagayoko Environmental Specialist Christophe Bertrand Messina Team Member Silvia Coni Team Member Angele Compaore Ouattara Procurement Team Tolidji Blaise Donou Environmental Specialist Mahamadou Ahmadou Maiga Social Specialist Mariam Denise Brain Team Member Lydie Anne Billey Team Member Pantaleo Creti Team Member Anne Hilger Team Member Page 53 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) Alice Diarra Sangare Procurement Team Johanna van Tilburg Safeguards Advisor/ESSA Benedicte Leroy De La Briere Team Member B. STAFF TIME AND COST Staff Time and Cost Stage of Project Cycle No. of staff weeks US$ (including travel and consultant costs) Preparation FY12 12.643 139,434.60 FY13 40.143 324,669.84 FY14 0 0.00 Total 52.79 464,104.44 Supervision/ICR FY13 0 410.00 FY14 16.771 167,110.02 FY15 18.315 317,799.67 FY16 10.350 77,373.21 FY17 37.000 133,256.07 FY18 8.250 79,338.86 FY19 14.469 70,019.65 FY20 51.874 203,848.52 FY21 33.313 153,850.10 FY22 25.587 137,900.93 FY23 27.502 116,281.68 FY24 11.560 34,142.55 Total 254.99 1,491,331.26 Page 54 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) ANNEX 3. PROJECT COST BY COMPONENT Amount at Approval Actual at Project Percentage of Approval Components (US$M) Closing (US$M) (US$M) 1. Cash transfer program and 112.80 86.2 76 accompanying measures 2. Establishment of a basic 8.30 5.82 70 safety net system 3. Project management 13.30 10.51 79 4. Contingency Emergency 0 0 100 Response Component Total 134.40 102.4542 76 42 The difference between the actual at closing of $102.45 million and the actual disbursed of $106 million is due to the change in exchange rates over the project life cycle. Year 2013 FCFA 501.45 = US$1 (parent project); Year 2016 FCFA 572.52 = US$1 (additional financing 1); Year 2018 FCFA 541.64 = US$1 (additional financing 2); Year 2024 (March 4) FCFA 606 = US$1. Page 55 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) ANNEX 4. EFFICIENCY ANALYSIS Ex Post Simulation – efficiency of the transfers value I. Introduction 1. The Emergency Safety Net Project (Jigiséméjiri) is a government project funded by the World Bank. It was launched in 2013 and lasted for 10 years until June 2023. Its development objective was to increase access to targeted cash transfers for poor and vulnerable households and build an adaptive national safety net system in the Recipient’s territory. 2. One of the project components was to provide cash transfers and AMs to the poor and most vulnerable. In the first phase (2014–2018) a budget of US$48.9 million was allocated to provide quarterly direct cash transfers to 62,000 poor households in the five regions of the south, plus the district of Bamako, for a period of 36 months. 3. At the design stage an ex ante simulation was conducted to compare the impact of different monthly transfer values (from FCFA 10,000 to FCFA 15,000) on poverty reduction. The simulation found that a monthly transfer value of FCFA 10,000 (FCFA 30,000 per trimester) targeting 62,000 chronically poor households was expected to reduce the poverty rate by 4.6 percent, the severity of poverty by 6.7 percent, and the poverty gap by 5.8 percent in the targeted southern regions, even accounting for targeting errors. 4. Additional financing made it possible to implement a second phase of cash transfers during the period 2020–2023, targeting 25,000 poor households in the southern and Gao regions. For this second phase, a new transfer value of FCFA 15,000 was introduced to compensate for inflation. Figure 2. Different waves of beneficiaries per project phase in relation to the periods of the poverty survey (EHCVM 2018–2019) 2014 2015 2016 2017 2018 2019 2020 2021 2022 Parent 13.444 13.444 13.444 Parent 30.169 30.169 30.169 Parent 17.201 17.201 AF1 1.290 1.290 1.290 AF1 1.710 1.710 1.710 AF2 18.810 18.810 18810 AF2 4364 3rd quarter 2nd quarter 2018 2019 National Poverty survey Source: Annual project reports. Page 56 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) 5. The present simulation focuses on the second phase (AF2). It uses data from the national poverty survey (EHCVM 2018–2019)43 to estimate the impact of the new transfer value on poverty compared to alternative values. An ex post simulation of the first phase was not conducted because beneficiaries were still receiving cash transfers during the national poverty survey 2018–2019.44 II. Objective 6. The objective of this ex post simulation is to verify whether the new monthly transfer value of FCFA 15,000 introduced in the second phase (2020–2023), which made it possible to target more than 25,000 poor households, was the most efficient option to reduce poverty or if there were other options that would have made the transfer more efficient (in terms of number of people lifted out of poverty). III. Methodology 7. The following parameters were considered for the simulation of the second phase (AF2): ▪ A budget of US$28.2 million was allocated for component 1A (cash transfers). ▪ The monthly transfer per household value was increased from FCFA 10,000 to FCFA 15,000 due to currency depreciation and inflation. ▪ The US$ / FCFA exchange rate was 561,644 (as per AF2 document). ▪ Targeting efficiency of 80%, meaning that 20% of the targeted households are not poor (based on the results of the impact evaluation of the first phase). ▪ An average household size of 7.8 (based on EMOP 17). 8. In the second phase, with a budget of US$28.12 million, the project could have reached about 30,000 beneficiary households with a monthly transfer of FCFA 15,000 (US$27) per household. Had the transfer value been maintained at FCFA 10,000 (US$10), it could have reached 43,800 beneficiary households. Table 7. Initial parameters for the ex post simulation Monthly transfer 2018 US$/FCFA Monthly transfer Annual transfer Budget component Potential value (FCFA) exchange rate value US$ value US$ 1A (AF2) US$ beneficiaries 10,000 561,644 18 640.97 28,120,000 43,871 11,000 561,644 20 705.07 28,120,000 39,882 12,000 561,644 21 769.17 28,120,000 36,559 13,000 561,644 23 833.27 28,120,000 33,747 14,000 561,644 25 897.37 28,120,000 31,336 15,000 561,644 27 961.46 28,120,000 29,247 Source: ToRs for the ex post simulation 9. The simulation was conducted in four steps, as presented below: 43Enquête Harmonisé sur les Conditions de Vies de Ménages (EHCVM). 44The poverty survey was conducted in two waves: the fourth quarter of 2018 and the second quarter of 2019. In the fourth quarter of 2018, 14,038 households were receiving cash transfers (first cycle) and in the second quarter of 2019, 7,718 households were receiving cash transfers (always first cycle). Page 57 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) 1st step: exclude from the simulation those households that had received cash transfers during the poverty survey (2018–2019)45 10. Within the sample of the national poverty survey, we identified the households that received cash transfers in 2018–2019.46 Out of 6,602 households, 127 households were identified and excluded, resulting in a simulation sample of 6,475 households. 2nd step: simulate the counterfactual scenario with different cash transfer values. 11. The poverty survey 2018–2019 set the national poverty line at FCFA 269,485.30 of annual expenditure per capita. We obtained national poverty indicators as well as poverty indicators for the southern regions.47 Our results suggest that in 2018, there were 872,635 poor households in the southern regions. When the food poverty line is used (FCFA 166,443), 224,284 households are classified as food poor. 12. The following table summarizes the number of households48 used in each simulation. Table 8. Parameters in the simulations using the national poverty line. Monthly payment Households Proportion of Poor Number of households (FCFA) (beneficiaries) household in the randomly selected in south region simulation (using the 2018/2019 survey)ii 10,000 35,096.51 4.0% 63 11,000 31,905.92 3.7% 57 12,000 29,247.09 3.4% 53 13,000 26,997.32 3.1% 49 14,000 25,068.94 2.9% 45 15,000 23,397.67 2.7% 42 Source: Numeric simulations using 2018 national survey data. Note: (i) With a US$28.2 million budget, the project targeted 25,000 poor households with quarterly transfers equivalent to FCFA 15,000/month for 36 months. The number of potential beneficiaries for the other transfer values was estimated considering the fixed budget. Eventually, the number of beneficiaries was adjusted to consider an exclusion error of 20%, as estimated in the project evaluation. 45 These households could not be included in the sample because the effects of the cash transfers were already reflected in their household’s consumption and they were not potential beneficiaries of the second phase (AF2). 46 The national poverty survey asked if the household or any of its members had received assistance in the previous 12 months and from which type of program. Government cash transfers was one of the possible programs. During the survey period Jigisémégiri was the only government cash transfer program. 47 The southern regions include Kayes, Koulikoro, Sikasso, Ségou, Mopti, and Bamako. 48 The scenario considers the actual number of poor beneficiary households (25,000) that were targeted in the second additional financing (2019–2022) with the transfer value of FCFA 15,000. Page 58 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) (ii) Using the national poverty line, we estimate that 872,635 households are classified as poor in the southern regions. Therefore, for the transfer of FCFA 10,000, the number of beneficiaries (35,096.5) would represent only 4 percent of the total poor households. Regarding the number of households in our 2018 survey, we have 1,571 households classified as poor (3,174 as non- poor). Hence, based on our survey sample, this column presents the potential number of households that would be selected to be integrated into the program. Table 9. Parameters in the simulations using the food poverty line. Payment (FCFA) Households Proportion of the Number of households (beneficiaries) food poor randomly selected in households in the simulation (using the southern regions 2018/2019 survey)ii 10,000 35,096.51 16% 64 11,000 31,905.92 14% 58 12,000 29,247.09 13% 53 13,000 26,997.32 12% 49 14,000 25,068.94 11% 46 15,000 23,397.67 10% 43 Source: Numeric simulations using 2018 national survey data. Note: (i) Regarding the food poverty line, we assumed that all beneficiaries would be included in the group of food poor households. (ii) By using the food poverty line, we estimate that in the southern regions 224,284 households are classified as poor. Regarding the number of households in our 2018 survey, we have 408 households classified as poor (4,337 as a non-poor). 3rd step: obtain the adjusted benefits for each randomly selected household i 13. The real benefits for each random selected household i in 2018 is obtained by applying the following formula: 1. Real Benefits_adjustedi = [(VT*12)/spatial deflator] 14. Where VT stands for the transfer value, that is = 10,000; 11,000; 12,000; 13,000; 14,000; 15,000... Spatial deflator represents the adjustment used to obtain real values (according with values presented by the national institute of statistic). 15. Therefore, the adjusted total consumption for each selected household is obtained as: Cons_ adjustedi = [Real_Consumptioni + Real Benefits_adjustedi Page 59 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) 4th step: calculate the new poverty indicators 16. We bring back the randomly selected households with the adjusted consumption values into the survey population to calculate the new poverty indicators: (i) Gini;49 (ii) FGT(O);50 (iii) FGT(1);51 and (iv) FGT(2).52 17. We used the percentage difference between poverty indicators in the 2018–2019 survey and those obtained with the simulation to present our results. IV. Results Figure 3. Average per capita consumption by each region (2018) Source: Data from 2018 survey. Figure 4. Poverty rate (Headcount ratio) by each region (2018) Source: Data from 2018 survey. 49 Index to measure the degree of inequality based on the distribution of expenditure among the population. 50 Poverty rate, an index to calculate the proportion of people living below the poverty line. 51 Poverty gap, an index to measure the expenditure discrepancy of poor people towards the poverty line. 52 Poverty severity, an index to describe the distribution of expenditure among poor people. Page 60 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) Figure 5. Poverty rate in the southern regions (2018) Source: Data from 2018 survey. Figure 6. Summary of the poverty indicators from the poverty survey 2018/2019 (national results) Poverty Line used Number of (%) (%) (%) Gini poor Poverty rate (with FGT(1)*100 FGT(2)*100 households the payment) – FGT (national) (0)*100 National poverty 2,758,632 42.1 11.858 4.617 0.3319 line Food poverty line 242,891 12.2 2.303 0.70 0.3319 Source: Data from the 2018/2019 survey Table 10. Summary of the poverty indicators for the southern regions from the poverty survey 2018–2019 Poverty Line used Number of (%) (%) (%) Gini poor Poverty rate (with the FGT(1)*100 FGT(2)*100 households (in payment) – FGT southern (0)*100 regions) National poverty 2,491,024 42.45 12.014 4.708 .3361 line Food poverty line 224,284 12.5 2.387 0.733 .3361 Source: Data from the 2018/2019 survey. Page 61 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) 18. The following tables show the results of the simulation conducted for the southern regions. Table 11. Simulation of the impact of different transfer values on the poverty indicators for the southern regions using national poverty line (imperfect targeting, 20% leakage) Benefits paid Number of Beneficiaries as (%) Poverty rate (%) (%) Gini for each households % over the total (with the FGT(1)* FGT(2)* household in receiving number of poor payment) – FGT 100 100 2018 benefits (20% households in (0)*100 leakage) south 10,000 35,096.51 4% 41.803 11.819 4.620 .33613 11,000 31,905.92 3.7% 41.803 11.819 4.621 .33613 12,000 29,247.09 3.4% 41.825 11.818 4.620 .3361 13,000 26,997.32 3.1% 41.825 11.819 4.620 .3361 14,000 25,068.94 2.9% 41.803 11.822 4.620 .3361 15,000b 23,397.67 2.7% 41.803 11.825 4.622 .33615 Source: Data prepared by the author using the EHCVM 2018–2019. Note: (i): the table presents results only for the southern regions by using the national poverty line (FCFA 269,485.30 of annual expenditure per capita). Table 12. Simulation of the impact of different transfer values on the poverty indicators for the southern regions using the food poverty line (imperfect targeting, 20% leakage) Benefits paid Number of Beneficiaries as % (%) (%) (%) for each households over the total Poverty rate FGT(1)* FGT(2) Gini household in receiving number of food (with the 100 *100 2018 benefits (20% poor households in payment) – leakage) south FGT (0)*100 10,000 35,096.51 16% 12.033 2.219 0.655 .33579 11,000 31,905.92 14% 12.065 2.220 0.654 .33579 12,000 29,247.09 13% 12.074 2.220 0.654 .33578 13,000 26,997.32 12% 12.122 2.218 0.652 .33576 14,000 25,068.94 11% 12.090 2.219 0.652 .33576 15,000 23,397.67 10% 12.128 2.215 0.651 .3357 Source: Data prepared by the author using the EHCVM 2018. Note: (i): the table presents results only for the southern regions by using the national food poverty line (FCFA 166,443). Page 62 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) Table 13. Differences between the simulation and the 2018 poverty indicators (no transfers) for the southern regions (imperfect targeting, 20% leakage) Benefits paid Number of Beneficiaries as FGT (0) FGT(1) GT(2) Gini for each households % over the total household in receiving benefits number of poor 2018 (20% leakage) households in south 10,000 35,096.51 4% -1.52% -1.62% -1.87% 0.01% 11,000 31,905.92 3.7% -1.52% -1.62% -1.85% 0.01% 12,000 29,247.09 3.4% -1.47% -1.63% -1.87% 0.00% 13,000 26,997.32 3.1% -1.47% -1.62% -1.87% 0.00% 14,000 25,068.94 2.9% -1.52% -1.60% -1.87% 0.00% 15,000 23,397.67 2.7% -1.52% -1.57% -1.83% 0.01% Source: Numeric simulations using 2018–2019 national survey data. Table 14. Differences between simulation and the 2018–2019 food poverty indicators (no transfers) for the southern regions (imperfect targeting, 20% leakage) Benefits paid Number of Beneficiaries as FGT (0) FGT(1) FGT(2) Gini for each households % over the total household in receiving benefits number of poor 2018 (20% leakage) households in south 10,000 35,096.51 16% -3.74% -7.04% -10.64% -0.09% 11,000 31,905.92 14% -3.48% -7.00% -10.78% -0.09% 12,000 29,247.09 13% -3.41% -7.00% -10.78% -0.10% 13,000 26,997.32 12% -3.02% -7.08% -11.05% -0.10% 14,000 25,068.94 11% -3.28% -7.04% -11.05% -0.10% 15,000 23,397.67 10% -2.98% -7.21% -11.19% -0.12% Source: Numeric simulations using 2018 national survey data. V. Final Note 19. The ex post simulation found that there is a tradeoff between giving a higher transfer value to fewer beneficiaries and giving a lower transfer value (FCFA 10,000) to a larger number of beneficiaries. Page 63 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) According to the simulation results, the poverty headcount is reduced by 1.52 percent either if 23,397 poor households are targeted with FCFA 15,000 or if 35,096 poor households are targeted with FCFA 10,000. 20. The lower transfer value (FCFA 10,000) to a higher number of poor households can lift a slightly higher number of people out of the food poverty line, which is lower than the poverty line. Giving FCFA 10,000 to 35,096 poor households reduces the food poverty headcount in the southern regions by 3.73 percent compared to 2.98 percent if a transfer of FCFA 15,000 is given to 23,397 food poor households. The higher value is more effective to reduce the severity and the gap of food poverty. Page 64 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) ANNEX 5. BORROWER, CO-FINANCIER, AND OTHER PARTNERS/STAKEHOLDERS’ COMMENTS 1. Donors have reviewed the report and provided comments which are synthetized below. In bold, the ICR team explains how the comments have been addressed. 2. Donors acknowledged the relevance of the lessons and recommendations for the World Bank operations and stakeholders' engagement in ASP in FCV settings. They expressed support to disseminating recommendations and discussing their implications for future program design. 3. Donors highlighted the importance of the USR as a foundation for a national safety net system. They suggested that recommendations address and highlight its institutionalization and the collaboration with partners, particularly regarding future registration and interoperability/data sharing possibilities. In response to this comment, further details were added to the recommendations. 4. Donors emphasized the significance of policy and technical dialogue between government, humanitarian, and development actors, especially in politically unstable and insecure environments. They also inquired about further lessons and recommendations regarding linkages between humanitarian interventions and social protection. An additional recommendation was elaborated to take this comment into account. 5. Donors highlighted that disability inclusion indicators were missing, even though the AF2 document mentioned counseling services in support to caregivers of elderly and disabled persons as part of home visits. The focus on disability was part of a broader support to vulnerable groups. The relevant indicator in this case was the proportion of households that have received three home visits. No impact evaluation, which could have provided further analysis, was conducted in the second phase of the project. 6. Donors asked whether the new PDO focus on 'vulnerable' (rather than ‘food insecure’) households included vulnerability to conflict-related protection issues. Although development objectives and outcomes do not explicitly address protection issues, they are included in project safeguard measures and throughout project implementation. 7. The impact of the program on violence reduction was noted, but the donors asked whether more analysis was available on the impact on polygamous versus non-polygamous households. The ICR provides an explanation for these findings and references the relevant study for further details. 8. Donors asked whether recommendations to simplify program design in FCV contexts, especially concerning the LIPW component, and to diversify grievance uptake channels to favor sensitive complaints could be generalized to other countries in the region. Both recommendations can be generalized, particularly at regional level. Page 65 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) 9. Donors appreciated mentioning the need to focus on increasing women's participation in cash transfers. They suggested including a more nuanced recommendation regarding women's empowerment through targeted cash transfers, such as home visits. Women’s participation is influenced by cultural and social norms. While suggesting a more proactive engagement of women, the project has also supported an evaluation of CT targeting women through the PDAZAM project. The findings of this evaluation are expected provide more nuanced recommendations for women’s engagement in the specific context of Malian rural areas. 10. The evaluation made by the borrower recognizes that the project has achieved its development objective by setting up a system and helping reducing poverty among the beneficiaries’ households. The borrower finds the ICR very comprehensive and rich in terms of lessons learned. The borrower has recognized difficulties during the implementation related to labor intensive public work (LIPW) and made some recommendations including a better planning of activities given the volatile environment and reduction of the number of actors involved in the implementation. Despite the setback in LIPW, the borrower encourages IDA to continue financing LIPW which have become popular choice among beneficiaries and stakeholders. Page 66 of 67 Official Use Only The World Bank Emergency Safety Nets project (Jigiséméjiri) (P127328) ANNEX 6. SUPPORTING DOCUMENTS Bibliography Colin A., de Montesquiou A., Arévalo Sánchez I., Dutta P., Varghese Paul B., Samaranayake S., Heisey J., Clay T., and Chaudhary S. 2021. The State of Economic Inclusion Report 2021: The Potential to Scale. Washington, DC: World Bank. doi:10.1596/978-1-4648-1598-0. License: Creative Commons Attribution CC BY 3.0 IGO Beegle, K.; Coudouel, A.; Monsalve, 2018. Realizing the Full Potential of Social Safety Nets in Africa. Africa Development Forum. Washington, DC: World Bank Grosh, Margaret, Phillippe Leite, Matthew Wai-Poi, and Emil Tesliuc, editors. 2022. Revisiting Targeting in Social Assistance: A New Look at Old Dilemmas. Washington, DC: World Bank. doi:10.1596/978-1-1814-1. License: Creative Commons Attribution CC BY 3.0 IGO Heath R., Hidrobo M., Roy S. (2018). Cash Transfers, Polygamy, and Intimate Partner Violence. Experimental evidence from Mali. IFPRI Discussion Paper 01785 December 2018 Poverty, Health, and Nutrition Division Hidrobo M., Roy S., Huybregts L., Njee-Bugha L., Sessou E., Kameli Y. 2018. Rapport de revue à mi-parcours du programme de filets sociaux (Jigiséméjiri) Hidrobo M., Roy S., Huybregts L., Njee-Bugha L., Karachiwalla N. Kameli Y. 2019. Rapport de revue finale du programme de Filets sociaux (Jigisémèjiri). IFPRI Hidrobo M., Huybregts L., Karachiwalla N., and Roy S. (2020a). Multi-Component Cash Transfer Programs: Evidence from Mali’s Social Safety Net Program (Jigisémèjiri). IFPRI and IRD Hidrobo M, Karachiwalla M., Roy S.(2020b). The Impacts of Cash Transfers on Mental Health and Investments Experimental Evidence from Mali. IFPRI Discussion Paper 01989 MEF (2021). Évaluation des chantiers des travaux à haute intensité de main d’œuvre (HIMO dans les cercles de Niono, Nara, Youwarou et Niafunké). Programme de Filets Sociaux. Ministry of Economy and Finance. MEF (2022). Rapport d’achèvement du programme “Jigisemejiri”. Ministry of Economy and Finance. Sokona M.E.B., (2022). Évaluation du processus de mise en œuvre des travaux à haute intensité de main d’œuvre du programme Jigisémégiri. Final report. WFP (2020). Note d’orientation et d’analyse - contribution à la conception et à l’utilisation du Registre Social par le PAM Mali. Internal document. World Bank (2022). Mali: Beneficiary experiences with mobile cash transfers. Insights from survey and qualitative data. G2Px / BMGF Page 67 of 67 Official Use Only