FOR OFFICIAL USE ONLY Report No: ICR00005366 IMPLEMENTATION COMPLETION AND RESULTS REPORT ON A CREDIT IN THE AMOUNT OF SDR 166.9 MILLION (US$250 MILLION EQUIVALENT) AND AN ADDITIONAL CREDIT IN THE AMOUNT OF SDR 36.8 MILLION (US$50 MILLION EQUIVALENT) TO THE REPUBLIC OF KENYA FOR THE NATIONAL SAFETY NET PROGRAM FOR RESULTS June 10, 2021 Social Protection and Jobs Global Practice Eastern and Southern Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS (Exchange Rate Effective June 10,2021) Currency Unit = Kenya Shilling (Ksh) Ksh 107.9 = US$1 US$1.4438 = SDR 1 FISCAL YEAR July 1 – June 30 ABBREVIATIONS AND ACRONYMS AF Additional Financing ASAL Arid and Semi-arid Land BOS Beneficiary Outreach Strategy CCTP Consolidated Cash Transfer Program CGM Complaint and Grievance Mechanism CT-OVC Cash Transfer for Orphans and Vulnerable Children CPS Country Partnership Strategy DfID Department for International Development DLI Disbursement-Linked Indicator DLR Disbursement-Linked Result DP Development Partner EFC Error, Fraud, and Corruption ESSA Environmental and Social Systems Assessment FCDO UK’s Foreign, Commonwealth and Development Office FM Financial Management G&CM Grievance and Case Management GDP Gross Domestic Product GoK Government of Kenya HSNP Hunger Safety Net Program HTM Harmonized Targeting Methodology ISR Implementation Status and Results Report IFAR Integrated Fiduciary Assessment Report IFMIS Integrated Financial Management System IPF Investment Project Financing IPRS Integrated Population Registry Service KENAO Kenya National Audit Office KIHBS Kenya Integrated Household Budget Survey KSEIP Kenya Social and Economic Inclusion Project M&E Monitoring and Evaluation MGCSD Ministry of Gender, Children, and Social Development MIS Management Information System MLEAA Ministry of Labour and East African Affairs MLSP Ministry of Labor and Social Protection MOF Ministry of Finance MTEEF Medium-Term Expenditure and Financing Framework MTR Midterm Review NDEF National Drought Emergency Fund NDMA National Drought Management Authority NEDI Northeastern Development Initiative NGO Nongovernmental Organization NRB National Registration Bureau NSNP National Safety Net Program NSPP National Social Protection Policy OM Operations Manual OPCT Older Persons Cash Transfer OPM Office of the Prime Minister PAD Program Appraisal Document PAP Program Action Plan PBC Performance-Based Condition PCK Postal Corporation of Kenya PDO Project Development Objective PforR Program for Results PIBS Program Implementation and Beneficiary Satisfaction PMT Proxy Means Test PSP Payment Service Provider PWSD-CT Persons with Severe Disabilities Cash Transfer RF Results Framework SAU Social Assistance Unit SDSP State Department for Social Protection SPS Social Protection Secretariat UFS-CT Urban Food Subsidy Cash Transfer UNICEF United Nations Children’s Fund WFP World Food Programme Regional Vice President: Hafez M. H. Ghanem Country Director: Keith E. Hansen Regional Director: Amit Dar Practice Manager: Paolo Belli Task Team Leaders: Yuliya Smolyar, Muhammad Iftikhar Malik ICR Main Contributor: Suzana N. de Campos Abbott TABLE OF CONTENTS DATA SHEET .......................................................................................................................... 1 I. PROGRAM CONTEXT AND DEVELOPMENT OBJECTIVES .................................................... 5 A. CONTEXT AT APPRAISAL AND THEORY OF CHANGE .................................................................5 B. SIGNIFICANT CHANGES DURING IMPLEMENTATION .............................................................. 13 II. OUTCOME .................................................................................................................... 15 A. RELEVANCE .......................................................................................................................... 15 B. ACHIEVEMENT OF PDOs (EFFICACY) ...................................................................................... 16 C. JUSTIFICATION OF OVERALL OUTCOME RATING .................................................................... 25 D. OTHER OUTCOMES AND IMPACTS (IF ANY) ........................................................................... 25 III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME ................................ 28 A. KEY FACTORS DURING PREPARATION ................................................................................... 28 B. KEY FACTORS DURING IMPLEMENTATION ............................................................................. 29 IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME .. 30 A. QUALITY OF MONITORING AND EVALUATION ....................................................................... 31 B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE ..................................................... 33 C. BANK PERFORMANCE ........................................................................................................... 35 D. RISK TO DEVELOPMENT OUTCOME ....................................................................................... 36 V. LESSONS AND RECOMMENDATIONS ............................................................................. 36 ANNEX 1. RESULTS FRAMEWORK, DISBURSEMENT-LINKED INDICATORS, AND PROGRAM ACTION PLAN ...................................................................................................................... 38 ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION ......................... 55 ANNEX 3. PROGRAM EXPENDITURE SUMMARY ................................................................... 57 ANNEX 4. BORROWER’S COMMENTS ................................................................................... 58 ANNEX 5. SUPPORTING DOCUMENTS .................................................................................. 59 ANNEX 6: LINK OF OUTCOME AND INTERMEDIATE OUTCOME INDICATORS WITH DLIS, AS APPROVED AND AS RESTRUCTURED .................................................................................... 60 ANNEX 7: KEY OUTCOME INDICATORS WITH DLIS IN RELATION TO THE PROGRAM’S THEORY OF CHANGEA ............................................................................................................................. 66 The World Bank National Safety Net Program for Results (P131305) DATA SHEET BASIC INFORMATION Product Information Program ID Program Name Financing Instrument P131305 National Safety Net Program for Results Program-for-Results Financing Country IPF Component Kenya No Organizations Borrower Implementing Agency Ministry of Labor and Social Protection, National Drought Republic of Kenya Management Authority Program Development Objective (PDO) Original PDO To support the Republic of Kenya's efforts to establish an effective national safety net program for poor and vulnerable. Page 1 of 69 The World Bank National Safety Net Program for Results (P131305) FINANCING FINANCE_TBL Original Amount (US$) Revised Amount (US$) Actual Disbursed (US$) World Bank Administered Financing 250,000,000 208,970,629 195,458,463 IDA-52870 50,000,000 34,110,190 35,777,459 IDA-59810 Total 300,000,000 243,080,819 231,235,922 Non-World Bank Administered Financing Borrower/Recipient 0 0 0 Total 0 0 0 Total Program Cost 300,000,000 243,080,819 231,235,922 KEY DATES Program Approval Effectiveness MTR Review Original Closing Actual Closing P131305 23-Jul-2013 16-Oct-2013 26-Feb-2016 31-Mar-2018 31-Dec-2020 RESTRUCTURING AND/OR ADDITIONAL FINANCING Date(s) Amount Disbursed (US$M) Key Revisions 27-Apr-2017 149.61 Additional Financing Change in Results Framework Change in Loan Closing Date(s) Reallocation between and/or Change in DLI Change in Legal Covenants 23-Jun-2020 229.43 Change in Results Framework Change in Loan Closing Date(s) Change in Implementation Schedule 03-Dec-2020 231.24 Cancellation of Financing KEY RATINGS Outcome Bank Performance M&E Quality Satisfactory Satisfactory Substantial Page 2 of 69 The World Bank National Safety Net Program for Results (P131305) RATINGS OF PROGRAM PERFORMANCE IN ISRs Actual No. Date ISR Archived DO Rating IP Rating Disbursements (US$M) 01 05-Jan-2014 Satisfactory Moderately Satisfactory 0 02 17-Jun-2014 Moderately Satisfactory Moderately Satisfactory 8.97 03 26-Dec-2014 Moderately Satisfactory Moderately Satisfactory 26.96 04 23-Jun-2015 Moderately Satisfactory Moderately Satisfactory 39.20 05 22-Jan-2016 Moderately Satisfactory Moderately Satisfactory 103.94 06 02-Sep-2016 Satisfactory Satisfactory 145.01 07 06-Apr-2017 Satisfactory Satisfactory 149.61 08 30-Oct-2017 Satisfactory Satisfactory 188.24 09 29-Jun-2018 Satisfactory Satisfactory 198.63 10 05-Mar-2019 Satisfactory Satisfactory 198.63 11 28-Oct-2019 Satisfactory Satisfactory 225.94 12 26-May-2020 Moderately Satisfactory Moderately Satisfactory 225.94 13 14-Dec-2020 Moderately Satisfactory Moderately Satisfactory 231.24 SECTORS AND THEMES Sectors Major Sector/Sector (%) Social Protection 100 Public Administration - Social Protection 100 Themes Major Theme/ Theme (Level 2)/ Theme (Level 3) (%) Finance 1 Finance for Development 1 Disaster Risk Finance 1 Page 3 of 69 The World Bank National Safety Net Program for Results (P131305) Social Development and Protection 95 Social Protection 95 Social Safety Nets 90 Social Insurance and Pensions 5 Urban and Rural Development 3 Disaster Risk Management 3 Disaster Response and Recovery 1 Disaster Risk Reduction 1 Disaster Preparedness 1 ADM STAFF Role At Approval At ICR Regional Vice President: Makhtar Diop Hafez M. H. Ghanem Country Director: Diarietou Gaye Keith E. Hansen Director: Ritva S. Reinikka Amit Dar Practice Manager: Lynne D. Sherburne-Benz Paolo Belli Yuliya Smolyar, Muhammad Task Team Leader(s): William David Wiseman Iftikhar Malik Suzana Nagele de Campos ICR Contributing Author: Abbott Page 4 of 69 The World Bank National Safety Net Program for Results (P131305) I. PROGRAM CONTEXT AND DEVELOPMENT OBJECTIVES A. CONTEXT AT APPRAISAL AND THEORY OF CHANGE 1. At the time of appraisal of the National Safety Net Program for Results (the PforR, the Program)1 in early 2013, Kenya had experienced a decade of sustained but volatile economic growth that was attributed to a stable macroeconomic environment, investments in infrastructure and education, and structural reforms in key sectors of the economy that had created a strong foundation for the country’s future. Kenya’s growth had been volatile: growth increased above 5 percent before 2007, declined in 2008 and 2009 as a result of the violence following December 2007 presidential elections, and the impacts of a drought and of the global food, fuel and financial crisis, then rebounded strongly in 2010 but fell again in 2011 and 2012 due to another drought and macroeconomic instability. 2. Despite progress, Kenya’s poverty rates remained high and the population continued to be vulnerable to shocks, undermining investments in human capital and limiting potential benefits of shared economic growth. The incidence of poverty declined from 52.2 percent in 2000 to 46.6 percent in 2005/06. Poverty was higher in rural areas (49.7 percent) and there were large regional differences, with 74 percent of Kenya’s poor residing in its northeastern counties. Poverty was higher among children, orphans and vulnerable children, the elderly, persons with disabilities, and those in female-headed households. Children living in households in the poorer quintiles were less likely than their better-off counterparts to attend school, advance to the next school grade, benefit from vaccination programs, and have good health outcomes. The 2011 drought in the Horn of Africa demonstrated the vulnerability of livelihoods, with poorer households experiencing the negative impact of shocks such as drought more often than others. 3. Kenya’s new Constitution (2010) aimed to move the country toward a more equitable and inclusive future. The Constitution focuses on protecting the rights of women, children, youth, disabled persons, the elderly, minorities, and marginalized groups and communities through a Bill of Rights that includes the right to social protection. This provides a robust foundation for social protection in Kenya. 4. Kenya had several social insurance schemes and safety net programs, but their coverage was low, and their effectiveness limited. The main form of safety net support for poor and vulnerable populations was humanitarian relief (often food aid), mobilized by the Government of Kenya (GoK) (the Government) and the international community in response to crises, such as droughts and floods. This type of response became common in the arid and semiarid lands (ASALs) where emergency relief was provided on a yearly basis to the chronically poor and food-insecure populations. 5. The Government’s expenditure on safety nets had risen rapidly and the coverage of cash transfer programs had grown significantly but was still low in relation to the country’s needy population. Spending on safety nets had almost doubled from 2005 to 2010 and was equivalent to 0.8 percent of gross domestic product (GDP) in 2010. At 53.2 percent of the total, emergency food aid still accounted for the largest share of net spending, but by the end of 2010 cash transfers increased from a low base to 25 percent of total safety net expenses. 1Throughout this report, the World Bank-financed National Safety Net Program for Results is referred to either as “the Program” or as the PforR; the overall Government-led National Safety Net Program is referred to as “the NSNP” or “the program”. Page 5 of 69 The World Bank National Safety Net Program for Results (P131305) 6. The five cash transfer programs faced several issues that needed to be addressed as these were likely to manifest as the programs continued to grow in scope. First, there were five cash transfer programs: (a) the Older Persons Cash Transfer (OPCT); (b) the Cash Transfer for Orphans and Vulnerable Children (CT-OVC); (c) the Hunger Safety Net Program (HSNP); (d) the Urban Food Subsidy Cash Transfer (UFS-CT); and (e) the Persons with Severe Disabilities Cash Transfer (PWSD-CT) (Table 1). But the programs tended to be small and fragmented. They were implemented by different entities and overlapped geographically. These different implementation arrangements led to inefficiency and duplication, and failed to optimize limited capacity or support a coordinated delivery of social assistance. Together, these five programs were estimated to reach 1.65 million people or 4 percent of the population. Still, cash transfers covered only a small fraction of the absolute poor (an estimated 9 percent, assuming perfect targeting), while food aid (although unpredictable and often late) supported twice the population targeted by transfers. Second, none of these programs was capable of expanding coverage in response to shocks, undermining their effectiveness in helping Kenya’s poor and vulnerable populations. Finally, the adequacy of the support provided to households was variable because the value of the cash transfers was rarely adjusted. Table 1: Five cash transfer programs: objectives, implementation responsibilities and characteristics Cash Implementing Implementing Implementation Transfer Objective Ministry Department characteristics Program to improve the welfare of poor households with Department of orphans and vulnerable children while supporting CT-OVC Children’s poverty reduction and income distribution among the Services poorest in society Both to strengthen the capacities of older people and departments improve their livelihoods by delivering cash transfers to Ministry of had dedicated OPCT poor households that included a member age 65 or Gender, staff at the older who was not already receiving a pension Children, and central and to increase access to food for vulnerable households in Social Department of county levels, urban poor communities in the short term and help Development Gender and but the numbers UFS-CT (MGCSD) were insufficient those households engage in longer-term food security Social and income-generating activities Development with skills mismatch to provide immediate relief from extreme poverty to PWSD- people with severe disabilities while enhancing their CT basic rights by providing them with a cash transfer every two months to reduce poverty in the drought-prone arid and Ministry of HSNP semiarid areas of northern Kenya by delivering regular Devolution and Implemented by HSNP Secretariat at cash transfers to extremely poor households in four Arid and Semi- NGOs NDMA counties Arid Land Areas 7. On the positive side, the programs’ operational systems were improving slowly, and best practices were emerging and being shared among the programs. Although the sophistication of their systems and the strength of fiduciary controls and governance varied, all five cash transfer programs had established systems and procedures for targeting, beneficiary registration, payments, and monitoring and evaluation (M&E). However, some payments in all five programs continued to be delayed, largely because of the lengthy process required to move funds through government systems. Two of the programs (CT- OVC and HSNP) had established accountability mechanisms, using a series of measures to develop strong Page 6 of 69 The World Bank National Safety Net Program for Results (P131305) ‘upward’ accountability (for example, fiduciary reviews, evaluations, internal controls on payments, and internal and external audits) and ‘downwards’ accountability (for example, community oversight committees, grievance and appeals mechanisms, and beneficiary service charters). Other programs were in the process of establishing these systems. 8. The Government’s National Social Protection Policy (NSPP), approved in May 2012, aimed to strengthen the delivery of social assistance to vulnerable populations, as mandated by the new Constitution. The NSPP stated that this would be achieved by: (a) ensuring that the design and implementation of programs was coordinated; (b) strengthening and scaling up existing social assistance programs; (c) putting in place the institutional frameworks to ensure consistent and adequate levels of support, and (d) conducting reviews based on standards agreed upon by stakeholders. The policy mandated the establishment of the National Social Protection Council, supported by the Social Protection Secretariat (SPS), which was already in place. As a first step, the Government was establishing its National Safety Net Program (NSNP, the program) and requested the World Bank’s support in view of its continuous engagement with social protection in Kenya. Theory of Change (Results Chain) 9. The Program Appraisal Document (PAD) did not describe a Theory of Change.2 For this Implementation Completion and Results Report (ICR), the Program’s Theory of Change is constructed ex post and has been retrofitted from the Program’s Results Framework (RF) and its description (Figure 1). 2According to the Operations Policy and Country Services guidelines, including the Theory of Change became mandatory for PADs as of May 2018. The PforR was approved by the World Bank’s Board of Directors in July 2013. Page 7 of 69 The World Bank National Safety Net Program for Results (P131305) Figure 1. Theory of Changea Outputs Intermediate Outcomes Intermediate Results Outcomes (Results Areas) • Prepare and implement an expansion plan. • Coverage expanded (IR 2, DLI 1) Cash transfers expanded to • Identify and enroll households. • Expansion among programs coordinated on the basis of objective promote more comprehensive and poverty and vulnerability criteria (IR 1) equitable coverage (PDO 2) • Establish baseline of beneficiaries. • Targeting and recertification process improved (PDO 3, DLI 2) • Strengthen implementation of targeting. • Develop a recertification plan. • Apply a harmonized targeting tool. • Introduce electronic payments with two-factor • Payments made in a timely fashion (PDO 4) authorization. (IR 4, DLI 4) • Improve the timeliness of disbursements to payment service providers (PSPs). (DLI 5) Program systems strengthened to ensure good governance Kenya has • Develop and implement a beneficiary outreach • Beneficiaries know program objectives and entitlements (IR 3) established an strategy (BOS). • Beneficiaries know how to make complaints (IR 5, 6, DLI 6-2) effective NSNP • Establish and maintain well-functioning complaints • Grievance and case management (G&CM) mechanisms are for poor and and grievance mechanisms at national and local levels. functioning at national and local levels. vulnerable (DLI 6-1) households • Adopt a single registry to share information on • Single registry fully operational and linked to IPRS. (IR 7, DLI 3) beneficiaries among five programs and with civil and registration system. • Standardize MIS’ internal payroll controls. • Internal controls on program payrolls improved. (IR 7) Consumption • Initiate government financing to the HSNP to promote • The NSNP increasingly financed by the Government (IR 8, 9) expenditure of the program’s financial sustainability. (DLI 9) beneficiary • Develop strategy for consolidating cash transfer • Strategy for consolidating cash transfer programs implemented. (DLI households programs. 8) increased (PDO 1) • Create capacity in the HSNP to scale up in response to • System for scaling up in response to shocks in place and financed by shocks. the Government (IR 10) • Develop a system for scaling up the NSNP as part of national drought risk management system. (DLI 7) Cash transfer programs harmonized • Develop a complementary services module in the • The NSNP households promoted access to complementary social to increase the coherence of the single registry. and productive services. (DLI 10) safety net sector • Design formal data sharing protocols for potential partners to access and use data. • Sign agreements with partners providing relevant complementary services to the NSNP beneficiary households. • Implement critical agreed financial management • Fiduciary systems for the NSNP have been strengthened. (DLI 11) (FM) actions. Note: IPRS = Integrated Population Registry Service; MIS = Management Information System. a. Bolded entries relate to changes to the Theory of Change following the additional financing (AF) and project restructuring. Page 8 of 69 The World Bank National Safety Net Program for Results (P131305) Rationale for PforR Support and Program Scope and Boundaries 10. The PforR was fully aligned with the World Bank Group’s Country Partnership Strategy (CPS) for Kenya for 2010–2013 (Report No. 52521 dated March 23, 2010), and the April 2012 CPS Progress Report (Report No 67224 dated April 5, 2012). The second pillar of the CPS aimed to address inequality and social exclusion by “establishing comprehensive and scalable social protection mechanisms,” and the Progress Report highlighted the World Bank’s (WB’s) intention to support the NSNP with a PforR instrument. A PforR was considered to offer several advantages over other instruments, including: (a) supporting the delivery of results in the sector; (b) creating incentives for the National Treasury and line ministries to make multi-annual budget commitments to the sector and ensure the timely allocation of funds, thus promoting sustainability of these programs; (c) continuing and expanding the use by cash transfer programs of government systems, thus strengthening these systems; and (d) harmonizing funding from many different sources, while promoting flexibility in the WB’s disbursements. Development Policy Financing would not provide the in-depth sectoral experience and program-specific dialogue, while a Specific Investment Loan would not be optimal in supporting the range of different cash transfer programs that existed within the NSNP nor support the shift to greater government ownership and financing. 11. The Government`s NSNP created a framework, around which the five cash transfer programs would be coordinated and harmonized; the PfoR boundary overlapped fully with the NSNP. The objective of the NSNP was to improve the welfare and resiliency of beneficiaries of these five programs, with the aim of reducing poverty and vulnerability in Kenya. It targeted poor and vulnerable households based on where they lived (the ASALs of northern Kenya or informal settlements in major urban centers) or because of their circumstances (households caring for orphans and vulnerable children, people with severe disabilities, and the elderly). Its objective was to be achieved by implementing activities that contributed to the following six results that are listed and described below according to the Program’s three Results Areas (para. 24). Program Results Area 1: Expanding cash transfer programs to promote more comprehensive and equitable coverage 12. First, the Government aimed to ensure that the poorest and most vulnerable households were enrolled in the NSNP, based on each of the transfer programs’ eligibility criteria supported by improvements in the implementation of the targeting. The Government sought to share functions across transfer programs as implementation of the NSNP progressed. All these programs targeted resources geographically and to eligible households within identified areas. Although assessments found that the programs were generally well targeted, some households that were eligible were not covered due to limited program funding, and targeting of the OPCT, PWSD-CT, and UFS-CT programs was poorly implemented due to limited operational resources. The NSNP aimed to improve targeting through revisions to a joint, revised Operations Manual (OM) that clarified and strengthened procedures to: (a) identify households that were eligible for support; (b) manage beneficiary waiting lists; and (c) retarget/recertify beneficiaries, which would be carried out in response to finding of irregularities in targeting (or evidence that households no longer required support) through the M&E system or complaint and grievance mechanisms. Additional measures included: (a) strengthening the HSNP’s targeting procedures through adoption of a proxy means test (PMT) together with community-based targeting; (b) providing each program with a dedicated operating budget starting in FY13/14 to cover, among others, Page 9 of 69 The World Bank National Safety Net Program for Results (P131305) the targeting process; and (c) strengthening the workflow in programs’ MISs from targeting of households to enrollment. Program Results Area 2: Strengthening program systems to ensure good governance 13. Second, the Government aimed to improve the predictability of payments to ensure that the NSNP beneficiaries received appropriate, reliable, and accessible payments. Although all five transfer programs were to deliver payments every two months, analysis revealed that these were often delayed due to: (a) the time it took to carry out reconciliations at the end of the payment cycle using the paper- based system, and (b) delays in the release of funds from the National Treasury (NT) to the MLSP. Payments for all programs were made through two independent PSPs: (a) the Postal Corporation of Kenya (PCK), which made payments to about 200,000 households under the CT-OVC, OPCT, PWSD-CT, and UFS- CT programs, and (b) Equity Bank, which made payments to about 80,000 households under the CT-OVC program and HSNP. Each of the PSPs applied different payment mechanisms: PCK delivered payments at post office branches, based on a one-factor authentication process using the National or Program ID, while Equity Bank delivered payments against a point of sale through a two-factor authentication process using the smart card and biometric data. Both the cost per transfer and cost as a percentage of the transfer value were higher for those made by Equity Bank. The Government intended to address payment issues by ensuring that future contracts with PSPs: (a) complied with the country’s new Anti-Money Laundering Act that required PSPs to be regulated financial service providers; (b) used mainstream payment infrastructure; and (c) included performance standards on the timely delivery of transfers to beneficiaries. The capacity of the MLSP’s SPS and four programs was strengthened to better manage contracts with then existing PSPs and identify and address constraints in the flow of funds through government systems. 14. Third, the Government aimed to ensure that all citizens were able to appeal and complain to improve program performance through a functional complaint and grievance mechanism (CGM) that strengthened demand-side accountability. Two of the programs, the HSNP and CT-OVC, had established CGMs and means of informing communities about the mechanisms’ existence and rules. Building upon the experience of the CT-OVC, the MLSP designed a CGM for the OPCT, PWSD-CT, and UFS-CT, which used a harmonized complaint and grievance system as documented in revised OMs. Locational committees (at the local level to conduct awareness raising on CGMs) and Beneficiary Welfare Committees (responsible for supporting program implementation once beneficiaries are enrolled, informing them of payments and forwarding any complaints to program managers) were being established and sensitized on how to file complaints. A hotline was being developed and operated and the three programs were formulating service charters to inform beneficiaries of their rights and entitlements. All the programs were in the process of developing a dedicated module in the MIS for recording CGMs, lodged at the national level or escalated to national program managers. This would allow for consolidation of CGM reporting for sharing with the WB at six-month intervals. The NSNP’s monitoring framework would monitor the functioning of the CGMs and the time taken to provide feedback on complaints. 15. Fourth, the Government aimed to have a fully functioning monitoring and learning system in place. The Government developed an M&E system for the NSNP that included improved transfer program MISs and was developing a single registry. The M&E framework included several indicators to monitor the Program’s progress, and the five programs’ MISs were being enhanced to have the capability t o produce regular monitoring reports. The programs’ existing MISs varied in complexity and efficiency. Therefore, the HSNP’s MIS was adapted for use by all the other programs, thereby creating common standards and Page 10 of 69 The World Bank National Safety Net Program for Results (P131305) functionality across the NSNP’s systems, including automatic links with the single registry, a common database within which information on the beneficiaries of all transfer programs were to be consolidated. The single registry that was under development at the time would: (a) provide policy makers with a consolidated source of information on cash transfer programs; (b) allow for cross-check beneficiary lists to reduce duplication across multiple programs; (c) strengthen the process for verifying beneficiaries’ identities by linking the single registry with the central civil registration system, IPRS; and (d) be a foundation for creating other common delivery systems, including payment systems. The OPCT, PWSD- CT, UFS-CT, and CT-OVC had migrated their data on beneficiary households to their program’s new MISs and had identified a number of households with duplicate names or ID numbers that were being investigated. The Government introduced a requirement that all beneficiary households needed to include at least one adult with a valid national ID for enrollment, allowing the programs to carry out cross- checks with the single registry (when it became operational) and with the IPRS. Program Results Area 3: Harmonizing cash transfer programs to increase the coherence of the safety net sector 16. Fifth, the Government was taking steps to ensure that the cash transfer sector was harmonized and government managed. The Government planned to carry out a functional review to inform consolidation of the four transfer programs managed by MLSP and formulate a plan for capacity building to the NSNP. The review would assess the current capacity of the programs to deliver on their objectives. The Government decided to move the financing of all the five cash transfer programs on-budget and begin financing the HSNP from its general revenues. The FY13/14 budget already included funding from the United Kingdom’s Department for International Finance (DFID) and its own contribution to the HSNP. 17. Sixth, the Government also planned to develop capacity to scale up the NSNP to respond to shocks. The NDMA and the National Drought Emergency Fund (NDEF), which was to be established as a designated source of contingency financing, would provide the foundation for this approach, including an established early warning system and local contingency plans. The NSNP could use the contingency funds to extend cash support to the population affected by shocks once an agreed set of early warning indicators was triggered. The Government intended to focus its initial efforts on the HSNP because it was managed by the NDMA and operated in drought-prone areas of the country. 18. The Government aimed to gradually expand the coverage of the NSNP to achieve the Program’s objective, and this presented several challenges. At appraisal, it was expected that the NSNP’s coverage would increase slowly, almost doubling from 1.65 million to 3.30 million people, in five years, at which time it would cover 17 percent of the country’s absolute poor and 8 percent of its population (assuming perfect targeting). To reach these targets, the Program’s financing would need to double, and there were strong commitments from the Government and its development partners (DPs)--DFID, Swedish International Development Cooperation Agency [SIDA], United Nations Children’s Fund [UNICEF], and the WB--to mobilize the required financing. The Government prepared a Medium-Term Expenditure and Financing Framework (MTEFF) and secured the agreement of DPs to ensure complementary resources. To ensure the NSNP’s long-term sustainability, the Government had committed to assume funding provided by DPs while increasing the number of Program beneficiaries. Although this would represent a more than fivefold increase in Government funding to the NSNP, from 39 percent of the total NSNP financing in FY12/13 to 85 percent in FY17/18, the additional funding was projected to require a small percentage of the Government’s annual additional revenue arising from economic growth. The Page 11 of 69 The World Bank National Safety Net Program for Results (P131305) Government had begun to formulate an expansion plan to guide the five programs’ coverage increases, with the program managers using the latest poverty and vulnerability data to identify new beneficiary households. This process was to be coordinated by the SPS to ensure harmonization and be based on the MTEFF to secure financing for expansion. The NSNP envisaged several activities to build the Government’s capacity to implement a significantly increased program, including those that would help improve the delivery of cash transfers while making the system more efficient by enhancing coordination and harmonization: (a) the adoption of a common M&E framework among the programs; (b) the rollout of a common MIS, with automatic links to the civil registry; (c) the adoption of a single OM for the OPCT, PWDS-CT, and UFS-CT and common implementation arrangements at the local level; and (d) the consolidation of a number of core processes, such as the complaint and grievance redress. 19. The SDR 166.9 million (US$250 million equivalent) credit was prepared and approved in response to the Government’s request to the WB for assistance in helping it achieve the key results required to deliver the objectives of the NSNP. The three assessments carried out during preparation of the PforR and the WB’s policy dialogue with the Government highlighted that to achieve this goal, the Government would need to also strengthen implementation and oversight of the five programs that constitute the NSNP to ensure good governance and increased coordination and thus improve sectoral harmonization. Program Development Objectives (PDOs) 20. To support the Republic of Kenya’s to establish an effective national safety net program for poor and vulnerable households.3 Key Expected Outcomes and Outcome Indicators 21. Progress toward achieving the PDO would be measured by the following outcome indicators detailed in the NSNP’s M&E framework: • Net change in the monthly consumption expenditures among program beneficiary households (reflecting the positive impact of the program on beneficiaries’ welfare); • Number of NSNP beneficiaries, disaggregated by sex (reflecting the increased coverage of the NSNP); • Percent of NSNP beneficiaries who conform to program targeting criteria (reflecting the improvements in the implementation of targeting methods); and • Percent of payments disbursed to payment service providers on time (reflecting the delivery of timely payments to beneficiaries). Program Results Areas and DLIs 22. The PforR was to focus on achieving results in the three results areas encompassing nine disbursement-linked indicators (DLIs). The DLIs were selected from the NSNP’s M&E framework and from 3This was the PDO as described in the PAD. In the Financing Agreement, the PDO was described as “to establish an effective national safety net program for poor and vulnerable households.” Page 12 of 69 The World Bank National Safety Net Program for Results (P131305) the action plans from the three assessments carried out during preparation. The nine DLIs, by results area, were as follows: • Results Area 1 - Expanding cash transfer programs to promote more comprehensive and equitable coverage o DLI 1: Number of additional households enrolled in the NSNP according to the expansion plan • Results Area 2 - Strengthening program system to ensure good governance o DLI 2: Percent of program beneficiaries who conform to the targeting criteria for the program in which they are enrolled o DLI 3: Single registry is fully operational with program MISs using agreed standards for internal payroll controls o DLI 4: Percent of NSNP payments made electronically using two-factor authentication o DLI 5: Percent of payments disbursed to Payment Service Providers on time o DLI 6: Functional complaint and grievance mechanisms and beneficiary awareness of transfer amount and complaint and grievance mechanisms • Results Area 3: Harmonizing cash transfer programs to increase the coherence of the safety net sector o DLI 7: System for scaling up the NSNP as part of the national drought risk management system o DLI 8: Strategy for consolidating the cash transfer programs (adoption and implementation) o DLI 9: The Government finances the NSNP in line with policy commitments. B. SIGNIFICANT CHANGES DURING IMPLEMENTATION 23. An additional credit of SDR 36.8 million (US$50 million equivalent, including US$35 million from the IDA Crisis Response Window) was approved on April 27, 2017. The Program Paper for the additional credit (the Program Paper) also included a project restructuring. Revised PDOs, Outcome Targets, Results Areas, and DLIs 24. The PDO was revised in the Program Paper to read “To support the Republic of Kenya’s efforts to establish an effective national safety net program for poor and vulnerable households.” The Program Paper proposed a project restructuring that adjusted the outcome targets to reflect revisions to outcomes expected in view of the AF and the two-year extension of the closing date as well as implementation experience and streamlined RF. The specific revisions introduced in the restructuring are provided in annex 6. The PforR’s three results areas were not revised. The Program Paper introduced two new DLIs, expanded one DLI, and modified three DLIs. As implementation progressed, it became necessary to revisit some of the DLIs. The Government began facing challenges around measuring improvements in the DLIs related to targeting conformity (DLI 2) and timeliness of payments (DLI 5). DLI 2, originally ‘percentage of program beneficiaries who conform to the targeting criteria for the program in which they are enrolled’ was modified to ‘system for recertification of NSNP beneficiaries is in place’. Although ensuring that the right people were benefiting from the NSNP continued to be a priority, DLI 2 needed adjustment in view of: (a) the massive expansion of the NSNP that took place before the baseline survey of the Program Page 13 of 69 The World Bank National Safety Net Program for Results (P131305) Implementation and Beneficiary Satisfaction (PIBS) was conducted (and the PIBS survey was required to measure achievement of the DLI), and (b) the conformity to targeting criteria measured by the PIBS baseline survey that proved much higher than originally contemplated and significant improvements were not expected because another massive nationwide expansion of the Program was not planned. Because of this, having in place a system that would regularly recertify beneficiaries to ensure up-to-date beneficiary registry with only eligible households was considered an acceptable replacement. DLI 5, ‘percentage of payments made on time to PSPs for transfer to Program beneficiaries’ remained the same and continued to monitor the timeliness of payments, but minor modifications were introduced to the DLI protocol and to the baseline and target (which increased from 65 percent to 75 percent with a new baseline of 12 percent).4 Finally, DLI 6, originally ‘functional Complaints and Grievance mechanism’ was modified to ‘Grievance and Case Management (G&CM) Mechanism for the Cash Transfer Programs are functional’ to recognize the Government’s priority of including case management (such as updates to beneficiary status) as a part of this function. The protocol for this DLI was also modified to enable decentralized recording of the G&CM updates through the MISs. The revisions to the DLI matrix are provided in annex 6. Other Changes 25. The four original programs under the responsibility of the MLSP (CT-OVC, OPCT, UFS-CT, and PWSD-CT) were not cited in either the PforR’s outcome targets, results areas, or DLIs but were considered within the scope of the NSNP. At the beginning of 2014, the MLSP decided to streamline its cash transfer programs by eliminating the UFS-CT, thereby prioritizing the remaining programs.5 The PforR’s closing date was extended by two years to June 30, 2020, together with the AF. The closing date was subsequently extended to December 31, 2020, in a project restructuring dated June 29, 2020. A total of US$50 million equivalent was cancelled from the undisbursed credit amount towards the Program closing date. Rationale for Changes and their Implication for the Original Theory of Change 26. The additional credit of SDR 36.8 million was approved to help the Government respond to the ongoing drought emergency in early 2017, build households’ resilience in the long term, and strengthen systems for more efficient implementation. The additional credit focused on improving safety net support to the households living in the most drought-affected counties: the north and northeastern region of Kenya that was already lagging behind in development outcomes. The revisions to the RF and DLI matrix reflected this new focus and the need to improve FM through compliance with newly agreed actions. The implications on the original Theory of Change are highlighted in bold in figure 1. The adjustment to the PDO was not substantial; it only adjusted the PDO of the original Financing Agreement to be consistent with the PDO in the PAD. The DLI matrix was adjusted to include and adapt the existing DLIs to the expanded scope of the program, and also to modify some of the existing DLIs for which the measurement 4 The main challenge with DLI 5 related to delays in submitting funding requests from the MLSP to the National Treasury, and subsequent delays in the release of funds from the National Treasury, which then delayed the overall payment cycle. Both agencies were working together to develop and implement a solution that would ensure timeliness of payments, and this, together with the monitoring of timeliness of payments throughout the Program’s life, was included as an additional action in the PAP. Therefore, for the DLI, a payment would be considered ‘on time’ if funds and payroll documentation were sent to the PSPs within 15 days of the due date (defined as the five working days before the start of the payment window to incentivize the Government to continue to make efforts to make payments as soon as possible after the intended date). The DLI would be considered as achieved if transfers to PSPs were made ‘on time’ for three consecutive cycles. 5 Transfers under the UFS-CT ceased in June 2014. Households that had been enrolled in the cash transfer program could be enrolled in another program provided that they were found to be eligible during a targeting exercise. Page 14 of 69 The World Bank National Safety Net Program for Results (P131305) of results posed challenges (section IV D). The rationale for closure of the UFS-CT was to consolidate beneficiaries among the MLSP’s three larger programs. The closing date was extended to provide additional time to implement the AF, and the restructuring that accompanied it was processed to reflect the additional two-year implementation period. The second closing date extension was approved to provide the Government with additional time to complete the Program’s activities in view of the impact of the COVID-19 pandemic on its implementation (section III B). The cancellation of the undisbursed credit proceeds in December 2020 reflected amounts allocated to DLIs that were not yet achieved at the time of the onset of the COVID-19 pandemic, but that did not affect the likelihood of achievement of the PDO, and could be reallocated toward urgent needs – in this case Additional Financing for the COVID-19 Health Emergency Response Projects – in line with the WBG COVID-19 Approach Paper recommendations. II. OUTCOME A. RELEVANCE Relevance of PDO 27. The relevance of the PDO is rated High. The PDO was relevant to the CPS for Kenya for 2010– 2013 and the April 2012 CPS Progress Report when it was approved, and continued to be relevant to the WB’s CPS for Kenya for FY14–FY18, and the CPS Performance and Learning Review FY20 throughout implementation and upon completion6. The FY14–FY18 CPS proposed WB support to the Government’s National Safety Net Policy and its NSNP that was designed to help achieve the policy’s goals, under a PforR instrument. The FY14–FY18 CPS, under its Domain two ‘Protection and potential—delivering shared prosperity’, specifically mentions that the “WBG will continue its strong engagement in social protection, and the CPS’ results matrix included two indicators to measure progress towards this goal: people covered by social safety nets (number) and beneficiaries for whom payments are made electronically using two factor authorization (percentage).” It also indicates that “WBG stands ready to supplement resources should that be merited in the latter part of the CPS period.” In approving the AF and later extending the closing date, the WB’s implementation support was responsive to emerging needs, first by providing additional funding to households affected by the drought and later providing additional time required to complete final activities in view of the COVID-19 pandemic. A €215.9 million (US$250 million equivalent) credit from IDA was subsequently approved on November 27, 2018 for the Kenya Social and Economic Inclusion Project (KSEIP, P164654), providing continuity to the initiatives begun under the Program and reinforcing its relevance. Relevance of DLIs 28. The relevance of DLIs is rated High. The DLIs were well articulated, each with a clear definition and protocol to evaluate achievement (with data source, verification entity, and procedure). They were reasonable in terms of expected achievement during implementation and clearly identified in terms of scalability. The DLI matrix was linked closely to the RF (in some cases overlapping). It included DLIs that recognized completion of activities and outputs required to further progress toward achieving the PDOs and key indicator targets for each of the Program’s three results areas (for example, DLI 8: strategy for consolidating the cash transfer programs, DLI 3: single registry is fully operational with program MISs using 6 World Bank, Kenya Country Partnership Strategy for FY14-FY18, Report No. 88940 dated June 2014. Page 15 of 69 The World Bank National Safety Net Program for Results (P131305) agreed standards for internal payroll controls, and so on) and DLIs that recognized progress that had been achieved (for example, DLI 1: number of additional households enrolled in the NSNP according to the expansion plan and DLI 2: percent of program beneficiaries who conform to the targeting criteria for the program in which they are enrolled). A matrix aligning the PDO and intermediate outcome indicators with the DLIs, both as appraised and as restructured, is provided in annex 6. 29. The phasing of results for the DLIs included several institutional milestones needed to achieve the PDO, the disbursement amounts and timing were calibrated to the importance and reasonably expected time required for achievement of the DLI targets. The DLIs provided strong incentives for the achievement of results, which in turn helped solidify the institutional backbone of the NSNP. Examples of such DLIs include: DLI 1: expansion plan adopted; DLI 3: MIS operational; DLI 6: functional C&GM; and DLI 8: strategy for consolidating cash transfer programs adopted and implemented. Finally, the DLI matrix maintained its relevance throughout implementation: it was adjusted in the AF to coincide with changes in the RF that were introduced to address emerging priorities on target beneficiaries (for example, DLI 1d, 1e(i) and 1e(ii)—reflecting expansion in the NEDI7 counties) and pressing needs to improve institutional performance for the NSNP (for example, DLI 11: fiduciary systems for NSNP have been strengthened). Rating of Overall Relevance 30. The Program’s overall relevance is rated High, based on similar ratings for relevance of the PDO and DLIs. B. ACHIEVEMENT OF PDOs (EFFICACY) 31. As the PDO statement was broad and lacked definition (Section IV A), the achievement of the PDO--to support the Republic of Kenya’s efforts to establish an effective NSNP for poor and vulnerable household--is assessed according to progress attained in each of the PforR’s three results areas and PDO Indicators.8 The evaluation of achievement of PDOs is not prepared following a split rating for the following reasons: (a) the Program’s scope was expanded; (b) the PDO was revised but only so that the definition in the Financing Agreement and PAD coincided; (c) the RF was restructured, but this was to streamline and adjust it to additional activities under the AF and to eliminate redundancies in indicators, and (d) the DLI/disbursement-linked result (DLR) matrix was also adjusted, again to reflect AF and implementation experience. 32. The Program provided a continuum of WB support for Kenya’s safety net system and building upon earlier efforts produced important results in establishing an effective NSNP for poor and vulnerable households that is now being supported under KSEIP. The earlier CT-OVC Project provided an initial effort to support this system. Under the second phase, the PforR helped the Government to assume financing of its safety net making this one of the few, if not only, safety net programs in Africa that is fully 7 Ten counties in northeast Kenya where poverty is significantly higher than the national average had been the most affected by the drought and famine in late 2016. The World Bank was working to identify how best to support this region through a multisectoral response under the North Eastern Development Initiative (NEDI), and the AF to the NSNP would be one of the operations to help by increasing coverage of the NSNP in these 10 NEDI counties. A completion explanation is provided in section III B. 8 The PDO and intermediate indicators related to each of the Results Areas are presented in boxes before the discussion of progress in each area. DLIs are also disaggregated by Results Area, but because of space limitations they are presented in more complete boxes that appear in Annex 8. Progress towards achievement of PDO indicator 1 targets is discussed separately from any of the Results Areas (Section IV A). Page 16 of 69 The World Bank National Safety Net Program for Results (P131305) financed (and owned) by the Government. During this phase, the Government successfully: (a) tripled coverage of cash transfers; (b) harmonized policy and program delivery through a consolidated strategy, including a streamlined targeting tool, payment system, and grievance management; (c) put in place a single registry of NSNP beneficiaries; (d) piloted a harmonized targeting tool for future targeting and recertification; (e) launched a new electronic payment system with two-factor authorization; (f) consolidated MISs for three transfer programs managed by MLSP;9 (g) consolidated a G&CM system and a Beneficiary Outreach Strategy (BOS); (h) institutionalized the financing of cash transfers within the Government; and (i) produced positive impacts, as reported by beneficiary households, on consumption, dietary diversity, school attendance, and performance and health. In 2018, 80 percent of beneficiaries consulted through the PIBS reported their satisfaction with the NSNP and its delivery systems. The PforR has laid the basis for the next phase of assistance of this Government-owned program that is being supported by the WB’s KSEIP (Section IV D). Assessment of Achievement of Each Objective or Outcome PDO Indicator 1: Net change in beneficiary household monthly per adult equivalent consumption expenditures (Baseline 2013=0.0; Target 2020=150.0; Achieved 2020=66.0) 33. The beneficiary household monthly per adult equivalent consumption expenditure increased by Ksh 66, against a target of Ksh 150 (PDO 1). Progress toward this target was to be measured through impact evaluations of the CT-OVC program and HSNP which, given the size of these programs, were assumed to be representative of the entire NSNP. Unfortunately, the results from the CT-OVC impact evaluation showed that, due to the quality of the consumption data collected for the end line and other reasons, it was difficult to accurately measure the impact of that program on household food consumption. The propensity score matching results of the HSNP impact evaluation confirmed an impact on monthly food expenditures, showing that households that had received a regular HSNP payment experienced an increase in monthly per adult equivalent food expenditure of around Ksh 66.10 Because of the spillover effects observed in the HSNP, the impact on consumption for its beneficiaries is likely underestimated. The evaluation estimates that for every Kenya shilling injected into the local economy by the HSNP, total incomes increase by somewhere between Ksh 1.38 and Ksh 1.93. Most of this additional income goes to non-beneficiaries of the program, hence producing significantly positive spillovers. 34. Nevertheless, the 2018 Poverty and Gender Assessment11 shows that the NSNP cash transfers are well targeted and have positive impacts on the beneficiaries` welfare outcomes. The assessment reports a significant and positive impact on child school enrollment. A total of 93.5 percent of households containing school-aged children (ages 6 to 16 years) had all those children enrolled in 2015/16, and the receipt of a cash transfer increased already high rates. The impact of a household being the beneficiary of any cash transfer program was an increase in the probability of all children being enrolled of about 1.4 percentage points; the CT-OVC had the largest impact, with a positive effect of three percentage points. Cash transfer programs also increased the probability that no child in a beneficiary household was working, with the OPCT having a close to seven percentage point increase. Finally, the HSNP transfer 9 The three programs under the MLSP are CT-OVC, PWSD-CT, and OPCT. The UFS-CT was discontinued in 2014. 10 Merttens et al (2017) Evaluation of the Kenya Hunger Safety Net Programme Phase 2, Impact Evaluation Final Report, Oxford Policy Management. 11 World Bank. 2018. Kenya Poverty and Gender Assessment 2015/16 Reflecting on a Decade of Progress and the Road Ahead. Report No. AUS0000466. Page 17 of 69 The World Bank National Safety Net Program for Results (P131305) recipient households had a 2.5 percentage point higher probability of being food secure as self-reported by them.12 35. The assessment also finds that the adequacy of transfers is low, and that the real value of transfers was decreasing over time. Since the launch of the NSNP, only HSNP benefits have been adjusted to account for inflation. The monthly benefit payments are equal to Ksh 2,700 for HSNP beneficiaries. The monthly payments under the three MLSP programs, collectively referred to as the Consolidated Cash Transfers Program (CCTP), were last adjusted in December 2011 from Ksh 1,500 to Ksh 2,000. As a result, the real value of transfers has eroded significantly over the life of the NSNP. When the Government assumed financing of the four cash transfer programs, the adjustment of payment amounts became difficult in view of fiscal constraints, and a trade-off between the increase in benefit size and coverage. Results Area 1: Expanding cash transfer programs to promote more comprehensive and equitable coverage Box 1. Results Chain for Results Area 1 PDO 2: Number of NSNP beneficiaries, by sex (Baseline 2013=1.65 m; Target 2019=4.28 m; Achieved 2020=4.08 m) IR 1: New households enrolled in the NSNP according to agreed expansion plan and paid for by GoK (Baseline 2017=114,384; Target 2019=792,000; Achieved=1.195 m) IR 2: Number of households on the NSNP payroll (deleted) 36. The Government made notable progress in expanding the coverage of cash transfer programs according to the expansion plan, but the expansion of the OPCT to individuals over 70 years helped increase the number further. An important output of the NSNP was to develop an expansion plan that would set out how each of the programs would expand during implementation. Using data from the 2009 census and 2005/06 KIHBS, the Government produced small area poverty estimates and estimated the number of beneficiaries who would meet categorical criteria of the cash transfer programs. Applying the above analysis, the Government adopted an expansion plan in October 2014, which outlined the expected expansion of each of the programs over 2014–2017. As the Government was finalizing the plan for approval, it proceeded with a large-scale expansion of the NSNP from 2013 to 2014. The expansion of about 220,000 beneficiaries resulted in an almost doubling of the coverage of the three programs managed by the MLSP. In 2014-2016, the OPCT, PWSD-CT, and CT-OVC programs expanded by about 130,000 beneficiaries, 8,000 beneficiaries, and 150,000 beneficiaries, respectively (and the UFS-CT program was discontinued). In the third phase, supported by the AF, the Government prepared a revised expansion plan in which it committed to expand the NSNP to an additional 120,000 beneficiaries in ten of the NEDI counties in 2017-2020. But the progress toward enrolling additional households has not been as successful. The registration process for enrollment of additional households (DLI 1e) was interrupted first because of the lack of funding and then by COVID-19 pandemic restrictions. Importantly, in mid-2017 the Government introduced a universal program for citizens aged 70 years and above under the OPCT.13 The NSNP expansion resources were used to fund that initiative. Over 523 thousand new beneficiaries or 12 The assessment reports food security based on a number of self-reported food adequacy questions in the Kenya Integrated Household Budget Survey (KIHBS) 2015/16 household questionnaire. A household is defined as food insecure if, in the last 12 months before being interviewed, members either missed meals, ran out of food, went hungry, or went without food for a whole day because of a lack of money/resources. 13 The previous beneficiaries were ages 65 years and above and could remain in the OPCT program. Under the revised program, individuals, not households, were enrolled in the program. As a result, more than one individual could be registered and enrolled. Page 18 of 69 The World Bank National Safety Net Program for Results (P131305) about 75 percent of persons 70 years and older are estimated to be enrolled during OPCT expansion. With this, the PforR fell only slightly short of the target number of NSNP beneficiaries (PDO 2), but it achieved this by targeting a different group of vulnerable population. As of June 30, 2020, over 1.19 million households were benefiting from the NSNP (table 2), which corresponds to an estimated 4.09 million individual beneficiaries. 14 Table 1. Number of NSNP Beneficiary Households Program Baseline Jun-14 Jun-15 Jun-16 Jun-17 Jun-18 Jun-19 Jun-20 UFS-CT 9,268 9,387 - - - - - - CT-OVC 152,331 249,486 225,984 355,469 349,601 340,087 287,278 293,867 OPCT* 58,113 162,604 157,044 315,454 314,504 783,089 753,314 766,424 PWSD-CT 14,518 24,963 25,505 47,460 46,917 42,633 29,584 34,032 HSNP 68,261 43,944 75,869 87,568 98,492 100,630 100,132 100,573 TOTAL 302,491 490,384 484,402 805,951 809,514 1,266,439 1,170,308 1,194,896 37. There has been a steady increase in Government financing of cash transfers. In mid-2016, the Government assumed financing of 40,000 CT-OVC beneficiaries who had previously been financed by DFID. Shortly thereafter, the Government assumed financing of a further 11,000 HSNP beneficiaries. For OPCT expansion implemented in 2017, additional Ksh 9.3 billion was allocated by the Government to provide cash transfers to persons aged 70 years and above. Finally, in FY19/20 the Government assumed financing of all HSNP beneficiaries, or over 100,000 households. In all, given the NSNP’s expansion, and increases in its financing, the Government was financing 1,194,896 households at end-2020, exceeding the target of 792,000 (Intermediate Result 1). Increased Government financing of the cash transfer programs contributes to their institutionalization and longer-term sustainability as this financing comes from general revenues and is reflected in the national budget. 38. Achievement of the expected outcomes under Results Area 1 is rated Substantial. The Government increased the coverage of cash transfer programs from 1.65 million beneficiaries at baseline to 4.09 million in June 2020, falling only slightly short of the targeted 4.28 million beneficiaries. Furthermore, the Government has gradually assumed financing of the NSNP beneficiaries to significantly exceed the target by closure of the operation. The one area that was less successful concerned the expansion of NSNP in the NEDI counties. Results Area 2: Strengthening program system to ensure good governance Box 2. Results Chain for Results Area 2 PDO 3: % of Program beneficiaries who conform to targeting criteria (deleted) PDO 3 (revised): System for recertification of NSNP beneficiaries is in place (Baseline 2017=No; Target 2019=Completion of the rollout of the recertification plan in at least 20 sub-counties; Achieved 2020=Recertification Plan has been adopted by not rolled out) 14 The number of individual beneficiaries is estimated based on the July–August 2020 payroll and considering the average household size, as the program MIS does not include information on the composition of beneficiary households. To arrive at these estimates, households in the CT-OVC and PWSD-CT were multiplied by 5, as were the OPCT beneficiary households enrolled before 2016. Those enrolled for the 70+ benefits are treated as individuals. The HSNP beneficiaries are multiplied by 7 to account for larger household sizes (payroll). Page 19 of 69 The World Bank National Safety Net Program for Results (P131305) PDO 4: % of all payments disbursed to payment service providers on time (Baseline 2017=12; Target 2020=90.0; Achieved 2020=19.47) Intermediate Result 3: % of beneficiaries who know correct program transfer amount for the program in which they are enrolled (Baseline 2017=56.3; Target 2019=86.0; Achieved 2020=70.0) Intermediate Result 4: % of program beneficiaries who can name two means of making a complaint (Baseline 2013=8.8; Target 2019=25.0; Achieved 2020=40.0) Intermediate Result 5: Grievance and Case Management mechanisms for Cash Transfer Program are functional (Baseline 2013=Not in place; Target 2020=BOS implementation completed in 30 sub-counties. G&CM module decentralization done; Achieved 2020=BOS implemented in at least 20 counties and G&CM module decentralized in 47 counties) Intermediate Result 6: % of complaints actioned (deleted) Intermediate Result 7: Single registry functional with program MISs using agreed standard for internal payroll controls (Baseline 2013=No; Target 2017=Yes; Achieved 2020=Yes) Intermediate Result 12: Fiduciary system for NSNP have been strengthened % (Baseline 2017=0; Target 2019=100; Achieved 2020=100) 39. Overall, targeting accuracy of cash transfer programs is good. Before consolidation of programs managed by the MLSP, each program followed different poverty targeting procedures in addition to the categorical criteria and community-based targeting, whereby communities themselves decide the beneficiaries. The Program Implementation and Beneficiary Satisfaction (PIBS) surveys carried out in 2016 and 2018 found that the four programs’ approach to targeting showed a strong compliance with categorical criteria and reasonable compliance with each of the programs’ poverty targeting procedures. The PIBS baseline survey (DLI 2 (i)) concluded that the programs’ targeting performance was relatively good, with an average of 82.4 percent of the NSNP beneficiaries falling below the harmonized PMT in 2017/18.15 40. A system for recertification of the NSNP beneficiaries is in place (PDO 3, DLI 2): a harmonized targeting tool was developed and piloted, and a recertification plan was adopted in 2019, but budgetary constraints have delayed the plan’s rollout. The new Harmonized Targeting Methodology (HTM) was developed to (a) have one tool to target all NSNP cash transfers; (b) coordinate the targeting process; (c) eliminate or minimize duplications across programs; (d) reduce errors of inclusion and exclusion; and (e) synchronize the approach to targeting, including for data collection. The HTM comprises registration and data collection for carrying out the PMT based on a Household Living Condition Score constructed using the Kenya 2009 Census data component analysis,16 community validation, and enrollment. It was piloted in three counties in 2016/17. A Government review of the HTM found that it was reasonably accurate in identifying households at the bottom of the welfare distribution. The HSNP applied the HTM to its ongoing re-registration exercise that started in 2017. A system for recertification of the NSNP beneficiaries is in place—the plan on how all four NSNP programs will maintain an up-to-date registry of beneficiaries through recertification was adopted in 2019 (DLI 2 (ii)). In addition, the MLSP has piloted a successful recertification for the OPCT based on the HTM. However, budget constraints have precluded the completion of the recertification rollout plan in at least 20 sub-counties (DLI 2 (iii)). 41. The single registry of beneficiaries under the NSNP that is linked to the IPRS became operational in 2015 (DLI 3 (ii)), supported by an integrated CCTP-MIS. The single registry of beneficiaries was under preparation at appraisal, housed in the MLSP, but it was not connected with individual program MISs. 15 By program, the percentage of beneficiaries below the cutoff of the harmonized PMT was as follows: CT-OVC, 81.4 percent; PWSD, 78.6 percent; OPCT, 85.3 percent; HSNP, 84.4 percent. 16 The PMT formula has been modified in 2020 to improve the targeting accuracy. Page 20 of 69 The World Bank National Safety Net Program for Results (P131305) Therefore, it was not possible to share beneficiary information among programs or provide timely and accurate monitoring. Under the PforR, the single registry was linked to the IPRS and program MISs, accessible at both the national and field levels. This linking has: (a) provided policy makers with a consolidated source of information on cash transfer programs; (b) permitted cross-checking of beneficiary lists to eliminate the duplication of benefits; (c) strengthened the verification of beneficiaries’ identity by linking with the IPRS; and (d) supported the creation of common delivery systems (for example, payment systems) thereby reducing delivery costs. The establishment of the Social Assistance Unit (SAU)17 in MLSP in 2016 improved coordination among the programs and facilitated the design and development of the CCTP-MIS, which is fully functional and strengthens internal controls such as payroll, beneficiary IPRS validation, funds reconciliation (Intermediate Result 7). The new HSNP MIS was significantly improved to support re-targeting, pre- and post-payroll checks, and a new approach to complaints and case management. It is being developed further to support M&E, including analysis of data. 42. A system of technology-based payments was introduced in 2015 that allowed beneficiary households to receive their cash transfers electronically under a two-factor authorization system, and this has been improved to allow beneficiaries to choose among payment providers. The HSNP had previously contracted payments to a commercial bank under a two-factor authorization system (with a debit card and either fingerprints or a PIN), as had the CT-OVC for some of its beneficiaries. By 2015, 100 percent of NSNP beneficiaries received payments through one of two commercial banks under a two- factor authorization process (DLI 4). This brought benefits to the NSNP recipients: by 2018, 84.9 percent of beneficiaries had to travel less than 5 km (20 km for the HSNP) to a pay point. A further four commercial banks signed contracts with the MLSP, and beneficiaries were allowed to choose service provider based on the quality of their service and change the provider at any time.18 The 2018 PIBS survey reports that over 90 percent of beneficiaries of all programs considered the treatment they received from the payment service providers as fair and polite. 43. Still, timeliness of payments continues to be a challenge, especially for the MLSP programs because of delays in the release of funds by the National Treasury. The HSNP was not affected by any significant payment delays early on because it was largely financed by DPs. However, the MLSP faced significant challenges related to delays in requesting funds from the National Treasury, preparing payrolls, receiving required authorizations, disbursements of funds to payment service providers, and reconciling the payments. By 2017, the Government was able to improve the timeliness of payments despite these challenges; it improved to 100 percent for the HSNP and 87 percent for programs under the MLSP (DLI 5). Timeliness of payments deteriorated, however, as the Government gradually assumed programs` financing. Government budget approval procedures kicked in requiring ex-ante compliance with several conditions. Starting in early 2020, budgetary constraints stemming from the economic impact of the COVID-19 pandemic made it difficult to disburse required amounts on time. Eventually, the percentage of payments made on time to PSPs for transfer to beneficiaries fell to 19.47 by end-2020, falling short of the targeted 90 percent (PDO 4). 17 The SAU became the Directorate of Social Assistance within the SDSP in December 2020. For consistency, it is referred to as the SAU throughout this report. 18 The accounts opened for NSNP beneficiaries by the payment service providers offer the following: (a) all transactions are authenticated biometrically by the beneficiary or registered caregiver; (b) transactions can be made at pay points across the country; (c) there are no ledger fees or other charges; (d) two free withdrawals are permitted per month, with additional withdrawals subject to fees as determined by the provider; and (e) two balance inquiries are permitted each month free of charge. Page 21 of 69 The World Bank National Safety Net Program for Results (P131305) 44. A fully functional G&CM mechanism is in place at the central level, but implementation of the decentralized case management was delayed due to COVID-19-related restrictions. Building upon experience with the HSNP and CT-OVC functional complaints management mechanisms, the Government documented procedures for making, receiving, recording, and responding to complaints and grievances in the CCTP OM (DLI 6). There were considerable challenges in documenting and reporting complaints under the MLSP’s programs, and the MLSP decided to focus on grievances and case management (G&CM), as opposed to strictly complaints and grievances.19 This broader approach required stronger reporting channels between the national and county and sub-county levels, along with the decentralization of MIS functions for G&CM to the county officers. A G&CM module was developed in the CCTP-MIS for local officers of the State Department for Social Protection (SDSP), based upon the procedures in the OM. In 2019, the G&CM mechanism was operationalized fully at the national level, and the Government planned to decentralize the module to all county offices in 2020. At this stage, local officers are being trained to use the module to register and handle complaints and beneficiary cases. But completion of training and operationalization at the county/sub-county level were affected by the COVID-19 pandemic and are facing budget constraints (Intermediate Result 5, DLI 6a (ii)). Given the importance of rolling out the decentralized G&CM throughout the country, this activity is being supported under KSEIP. The SAU maintains a toll-free hotline for anyone desiring to make a complaint about the NSNP at the national level. 45. A BOS was designed and delivered, but continued efforts are required to improve knowledge of the cash transfer programs, including awareness of program objectives, transfer amounts, and the G&CM system. The BOS was developed and adopted in 2018 (DLI 6b (i)) and was rolled out to 20 sub- counties in 2019 (DLI 6b (ii)). Nevertheless, despite rollout of the BOS, further outreach efforts are needed to increase beneficiaries’ awareness of the programs’ requirements and procedures. The 2018 PIBS survey reported that 86 percent of beneficiaries were aware of the correct transfer amount for the program in which they were enrolled, exceeding the target of 70 percent (Intermediate Result 3). The PIBS survey reported that only 30.4 percent of households could name two means of filing a complaint, against a target of 30 percent20 (Intermediate Result 4, DLI 6b (iii)). In 2019, a follow-up survey carried out by the Government found that both these numbers had deteriorated: only 75 percent of households knew the amount to which they were entitled, and only 25 percent of households knew of two ways of contacting the program, with only respondents from the HSNP exceeding the target of 30 percent. Although the expected targets were not met by completion, the roll-out of the decentralized G&CM and the BOS will continue under KSEIP to improve beneficiary awareness. 46. The MLSP has implemented fully a time-bound action plan with critical FM actions that systematically address the NSNP’s fiduciary risks (Intermediate Result 12, DLI 11). Measures to mitigate fiduciary risks were identified during preparation and included in the Program Action Plan (PAP) (Annex 1 C). However, during implementation additional efforts were identified to further strengthen fiduciary systems; and a DLI was introduced in the restructuring. The identified issues related to: (a) governance and management of information technology functions; (b) oversight of beneficiaries’/caregivers’ enrollment and payments; (c) transition to two PSPs; and (d) movement to the electronic payment system, including oversight of the payment process and automated reconciliation of payrolls. The Government 19Because of this new focus, the intermediate outcome “% of complaints actioned” was replaced by “Grievance and Case Management mechanisms for Cash Transfer Program are functional”. 20The 30 percent was the minimum level to trigger the DLR disbursement. The DLR was scalable and could disburse more above that percentage. The 40 percent was set as an end target through AF but at that time this was no longer a DLR but an Intermediate Result indicator. Page 22 of 69 The World Bank National Safety Net Program for Results (P131305) prepared and successfully implemented a comprehensive time-bound action plan to address the identified issues that covered, among others, (a) improved reconciliation process of the payroll, (b) strengthened payroll verification and approval workflows, (c) better segregation of duties, and (d) compliance of MISs with OMs. 47. Achievement of the expected outcomes under Results Area 2 is rated Substantial. Significant progress was made on several initiatives to improve governance of the NSNP, but full benefits are delayed mostly because of budget constraints as well as limitations in implementation due to COVID-19. There is no doubt, however, that the NSNP’s governance has shown major improvements by completion. A new HTM was adopted for all the NSNP programs, and a recertification plan was adopted in 2019 and piloted, but budgetary constraints have delayed its rollout. The targeting accuracy of the NSNP transfers was found to be good, with about 82 percent of beneficiaries falling below the harmonized PMT according to the PIBS baseline survey. The single registry of beneficiaries linked to the IPRS and supported by an integrated CCTP-MIS is operational, which together with the establishment of the SAU, has improved the programs` coordination and strengthened internal controls. Payments to beneficiary households are made electronically under a two-factor authorization system, by one of four possible PSPs. While beneficiaries report satisfaction with the treatment afforded by these providers, the timeliness of payments continues to be a challenge, mostly because of delays in the release of funds by the NT. A fully functional G&CM is in place at the central level, but implementation of decentralized mechanism has been delayed due to COVID-19. A BOS has been designed and delivered in some parts of the country. Further efforts are required to improve awareness of program objectives, transfer amounts and processes, and the G&CM, which are pursued under a follow-on KSEIP. Finally, implementation of critical FM actions has reduced the NSNP’s fiduciary risks. Results Area 3: Harmonizing cash transfer programs to increase the coherence of the safety net sector Box 3. Results Chain for Results Area 3 Intermediate Result 8: % of NSNP spending which is government financed (Baseline 2013=39; Target 2019=96; Achieved 2020=100) Intermediate Result 9: Amount spent on NSNP by GoK as % of GDP (deleted) Intermediate Result 10: Number of households per year paid through the NSNP out of NDEF (Baseline 2013=0; Target 2019=not specified; Achieved 2020=0) Intermediate Result 11: NSNP beneficiaries’ access to complementary social and productive services is promoted (Baseline 201=not available; Target 2019=Agreements signed with at least two partners providing relevant complementary services to NSNP beneficiary households; Achieved 2020=Complimentary services MIS module has been completed but agreements with service providers are yet to be signed) 48. In 2016, the Government adopted and implemented its consolidation strategy that allowed all cash transfer programs to use a streamlined targeting tool, payment system, and G&CM system (DLI 8). The strategy responded to the NSNP’s objective of harmonizing cash transfer programs to improve the sector’s coherence. It provided for the creation of the SAU to merge the management of the three programs in MLSP and thereby helped reduce fragmentation and promote efficiency. The consolidation helped establish a more robust system for targeting, enrollment, payments, and monitoring, expand coverage of programs in a coordinated fashion, minimize and eliminate duplication among programs and strengthen the programs’ governance. Page 23 of 69 The World Bank National Safety Net Program for Results (P131305) 49. A recent analysis of some of the cash transfer programs concludes that while the operating costs of various programs vary, they are reasonable on average. Administrative costs have fallen steadily from 7.41 percent of program budget in 2014/15 to 2.64 percent in 2019/20.21 While this invariably reflects efficiency improvements through program consolidation and standardization, it is also the result of severe budget constraints that have delayed recertification and other program improvements. 50. The Government has taken over financing of all the HSNP households, thus assuming full funding of the NSNP cash transfers in FY19/FY20 as other three programs have been financed by the Government (Intermediate Result 8, DLI 9). Budget constraints in FY20/21 have resulted in a budget shortfall of about Ksh 4.7 billion for the cash transfer programs implemented by the MLSP. Furthermore, the HSNP funding has been cut through amendments to the FY20/21 budget reflecting the dire revenue situation. Discussions are ongoing with the NT to bridge this gap through a supplementary budget as this would affect the financing of transfers to the existing beneficiaries toward the end of the fiscal year. 51. The Government has made important progress toward putting in place a cash transfers scale- up mechanism as part of the national drought risk management system. The HSNP has developed systems and procedures to scale up the payments to additional households during time of drought once the pre-agreed triggers are met. These include pre-registering households in drought-prone areas of four counties where the program operates, opening bank accounts for those households, activating the emergency triggers and disbursing cash transfers to the accounts of the affected population. A final requirement for a fully functioning scale-up system was the establishment of NDEF as a contingency fund that would provide sustainable financing of the HSNP through ringfencing some budget money and channeling funds from other sources, including DPs. Although the NDEF’s Regulations have been approved, additional time is required to allocate a seed capital for operationalization of this strategic instrument (Intermediate Result 10, DLI 7). The Government aims to do so through the FY21/22 budget. While this DLI 7 remains to be fully achieved, the HSNP scaled up 31 times during the NSNP’s implementation period, including pilot scale-up and flood risk payments in 2015 and 2016.22 Although according to the HSNP impact evaluation, beneficiaries found the payments intermittent and unpredictable, they recognized them useful in supporting basic needs. 52. A Complementary Service Module has been completed in the single registry (DLI 10a), and agreements are being discussed with two potential partners. The linking of beneficiaries with complementary services was expected to increase the coherence of the social protection system and contribute to increased resilience of beneficiary households in the long term. It was to do so by helping to diversify provision of broader social protection interventions, referring beneficiaries to services delivered by other partner organizations. The SDSP has been working with the Kakamega County Government and the National Hospital Insurance Fund to develop partnership agreements for complementary services but has not been able to formalize these agreements by the Program closing date (Intermediate Result 11, DLI 10b). 53. Achievement of the expected outcomes under Results Area 3 is rated Substantial. The consolidation strategy improved the coherence of the Government’s social protection system by harmonizing cash transfer programs and applying a streamlined targeting tool, payment system, and G&CM. The SAU coordinates and consolidates the management of the three programs under the MLSP, 21 Ministry of Labor and Social Protection, NSNP Completion and Assessment Report, March 2021. 22 Ministry of Labor and Social Protection, NSNP Completion and Assessment Report, March 2021. Page 24 of 69 The World Bank National Safety Net Program for Results (P131305) thereby reducing fragmentation and promoting efficiency in their administration. The Government has assumed financing of all the NSNP cash transfers in FY19/20, although the pandemic has placed a strain on the national budget. Finally, the legal requirements, including regulations, for establishing a system to underpin sustainable financing of the NSNP scale-up in response to climactic events have been approved, but the seed funding of the NDEF is affected by budgetary constraints. Rating of Overall Efficacy 54. Overall efficacy is rated Substantial. The PforR produced significant results in terms of putting in place a coordinated, effective, and sustainable safety net system in Kenya. Greater impact was not achieved, or documented, because of issues with the CT-OVC impact evaluation and the onset of the COVID-19 pandemic during the final stages of implementation, which affected completion of the last activities and the country’s fiscal situation, which impacted budget allocations to the NSNP. Nevertheless, the results achieved in relation to the uncoordinated systems in place at the time of appraisal are impressive. Kenya has developed a coordinated, targeted, and well-governed system of social safety net that, despite temporary budget shortfalls precipitated by the COVID-19 pandemic, is fully owned by the Government. Poverty estimates and beneficiaries’ responses in the PIBS survey confirm important impacts of the NSNP. The activities that were not completed by the closing date are being supported under KSEIP (Section IV D). C. JUSTIFICATION OF OVERALL OUTCOME RATING 55. The overall outcome is rated Satisfactory based on the Program’s continued High relevance, and Substantial efficacy, as described above. D. OTHER OUTCOMES AND IMPACTS (IF ANY) Gender 56. The Program benefited a large number of women and generated positive impacts on gender outcomes. According to the MLSP estimates, about 2.6 million of the Program’s 4.1 million beneficiaries are women (64 percent). Although women are not explicitly targeted as the main transfer recipients, the vulnerability of female-headed households is recognized in the program targeting of both the CT-OVC and HSNP. This has enabled women to assume more control of the household budgets and participate more in income-generating activities. The HSNP has indicated that there may also be small spillover effects into household decision-making. Evaluations have found that the cash transfer programs have resulted in young women being less likely to become pregnant (5 percentage points lower) and more likely to enjoy better mental health outcomes, greater belief in their self-agency and self-efficacy, and a more positive view of the future. Still, the NSNP’s Operational Monitoring Reports found that, in several counties, single mothers suffered from social exclusion because female-headed households are not treated as equally legitimate as the ones headed by a male, and this situation is compounded where women are not allowed to own property.23 23 National Safety Net Program (NSNP) Operational Monitoring Report, Fourth Monitoring Cycle, dated March 2018. Page 25 of 69 The World Bank National Safety Net Program for Results (P131305) Institutional Strengthening 57. Although technical assistance for the NSNP was financed through other vehicles (the WB- financed CT-OVC Project, DFID, UNICEF, and World Food Programme [WFP] funding), the PforR had an important impact on the institutional framework for delivery of social assistance in Kenya. The PforR’s three results areas supported policy actions that were directed at the institutional strengthening needed to deliver an effective NSNP for poor and vulnerable households. All the institutional achievements, including significant improvements in the Program’s FM, are detailed in section II B. Two of the NSNP’s achievements are worth highlighting: (a) consolidation of cash transfer programs, and (b) progress toward establishing the NDEF for sustainable financing of Kenya’s shock-responsive safety net. 58. The Government’s adoption of a consolidation strategy and action plan that resulted in the consolidation of three of the NSNP’s four cash transfers was a huge step: these three programs are now managed by the SAU through a synergized and coordinated approach (section II B). As described in greater detail in section II B, significant progress made by the Government in implementing its strategy and action plan has strengthened the institutional setup for the NSNP by (a) creating and staffing the SAU; (b) preparing the consolidated OM; (c) developing consolidated work plans, budget, performance targets, and communication strategy; (d) harmonizing the M&E framework and producing consolidated reports; (e) developing and implementing the CCTP-MIS, incorporating data from former programs’ data bases, and linking to the single registry; and (d) developing, piloting, and adopting the HTM for all four cash transfer programs supported by the NSNP. It has also put in place the G&CM mechanism at the national level and is decentralizing the mechanism and implementing the BOS. Poverty Reduction and Shared Prosperity 59. Kenya saw a steady but modest reduction in the poverty rate between 2005/06 and 2015/16, and recent calculations by the WB based on the 2015/16 KIHBS estimated a further reduction by 2019.24 The Kenya Poverty and Gender Assessment found that robust economic growth allowed the country to reduce the poverty headcount rate by more than 10 percentage points, from 46.8 percent in 2005/06 to 36.1 percent in 2015/16. Still, during the same period, the absolute number of poor people declined only marginally, from 16.6 million 16.4 million. The number of people living below the poverty line increased in urban and NEDI counties, from 2.3 million to 3.8 million and from 2.4 million to 3.2 million, respectively, while it decreased in rural and non-NEDI countries. Prior to COVID-19, the WB’s calculations projected further decline in poverty headcount rate to 32.3 percent in 2019. 60. The Program reached approximately 4.1 million individual beneficiaries in the poorest and most vulnerable households, but the overall coverage remains low, especially given the needs of Kenya’s population. The Kenya Public Expenditure Review (forthcoming) found that as estimated based on the 2015/16 KIHBS data, social assistance covers only 6.8 percent of the population. According to 2020 administrative data, NSNP transfers are estimated to cover up to 10 percent of all Kenya`s households. This coverage is higher than in several Western African countries but is below East Africa comparators such as Rwanda and Tanzania.25 The adequacy of transfers is quite low as the nominal value of cash transfers in three of the four NSNP programs has not increased since 2012, resulting in deterioration of their real value. Relatively small coverage of social assistance programs coupled with low adequacy of 24 World Bank. 2020. Kenya Economic Update 22: Special Focus on the Socioeconomic Impact of the COVID-19 Pandemic. 25 World Bank. 2018. Realizing the full potential of social safety nets in Africa. Page 26 of 69 The World Bank National Safety Net Program for Results (P131305) transfers translates in a low impact of these programs on reducing Kenya’s poverty rate: estimates from the 2015/16 KIHBS data show that overall, social assistance results in poverty reduction of some 0.3 percentage points. 61. On the positive side, surveys indicate that Kenya’s cash transfer programs have important impacts on beneficiaries’ welfare. The second Program Implementation and Beneficiary Satisfaction (PIBS II) survey reported the following positive impacts from the programs: (a) 94.7 percent of beneficiaries reported positive impacts on household consumption, (b) 93.8 percent reported positive impacts on household dietary diversity, (c) 92.7 percent reported positive impacts on household members’ school attendance and performance, (d) 90.2 percent reported positive impacts on household health, (e) 84.8 percent reported positive impacts on household members’ self-esteem, and (f) 59.1 percent reported positive impacts on household asset ownership. Ratings varied by program, with the HSNP beneficiary households surveyed reporting consistently higher impacts (including 74.54 percent reporting positive impacts on household income-generating activities). 26 62. The COVID-19 pandemic that affected the final stages of the Program’s implementation (Section III B) is estimated by the WB to have increased poverty, but it affected urban and rural households differently.27 The calculations project that poverty increased by just under 6 percentage points to a headcount rate of 38 percent, removing the last five years of poverty reduction, and that the severity of poverty also worsened with just under a 3 percentage point increase in the poverty gap. The impact on the poverty headcount and gap are larger in urban areas, resulting in a closing of the gap between urban and rural poverty. The initial impact of the pandemic is estimated to have increased poverty in urban areas by 18 percentage points compared to 9 percentage point in rural areas. The new poor affected by COVID- 19 tend to be younger, better educated, and live in households with fewer members and more often in urban areas. They have a greater dependency on employment in the services sector, than the existing poor who depend largely on the agricultural sector. 63. Given the characteristics of the new poor due to COVID-19, the Government opted for policy responses mostly outside the framework of the NSNP, despite availability of additional support under the PforR. The Government’s policy responses included reduced taxation, emergency cash transfers and extended public works. As the NSNP’s beneficiaries were not the group most affected by COVID-19, the Government opted to provide emergency cash transfers through a separate, temporary mechanism targeting the urban poor. Implementation started as a pilot paying a weekly stipend of Ksh 2,000 to over 108,000 households in the informal urban settlements of Nairobi, Mombasa, Kilifi, and Kwale. By April through June 2020, 342,000 households received the cash support. The scale-up continued at the beginning of FY20/21 with the goal of reaching the target coverage of 669,000 households. The WB estimates that all the policies applied by the Government may have reduced the overall poverty headcount by 1 percentage point and by just under 5 percentage points below pre-COVID levels in urban areas, given the policy focus on informal urban settlements. 26MLEAA, PIBS Survey for the Kenya NSNP, PIBS II Final Report dated October 2018. 27 World Bank, Technical Note, Informing Targeting of Households by COVID-19 Socio-Economic Impacts in Kenya. Estimations were based on the KIHBS 2015/16 and Kenya Macro Poverty Outlook. Using the 2019 baseline, the size of the impact was modelled considering (a) the direct loss of labor earning due to illness, (b) the indirect loss of labor earnings (either wages or job losses), and (c) a reduction in remittances. Real-time data collected from households, establishments, and micro-enterprises are used to adjust the size of the impact. Page 27 of 69 The World Bank National Safety Net Program for Results (P131305) III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME A. KEY FACTORS DURING PREPARATION 64. The Program was prepared in response to the Government’s strong commitment and drew the WB`s strong engagement with social protection in Kenya. The focus on the five cash transfer programs was justified as these programs shared several commonalities, which made a coordinated approach possible. First, the programs shared similar objectives and implementation requirements. Second, increased coordination among these programs would allow the Government to provide effective safety net support for the vulnerable groups while common implementation systems and procedures would make their delivery more efficient. Finally, focusing on these five programs would present an important step toward a more integrated social protection system, possibly extending the framework to other poverty-targeted programs over the longer term. The Government requested the WB’s support for the NSNP given its ongoing analytical work to help strengthen Kenya’s social protection policies. Also, the CT- OVC Project, financed by an SDR 33 million IDA Credit (and an SDR 6 million additional credit), was under implementation. This project was supporting an increase in social safety net access, through an effective and efficient expansion of the CT-OVC program, one of the five programs under the NSNP. 65. The choice of the PforR instrument was justified to ensure effective support for the Government- led social protection agenda, and sources for the financing of technical assistance that would be needed for institution and system strengthening had been identified. The PforR would support delivery of results for a Government program at the sectoral level; expand the use of Government systems by that program, thereby strengthening them; and harmonize funding for the program from different sources, including DFID, WFP, and UNICEF. The Government’s MTEFF supported the NSNP’s expansion based on donor partner commitments and a gradual assumption of exclusive government financing. As required, drawing on recommendations of the Technical and Integrated Fiduciary Assessments and Environmental and Social Systems Assessments (ESSAs), the preparation team developed a detailed PAP to ensure adequacy of the Government’s systems. The PAP activities aimed to: (a) improve the technical dimensions of the five cash transfer programs (targeting, payment, C&GM, and M&E); (b) increase the capacity of the SDSP within the MLSP and NDMA for NSNP`s oversight and implementation; and (c) mitigate risks, including by addressing fiduciary and social concerns. 66. The preparation team highlighted the needed investments in staffing, systems and infrastructure, and technical assistance to support implementation in the short and medium terms and identified sources for financing these. The Government would finance the operating costs to support effective program delivery, particularly activities related to targeting, payments, complaints and grievances, and M&E. Based on experience with the operating costs of the CT-OVC program, the Government allocated 15 percent of the total program costs to its operating budget in FY13/14. The WB planned to help finance technical assistance needs that were identified in the ESSAs under AF for the CT- OVC Project that was to be extended by three years. Also, additional US$5 million in grants from DFID were in the process of being signed to help fund technical assistance, and the WFP support was expected to continue. Page 28 of 69 The World Bank National Safety Net Program for Results (P131305) B. KEY FACTORS DURING IMPLEMENTATION 67. The SDR 166.9 million (US$250 million equivalent) IDA Credit for the National Safety Net Program for Results was approved by the WB’s Board of Directors on July 23, 2013; signed on September 9, 2013; and became effective on October 16, 2013. A midterm review (MTR) was held on February 12– 26, 2016 (as originally scheduled). Although implementation faced some challenges, the PforR was never considered at risk as reflected by its Implementation Progress or Development Objectives ratings in its Implementation Status and Results Reports (ISRs), which were consistently either Satisfactory or Moderately Satisfactory. 68. Despite some challenges and a ministerial restructuring, initial implementation expanded the NSNP cash transfer programs at a much larger scale than originally expected, and by the MTR, the Program had made solid progress. Following a general election in early 2013, the number of ministries was reduced, affecting the ministerial home of the NSNP programs.28 The CT-OVC, OPCT, PWSD-CT, and UFS-CT were housed in the new MLEAA, and the HSNP was housed under the Ministry of Devolution and Planning. The coordinating departments remained the same,29 and the staff moved to the new ministries without disruption in implementation arrangements. At the same time, the programs managed by the MLSP received a significant increase in budgetary allocations for FY13/14. The size and speed of the scale- up of those programs was seen as possibly stretching the ministry`s capacity. Therefore, the WB and DFID technical assistance was intensified to help address possible issues before they affected implementation. By the time of the MTR, several DLIs had been achieved and verified, and by December 2015, disbursements totaled US$100 million (44 percent of credit proceeds). The MTR identified some indicators that could eventually require restructuring but agreed that the need to restructure would be reevaluated in June 2016 (section IV A). 69. Despite this progress, several challenges continued to affect the Program. Timeliness of payments to beneficiaries remained a major challenge, with only the HSNP managing to make payments on time. Payment delays were particularly significant in the first part of the fiscal year when revenue collection tended to be low. Also, a payment tracking tool was developed and was being incorporated as part of the MIS to ensure proper tracking of payment delays in payment decision points to allow taking corrective measures on time. There were delays in carrying out the PIBS survey, which was critical to measure achievement of the DLIs relating to targeting accuracy and beneficiaries’ awareness with grievance redress mechanisms. When contracted, the firm to carry out the PIBS survey faced several challenges, including heavy rains throughout the country that made travelling difficult, insecurity in some parts of the country, and difficulties in tracking down some beneficiaries. Then, the initial draft of the firm’s baseline report was found to be less than satisfactory and had to be revised. The NSNP consolidation strategy, originally scheduled to be completed in July 2014, had not been finalized, partly due to resistance from program-level staff and the uncertainties that often surround organizational changes. There had been various audit queries related to internal controls and systemic weaknesses that posed high risks to the program (section IV B). 28 This ministerial restructuring hadbeen mentioned in PAD, although the specifics had not been decided at the time of appraisal. A full description of changes to the ministerial home of the NSNP programs is provided in footnote 2 of this ICR. 29 The CT-OVC continued under the Department of Children’s Services, th e OPCT, PWSD-CT, and UFS-CT continued under a renamed Department of Social Development, and the HSNP continued under the NDMA. Page 29 of 69 The World Bank National Safety Net Program for Results (P131305) 70. A severe drought in late 2016/early 2017 affected Kenya’s north and northeastern regions. In response, Kenya’s President declared a National Disaster on February 10, 2017. Of Kenya’s 47 counties, 23 were severely affected by the crisis. Most of these belonged to ASALs and had high levels of poverty representing about 30 percent of the country’s population. An estimated 2.7 million people became acutely food insecure according to an assessment by United Nations Office for Coordination of Humanitarian Affairs. Food prices reached a record high due to shortages. The President’s declaration included a direct appeal to Kenya’s DPs to assist and supplement national disaster response efforts, including a specific request to the WB to mobilize IDA resources from the Crisis Response Window. The SDR 36.8 million (US$50 million) AF was approved to address the additional needs. 71. The AF to increase coverage in NEDI counties, was approved on April 26, 2017, together with a project restructuring that adjusted the RF and the DLI matrix, and extended the closing date. The AF would allow the Government to continue increasing coverage in the north and northeastern counties, by reaching 50,000 new beneficiary households and taking over the financing of 20,000 and 40,000 existing households under the HSNP and CT-OVC respectively. At the time, four issues continued to pose challenges to implementation: (a) timely delivery of cash transfers; (b) establishment of the functional G&CM system at lower levels of implementation, together with a program of beneficiary outreach; (c) continued FM issues; and (d) delays in the approval of regulations for the NDEF due to the tense political environment in the country.30 Project restructuring helped address these challenges by adjusting some of the existing DLIs for which the measurement of results posed challenges (section IV A) and including a few additional activities, supported by new DLIs. 72. Starting in 2017, the PforR was affected by two factors: the Government’s decision to expand the OPCT’s beneficiary group and the COVID-19 pandemic. The decision to expand the OPCT to beneficiaries aged 70 and above was made by the Government in the tense political environment preceding Presidential elections. The pandemic affected the Government’s efforts to complete outstanding DLRs toward the end of the Program implementation. The measures announced and enforced to contain the spread of the virus included domestic travel restrictions and the suspension of all fieldwork, which made it impossible to make progress on expansion of the NSNP benefits coverage in north and northeastern counties (DLR 1e) and recertification of program beneficiaries in 20 sub-counties (DLR 2c). Further, the budgetary constraints precipitated by the pandemic negatively affected financing of the NSNP cash transfers and the achievement of DLIs and some key indicators including timeliness of payments (PDO 4) and operationalization of NDEF (DLI 7). The Government introduced an emergency cash transfer intervention for the households who lost income due to the COVID-19 containment measures, but that intervention was not financed under now traditional cash transfer programsas the authorities viewed the immediate income shock as having a primarily urban impact, and potentially as a security issue associated with a rising risk of social unrest in urban informal settlements while the NSNP supported programs targeted the extreme poor and vulnerable in mostly rural areas (section II B). IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME 30This was due to events following the elections in August 2017. The results of the Presidential elections in August 2017 were contested whereupon the Supreme Court called for a rerun. The new elections in October 2017 confirmed the incumbent President’s victory (the same outcome of the August 2017 elections). Page 30 of 69 The World Bank National Safety Net Program for Results (P131305) A. QUALITY OF MONITORING AND EVALUATION M&E Design 73. The design of M&E for the PforR was comprehensive and drew extensively on NSNP monitoring framework. It addressed/included the following: (a) progress toward the NSNP outcomes, outputs, and activities based on the NSNP’s M&E framework; (b) financial reporting; (c) independent impact evaluations of the CT-OVC and HSNP; (d) a PIBS survey; (e) detailed protocols to monitor achievement of DLIs; and (f) progress in PAP implementation. In addition, by design, the Program included activities that would strengthen monitoring through MISs of targeting, enrollment, payments, and complaints and grievances, as described in section II B. 74. The PforR’s DO statement and RF presented some challenges. First, the DO statement was very broad and defined neither what constituted an effective safety net program nor who comprised the poor and vulnerable populations. This ICR has interpreted an effective safety net program to be one that addresses the operation’s three Results Areas, as measured by the PDO indicators for each of them. Also, the first PDO indicator in the RF could not have been expected to impact any of the three Results Areas, but rather be impacted by progress towards successful implementation of measures in all of them together. 75. The SAU was responsible for monitoring progress toward achieving the NSNP’s outcomes, outputs, and activities, based on a set of 39 indicators from the M&E framework developed for the NSNP with donor partner support. Routine monitoring would include bimonthly, quarterly, half-yearly, and annual reports to inform program managers, government ministries, and DPs, and annual reports were to be made public on the MLSP’s website. Operational Monitoring Reports were delivered semiannually by an external consultant, and focused on different topics (targeting, vulnerable groups, communications, case management, and so on) based on a desk review of various documents, beneficiary data, and geographical coverage information, complemented by inputs from various key informants.31 Financial Reports were to be produced from the Integrated Financial Management System (IFMIS) on a semiannual and annual basis, based on the Government’s standard chart of accounts that includes the NSNP. Impact evaluations of the CT-OVC and the HSNP were being carried out at appraisal with financing under the AF of the CT-OVC Project. The PIBS survey would be conducted yearly by an independent contractor (also with financing under the AF of the CT-OVC Project) in a representative sample of locations for each of the programs and would interview beneficiary households, members of local committees, and frontline implementers.32 31 The Operational Monitoring Reports counted upon data collected through individual interviews with key informants. In each sub-county visited, in addition to discussion with carefully selected focus groups consisting of between 6 to 10 beneficiary respondents, interviews were conducted with (a) 10 non-beneficiaries; (b) three PSPs, including agents and banks; (c) three to four Country Program Officers, including one officer for each program; (d) three to four Sub-County Program Officers, including one officer for each program; (e) three Constituency Social Assistance Committee members; (f) 10 Beneficiary Welfare Committee/Local Orphans and Vulnerable Children Committee members; and (g) three local leaders. 32 The PIBS aimed to (a) measure the quality of program implementation, (b) assess the characteristic of beneficiary households against the program’s eligibility criteria, (c) gauge beneficiaries` awareness about their entitlements, program systems, such as the complaint and grievance mechanism, and perceptions of the quality of program implementation; (d) seek beneficiaries` view of the program impact; and (e) verify that the programs are being implemented according to the OMs. Page 31 of 69 The World Bank National Safety Net Program for Results (P131305) M&E Implementation 76. Routine monitoring was implemented by the SAU, based on information provided by programs’ secretariats. Each of these secretariats had staff responsible for preparing reports based on information on program implementation provided through county and sub-county officials and staff responsible for ensuring that up-to-date information was in the programs’ MISs. During implementation, an integrated beneficiary registry for all four cash transfer programs was developed and put under implementation, thereby facilitating comparison and eliminating duplication of benefits among programs. 77. The main M&E challenges related to implementation of impact evaluations of the HSNP and CT- OVC and the PIBS, all of which were financed under the AF for the CT-OVC Project but were required to document the PforR results. Implementation of these tasks suffered initial delays in procuring external technical assistance. Further delays were encountered during the fieldwork and data analysis for the PIBS. A first draft of the PIBS baseline report was found to be less than satisfactory. Therefore, WB and DFID engaged technical support to help the Government improve the quality of the revised report and ensure that follow up surveys produce timely and robust results. The baseline report of the PIBS (PIBS I) carried out in 2015/16 was delivered in June 2016, two years later than planned. The second PIBS survey (PIBS II) was completed in 2018. The baseline for the CT-OVC impact evaluation also faced delays and was only completed in 2015. The midline for the CT-OVC impact evaluation never took place due to contracting issues while the end line was completed only in 2019. Both studies produced valuable analysis for substantiating the NSNP’s outcomes, although there were some limitations in the quality of data collected for the end line of the CT-OVC impact evaluation (Section II B). Furthermore, the HSNP evaluation study commissioned with DFID support used a mixed-methods approach to provide a nuanced assessment of the program`s benefits and impact.33 M&E Utilization 78. The PforR’s RF and monitoring reports were used routinely during Joint Review and Implementation Support (JRIS) missions34 to: (a) discuss achievement of results, including DLIs, and identify key action to ensure that the NSNP would attain them on time, (b) identify measures to advance Program implementation, and (c) identify areas of technical and operational support to advance key activities and how development partners could provide that. The availability of timely and coordinated monitoring data was made possible, in part, through the integrated MIS for the NSNP. The JRIS reviews, based on the monitoring information, helped identify the need for and structure of the Program’s AF, including adjustments to key indicators and DLIs, as mentioned earlier. The evidence-based reviews also served to guide an informed discussion of next steps (for example, the ‘cash plus’ agenda) to be included as an integral component of the successor KSEIP (section IV D). Results of these reviews were routinely summarized in Aide Memoires and updated in each of the 13 ISRs filed. 33 Oxford Policy Management. 2018. Evaluation of the Kenya Hunger Safety Net Programme Phase 2: Impact Evaluation Final Report Executive Summary. https://www.opml.co.uk/files/Publications/a0013-evaluation-kenya-hunger-safety-net- programme/impact-evaluation-final-report-ex-summary.pdf?noredirect=1. 34 These review missions counted upon the participation of representatives of the cash transfer programs that constitute the NSNP; representatives of Government ministries, including the National Treasury; and representatives of DPs, including representatives from DFID, the Embassy of Sweden, the WFP, UNICEF, and the World Bank. Page 32 of 69 The World Bank National Safety Net Program for Results (P131305) Justification of Overall Rating of M&E Quality 79. The overall rating of M&E quality is Substantial. The Program’s M&E, including its RF, DLI Matrix, and M&E plans and procedures was (a) well designed, with coherence and complementarity among results areas, key indicators, and DLIs; (b) thoroughly implemented, including timely adjustments to reflect experience, albeit with issues with respect to contracting delays and quality of analysis for the evaluations; and (c) used reliably to inform issues and areas of concern as well as credit disbursements and next steps in the Government’s social protection agenda. B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE Environmental and Social 80. The ESSA required for OP/BP 9.00 focused on potential issues related to equitable access to the five cash transfer programs that constituted the NSNP and the ability of these programs to meet the needs of vulnerable and marginalized groups. This is because of the absence of any direct environmental impacts and the limited range of potential social effects. The ESSA concluded that Kenya’s policy environment provided an appropriate framework for management of the potential social effects of the NSNP and that the institutional arrangements for the five cash transfer programs were adequate, although capacity-constrained at all levels. The fieldwork carried out to inform the ESSA assessed the extent to which marginalized communities were included or excluded from the four programs managed by the MLSP. The findings of the ESSA confirmed that neither the NSNP nor the PforR operation included any activities that were judged to be likely to have a significant adverse impact that would be sensitive, diverse, or unprecedented on the environment and/or affected people. Although ESSA found instances of potentially eligible households from marginalized communities not being included in a program, it was not possible to determine whether this was a consequence of their marginalized situation or the low program coverage in those areas. However, there were some indications of potential bias given the limited number of households benefiting from the cash transfers. 81. A review of social safeguards was carried out from April-July 2016 to assess the NSNP’s compliance with WB social safeguard policies and the status of the ESSA. The review found no evidence of any consistent pattern of exclusion of marginalized groups, although low program`s coverage in remote areas continued to be a challenge. The review found that if there were greater efforts to implement the recommendations of the ESSA, especially with regard to increasing the understanding and documentation of the marginalized groups and where they were located, a social risk rating of highly satisfactory could be achieved. To promote this, the review recommended (a) strengthening the implementation of the sections of the OM on marginalized communities, (b) improving understanding and documentation of where minority groups are located and the challenges and successes in reaching them, (c) including in the communications strategy measures to reach minority and hard-to-reach groups; and (d) including indicators to measure potential exclusion in internal and operational monitoring. 82. The process to strengthen the cash transfer programs’ operational systems and complaints and grievance mechanisms was expected to ensure the voice of the poor and vulnerable in the programs. The OMs for the five programs were revised to include guidance on how to ensure that marginalized communities would be considered equally during the targeting process and participate during program implementation, and staff were trained in the use of the new guidelines. The application of these guidelines was assessed during regular program monitoring. As described in section II B, the G&CM Page 33 of 69 The World Bank National Safety Net Program for Results (P131305) mechanisms at the national level were strengthened based on the experience of the CT-OVC program and a module for the consolidated MIS was designed to track the number and type of complaints that were lodged and the actions taken. Training was delayed because of COVID-19 but has resumed online. Once the training is completed, the G&CM will be run in a decentralized manner through country officers. Fiduciary 83. The Integrated Fiduciary Assessment Review (IFAR), carried out during preparation, concluded that the NSNP’s FM and procurement systems were adequate but identified several weaknesses in the fiduciary systems of the five cash transfer programs, thus the PforR’s fiduciary risk was considered Substantial. The weaknesses related to the need for cash transfer programs to ensure that the right people are being paid the right amount at the right time were addressed during preparation in the design of the Program, specifically its governance Result area. The IFAR also identified risks relating to weaknesses in the public financial management (PFM) system upon which the NSNP would rely.35 The PAP included several activities to address PFM issues as they relate to the NSNP, specifically the extended coverage of internal and external audits, the adoption of specific codes for the NSNP based on the standard chart of accounts to enable program-based budgeting and reporting, and improving the functionality of the IFMIS. The IFAR also identified a risk related to possible delays in procurement and contract award of a new PSP for the NSNP due to the MLSP’s insufficient technical and procurement capacity to carry out procurement without external support. An expert on payment services was hired to help manage the procurement process. 84. The NSNP’s FM reviews and internal/external audits periodically identified routine internal control and systemic weakness that posed high risks for the program, but strong leadership and commitment at the highest levels of the MLSP, together with technical support from DPs and the WB, made it possible to address the FM challenges. Immediately after the MTR, the Government and DPs initiated a broader stream of work on preventing, detecting, and deterring error, fraud, and corruption (EFC) in NSNP in a more systematic fashion. The Government team engaged an international expertise to advance the work on EFC. By the end of July 2016, the MLSP had prepared a draft EFC strategy and action plan. The MTR also identified specific EFC capacity requirements to be incorporated in the NSNP technical assistance plan (financed under the CT-OVC Project AF), including study and exchange visits, and operationalization of the EFC implementation strategy. Then, a forensic audit of the CT-OVC program conducted by the Kenya National Audit Office (KENAO) raised several fiduciary issues, including those related to the PSPs and refunds that had not been processed and reconciliations that were still pending, and also several operational issues related to the MIS. The MLSP worked closely with the WB and DFID teams to address each of the observations raised in the audit and provided KENAO with a management response to the audit’s observations. In addition, a new DLI 11 was included through the AF and restructuring to address the Program’s critical FM issues in a Comprehensive Action Plan. The ministry, with technical assistance from the WB, had complied fully with the FM Action Plan. 35 The PFM issues included (a) inadequate segregation of total program-level expenditures and the lack of ‘ring-fenced’ resources for operational costs for three of the cash transfer programs implemented by the MLSP that were budgeted under a single budget line; (b) the IFMIS not being reliably used for program-based budgeting, accounting, and financial reporting; (c) inadequate segregation in the management of program payrolls; (d) inadequate internal audit coverage of the programs and limited use of risk-based audits; (e) inadequate FM and operations staff in the departments managing the program; (f) weak links between technical departments and finance/accounting departments within the MLSP; and (g) funds and cash flow management inefficiencies at the ministry and program levels. Page 34 of 69 The World Bank National Safety Net Program for Results (P131305) C. BANK PERFORMANCE Quality at Entry 85. The WB’s Performance in ensuring quality at entry is rated Satisfactory. Building upon financial and technical assistance under the then ongoing CT-OVC Project, the team established a strong dialogue with the Government aimed at strengthening Kenya’s social protection system. That dialogue contributed to the formulation of the Government’s NSNP which became the backbone of the Program. The operation was processed under the new PforR lending instrument, the first of its kind in Kenya and in the Social Protection Practice. The PforR instrument had the advantage of allowing delivery of results for a government program at the sectoral level; expanding the use of government systems thereby strengthening them; and harmonizing funding for the program from different sources, including DPs. In this sense, the team built upon a successful engagement by breaking new ground and taking the WB’s support to a broader level. The team ensured coordination of needed funding for technical assistance and support under both the CT-OVC Project and DPs (the PforR was not structured to provide this funding). This was due to the fact that the same WB team was responsible for both implementation support under the CT-OVC Project and preparation of the PforR. The Results Framework and DLI matrix were concise and the key indicators and DLI protocols were well formulated, with measurable targets within the established M&E design, all of which facilitated subsequent monitoring of the Program and the credit’s disbursements. Quality of Supervision 86. The WB’s performance in ensuring quality of supervision is rated Satisfactory. The WB’s implementation support team—initially led by the same team that prepared the PforR—inherited a well- designed program ready for implementation. Implementation support was timely, results focused, and well documented, both in Aide Memoires and ISRs. The Program’s AF and restructuring—again a first for a PforR Instrument—was prepared under a strict timeline in view of the drought’s impact and availability of IDA CRW funding. The AF and restructuring were well prepared, responsive to the Government’s needs, and adjusted the Program’s implementation and RF to the increased scope and the progress in and challenges of implementation. In the final stages of implementation, the team worked closely with the Government to identify how the PforR could support its response to the COVID-19 crisis but concluded that the restructuring was impossible. The team focused on helping to accelerate outstanding activities, to extend the closing date to permit completion of as many activities and results as possible, and eventually to cancel the undisbursed credit balance to allow the Government to reallocate the IDA funds for other development needs. Justification of Overall Rating of Bank Performance 87. The overall rating of WB performance is Satisfactory. Building upon a traditional investment project supporting one of the Government’s cash transfer programs, the WB developed a strong sectoral dialogue through analytical work and preparation of the PforR, which contributed to the adoption of the Government’s NSNP. The WB’s team was bold, opting for a PforR operation, the first in Kenya and in the Global Practice which, given the availability of financing for technical assistance under other vehicles, turned out to be a good decision with a strong results focus. Implementation support was hands-on and responsive to the client’s and the Program’s needs. Page 35 of 69 The World Bank National Safety Net Program for Results (P131305) D. RISK TO DEVELOPMENT OUTCOME 88. The risk to development outcome is Moderate. The Government has shown continuous commitment to the Program and has made solid progress toward achieving the PDOs, as measured by progress toward key indicator targets, disaggregated by each of the three results areas. The Government has assumed full financing of the NSNP cash transfers, although currently being affected by budgetary constraints. The constraints concern the Government’s ability to make payments to programs’ beneficiaries but also delay achievement of some of the key indicators and DLIs. Funding gaps, as well as implementation delays, were precipitated to a large extent by the COVID-19 pandemic that led to slowdown in economic activity and exacerbated the Government’s fiscal pressures. The factors that have affected the program are exogenous, and do not reflect the Government’s commitment (or lack thereof) to its NSNP. The Government is committed to further strengthen the program`s delivery systems, expand the coverage of cash transfers, and link them with other social protection interventions and services that would help enhance human capital of the beneficiary households. To that end, it counts upon KSEIP financed by the €215.9 million (US$250 million equivalent) IDA credit and structured as an Investment Project Financing (IPF) with performance-based conditions (PBCs).36 The KSEIP (P164654) was approved on January 31, 2019 and is currently under implementation providing continuity to initiatives begun under the PforR namely G&CM decentralization, recertification/update of beneficiary eligibility through collection of data for a national social registry, linking program beneficiaries to complementary services, and expansion of the HSNP with strengthening its the shock response mechanism. The Government continues to draw on financial and technical support from UK’s Foreign, Commonwealth and Development Office (FCDO)37, UNICEF, and WFP. V. LESSONS AND RECOMMENDATIONS 89. Where Government programs require institutional strengthening, well-designed performance- based operations, whether applying a PforR or an IPF instrument, require a confirmed source of financing for technical assistance. The PforR counted upon technical assistance provided under the WB- financed CT-OVC Project and by other DPs. The synergies among the various sources of financing, and the close coordination of the teams, including through joint review missions, ensured the timely availability of technical support for advancing the NSNP’s activities. 90. The activities to build broad-based political support and coalition around the value and priorities of the national safety nets are as important as investing in government`s technical expertise and delivery capacity. In the lead up to the 2017 elections, the Government announced the expansion of the OPCT to beneficiaries aged 70 years and above and implemented it afterwards. That allowed the national safety net to expand significantly and become more visible at the national scale. However, given the limited budget envelope, the expansion of poverty targeted NSNP benefits in the NEDI counties had to be postponed. Social assistance policies may shift along political cycles with universal cash transfers being prioritized for ease of implementation and communication to the public. Maintaining political consensus 36 The KSEIP is structured as an IPF with PBCs as sources of financing for technical assistance were not available in the amounts required (as was the case when the PforR was prepared). A US$58 million equivalent of the US$250 million equivalent credit is allocated to technical assistance under regular IPF (disbursement against payments) needed to support achievement of PBCs. The KSEIP’s PDO is to strengthen delivery systems for enhanced access to social and economic inclusion and shock-responsive safety nets for poor and vulnerable households. 37 Formerly DFID. Page 36 of 69 The World Bank National Safety Net Program for Results (P131305) around poverty-targeted programs requires taking policy dialogue beyond the usual technical counterparts to build broader government support. Moreover, it is important that the DPs engaged in social protection join forces for such advocacy efforts. On the other hand, implementation support teams should be flexible to respond to changed policy priorities as long as the government can demonstrate their technical soundness and implementation readiness. 91. Major economic shocks offer the opportunity to raise the importance of investments in adaptation and preparedness of the safety nets for the subsequent crises. As discussed, the Government opted to cushion households from the impact of COVID-19 pandemic outside the NSNP framework and past the PforR. Absent of country-wide shock response mechanisms, NSNP was not seen as a vehicle to deliver assistance to the emerging group of new poor in informal urban settlements. Thus, the Government rolled out emergency cash transfers through a separate, temporary program. In a way, the COVID-19 crisis exposed a weakness of the existing social assistance system, mainly focused on alleviating chronic poverty particularly in rural areas, rather than responding to shocks in urban areas. However, the COVID-19 pandemic also provided an opportunity to raise the profile of the national safety nets, design strategies to expand their coverage and intensified discussions around adaptive social protection systems. The ongoing KSEIP is well placed to support this agenda by building a national social registry to host the data about the country`s poor and vulnerable households. The registry will serve as a platform to target and deliver assistance to the affected population in times of crisis. 92. When issues are identified through risk assessments during preparation, teams need to consider carefully whether they should be included in PAP or supported through DLIs. The issues concerned with fiduciary strengthening are normally included in the PforR’s PAP, but at times they require financial incentive to ensure compliance. Under the PforR, several FM concerns were identified and included in the operation’s PAP. However, when lack of progress in addressing those issues started to increase the risks to attaining the Program`s results, the team opted to add a DLI with a comprehensive plan of FM actions that were critical for NSNP governance and implementation. With the incentives linking progress to disbursements, these actions were implemented satisfactorily. . Page 37 of 69 The World Bank National Safety Net Program for Results (P131305) ANNEX 1. RESULTS FRAMEWORK, DISBURSEMENT-LINKED INDICATORS, AND PROGRAM ACTION PLAN Annex 1A. RESULTS FRAMEWORK (i) PDO Indicators Objective/Outcome: To support the Republic of Kenya's efforts to establish effective NSNP for poor and vulnerable HHs Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Net change in beneficiary Number 0.00 150.00 66.00 household monthly per adult equivalent consumption 26-Jun-2013 31-Dec-2020 31-Dec-2020 expenditure Comments (achievements against targets): This indicator was to be measured by the CT-OVC Impact Evaluation and the HSNP Impact Evaluation. Unfortunately, the results from the CT-OVC Impact Evaluation show that due to a number of reasons (the quality of the consumption data collected for the end line with food consumption likely to be significantly underreported as well as the high attrition rate), it is difficult to detect the true impact of the CT-OVC program on the household food consumption. According to the HSNP Impact Evaluation, the Propensity-score matching (PSM) results confirm an impact on monthly food expenditure, showing that households that have received a regular HSNP payment experience an increase in monthly per adult equivalent food expenditure of around Ksh66. However, the spill-over effects observed in the HSNP mean that it is likely that the impact on consumption for HSNP beneficiaries is underestimated. For every shilling injected into the local economy by the HSNP, total incomes are raised by somewhere between 1.38 and 1.93 shillings. Most of this additional income goes to non-beneficiaries of the program, hence producing spill-overs. Page 38 of 69 The World Bank National Safety Net Program for Results (P131305) Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of NSNP Number 1,650,000.00 2500000.00 4,280,000.00 4,082,704.00 beneficiaries, by sex 28-Jun-2013 29-Dec-2017 31-Dec-2019 31-Dec-2020 Comments (achievements against targets): This is estimated based on the total number of households on the payroll and benefiting from the four programs. Households in CCTP were multiplied by an average size of a household of 5 people. OP-CT beneficiaries above 70 year old are treated as individual cash transfer recipients. HSNP beneficiaries are estimated based of an average household size of 7. The data available in MISs is insufficient to identify the gender composition of households. The only accurate data available is for the primary recipient/account holder. This data has been used to estimate a gender breakdown but it may results in an over- estimation of the number of females because they are primary/account holders for CT-OVC. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion System for recertification of Text No Completion of the Recertification Plan NSNP beneficiaries is in place rollout of the has been adopted but recertification plan in not rolled out. at least 20 sub- counties 05-Apr-2017 31-Dec-2019 31-Dec-2020 Comments (achievements against targets): The Recertification Plan has been adopted and implementation tools developed. However, the roll-out of recertification has not started due to lack of funding. Page 39 of 69 The World Bank National Safety Net Program for Results (P131305) Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion % of all payments are made Percentage 12.00 70.00 90.00 19.47 on time to Payment Service Providers for transfer to 05-Apr-2017 29-Dec-2017 31-Dec-2020 31-Dec-2020 Program Beneficiaries Comments (achievements against targets): The timeliness of payments has fluctuated significantly during the project implementation. At the end of 2019 the indicator approximated 70 percent. However, due to the tight fiscal situation, in 2020, most of disbursements were delayed hence the drop in the percentage of payments made on time. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Beneficiaries of social safety Number 1,650,000.00 2500000.00 4,280,000.00 4,082,704.00 net programs 26-Jun-2013 29-Dec-2017 31-Dec-2019 31-Dec-2020 Beneficiaries of social safety Number 825000.00 2140000.00 2,675,735.00 net programs - Female Comments (achievements against targets): This is estimated based on the total number of households on the payroll and benefiting from the four programs. Households in CCTP were multiplied by an average size of a household of 5 people. OP-CT beneficiaries above 70 year old are treated as individual cash transfer recipients. HSNP beneficiaries are estimated based of an average household size of 7. The data available in MISs is insufficient to identify the gender composition of households. The only Page 40 of 69 The World Bank National Safety Net Program for Results (P131305) accurate data available is for the primary recipient/account holder. This data has been used to estimate a gender breakdown but it may results in an over- estimation of the number of females because they are primary/account holders for CT-OVC. (ii) Intermediate Results Indicators Results Area: Expanding cash transfer programs to promote more comprehensive and equitable coverage Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of households paid Number 114,384.00 792000.00 1,194,896.00 for by the Government 05-Apr-2017 31-Dec-2019 31-Dec-2020 Comments (achievements against targets): Data on number of households derived from July/August 2020 payrolls. Results Area: Strengthening program systems to ensure good governance Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion % of beneficiaries who know Percentage 56.30 70.00 75.00 correct program transfer amount for the program in 05-Apr-2017 31-Dec-2019 31-Dec-2020 Page 41 of 69 The World Bank National Safety Net Program for Results (P131305) which they are enrolled Comments (achievements against targets): This indicator has been measured and found as overachieved at 86% as part of PIBS II. The overachievement of the target was also confirmed by the 2019 survey carried out by OPM that showed about 75% of households know the amount to which they are entitled. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion % of beneficiaries for whom Percentage 40.00 90.00 90.00 100.00 payments are made electronically using Two- 28-Jun-2013 29-Dec-2017 31-Dec-2019 31-Dec-2020 Factor Authentication Comments (achievements against targets): All payments are made to bank accounts and rely on two factor authentication. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Grievance and Case Text Not in place BOS implemented in BOS implementation Management mechanisms at least 20 counties. completed in 30 sub- for the Cash Transfer counties. Programs are functional 05-Apr-2017 31-Dec-2019 31-Dec-2020 Comments (achievements against targets): Page 42 of 69 The World Bank National Safety Net Program for Results (P131305) The Beneficiary Outreach Strategy (BOS) was developed and implemented in 30 sub-counties. Further implementation is being constrained by lack of funding. The central level G&CM mechanism is fully functional while its deployment at the local level is still underway. The corresponding MIS module for local SDSP officers has been developed. The local officers are being trained to apply the module to register and handle complaints and update the beneficiary cases. It is expected that after completion of training, the G&CM will be run in a decentralized manner i.e. through country and sub-county officers. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion % of Program Beneficiaries Percentage 15.00 65.00 40.00 25.00 who can name two means of making complaint 26-Jun-2013 29-Dec-2017 31-Dec-2019 31-Dec-2020 Comments (achievements against targets): PIBS II reported awareness at 30.4%. Another survey completed in 2019 suggested that about 25% of households know two ways of contacting the programs. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Single Registry is operational Yes/No No Y Yes and progam MISs use the Standards for Internal Payroll 26-Jun-2013 29-Dec-2017 31-Dec-2020 Controls Comments (achievements against targets): Page 43 of 69 The World Bank National Safety Net Program for Results (P131305) Results Area: Haromonizing cash transfer programs to increase the coherence of the safety net sector Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion % of NSNP funding financed Percentage 39.00 79.00 96.00 100.00 by the Government 26-Jun-2013 29-Dec-2017 31-Dec-2019 31-Dec-2020 Comments (achievements against targets): The Government has taken over financing of all HSNP households as part of the July/August 2019 beneficiary payment cycle achieving thus full funding of NSNP cash transfers. Other three programs were financed by the Government prior to FY19/20. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of households per Text 0.00 not available 0.00 year paid through the NSNP out of NDEF 26-Jun-2013 31-Dec-2020 31-Dec-2020 Comments (achievements against targets): The establishment of NDEF was completed in May 2021. However, the Government is yet to allocate budget to enable the entity to pay households in response to the drought-induced shock. Indicator Name Unit of Measure Baseline Original Target Formally Revised Actual Achieved at Page 44 of 69 The World Bank National Safety Net Program for Results (P131305) Target Completion NSNP beneficiaries' access to Text n/a Agreements signed Complimentary complementary social and with at least two services MIS module productive services is partners providing has been completed promoted relevant but agreements with complementary service providers are services to NSNP yet to be signed. beneficiary households 05-Apr-2017 31-Dec-2019 31-Dec-2020 Comments (achievements against targets): Complementary services MIS module has been completed. However, the partnership agreements with the providers of complementary services have not been signed yet. SDSP is in process of developing agreements with Kakamega county government and the National Hospital Insurance Fund. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Fiduciary systems for NSNP Percentage 0.00 100.00 100.00 has been strengthened 05-Apr-2017 31-Dec-2019 31-Dec-2020 Comments (achievements against targets): The MLSP has implemented 100% of the time-bound FM action plan that allows to strengthen fiduciary systems for the NSNP and address the risks and systemic weaknesses identified by previous audits. Page 45 of 69 The World Bank National Safety Net Program for Results (P131305) ANNEX 1B. DISBURSEMENT LINKED INDICATORS DLI IN00904044 TABLE DLI 1: Additional households are enrolled in the NSNP (Number) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values 0.00 0.00 0.00 0.00 0.00 0.00 Actual values 445,000.00 275,000.00 305,000.00 345,000.00 221,000.00 Allocated amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 103,000,000.0 5,000,000.00 11,570,000.00 15,430,000.00 221,000.00 Disbursed amount ($) 135,221,000.00 0 Comments (achievements against targets): The Government exceeded the target for this DLI. However, for the purpose of the DLI disbursements the target was 135,000 (under original IDA credit) plus 70,000 (AF IDA credit) = 205,000 additional households. The Government has achieved the expansion of 135,000 planned for the original PforR (worth around US$ 100 million) as well as taken over financing of 51,000 households (40,000 CT-OVC and 11,000 HSNP). Furthermore, as of FY19/20 the Government has taken over financing of all HSNP beneficiaries or over 100K households. This means the target of 74,000 households (54,000 baseline + 20,000 target) set under DLR 1e has been exceeded by far. The remainder of about 30,000 households refers mainly to extra CT-OVC beneficiaries to be enrolled in NEDI counties. Registration process for NEDI expansion started in February 2020 but was not completed because of Covid pandemic restrictions. Number of existing households in the program is much higher, but we have here amended the reporting and confirm that for the purpose of the DLI the "actual current" value is 221,000. DLI IN00904045 TABLE DLI 2: System for recertification of NSNP beneficiaries is in place (Yes/No) Page 46 of 69 The World Bank National Safety Net Program for Results (P131305) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values No Actual values No No Yes Yes Allocated amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 Disbursed amount ($) 0.00 0.00 5,000,000.00 10,000,000.00 0.00 15,000,000.00 Comments (achievements against targets): The re-certification plan was adopted in June 2019. Roll-out of the plan has not taken place absent of funding. DLI IN00904046 TABLE DLI 3: Single Registry is fully operational with program MISs using agreed standards for internal payroll controls (Yes/No) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values No Actual values Yes Yes Yes Yes Allocated amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 Disbursed amount ($) 25,000,000.00 0.00 0.00 0.00 0.00 25,000,000.00 DLI IN00904047 TABLE DLI 4: Percent of NSNP payments made electroincally using two-factor authentication (Percentage) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Page 47 of 69 The World Bank National Safety Net Program for Results (P131305) Original values 40.00 0.00 0.00 0.00 0.00 0.00 Actual values 90.00 90.00 90.00 90.00 0.00 Allocated amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 Disbursed amount ($) 15,000,000.00 0.00 0.00 0.00 100.00 15,000,100.00 DLI IN00904048 TABLE DLI 5: Percent of all payments are made on time to Payment Service Providers for transfer of Program Beneficiaries (Percentage) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values 12.00 0.00 0.00 0.00 0.00 0.00 Actual values 0.00 0.00 75.00 0.00 0.00 Allocated amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 Disbursed amount ($) 0.00 0.00 15,000,000.00 0.00 0.00 15,000,000.00 DLI IN00904049 TABLE DLI 6: Grievance and Case Management mechanisms for the Cash Transfer Programs are functional (Text) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values Not functional Beneficiary Beneficiary Beneficiary G&CM Outreach Outreach awareness of Actual values functional at Strategy Strategy transfer all levels developed implemented amount and and adopted in at least 20 G&CM system Page 48 of 69 The World Bank National Safety Net Program for Results (P131305) sub-counties Allocated amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 Disbursed amount ($) 15,000,000.00 3,000,000.00 10,000,000.00 7,000,000.00 0.00 35,000,000.00 Comments (achievements against targets): The G&CM system is functional nationally, and efforts to develop a decentralized G&CM MIS module continue. To complete the decentralization process, the Social Assistance Unit proceeded to train local SDSP officers on utilization of G&CM module of CCTP MIS. The training is underway. The Beneficiary Outreach Strategy has been developed and implemented in 30 sub- counties in 4 counties. Despite the outreach activities, the target for DLR 6 b(iii) on beneficiary`s awareness of transfer am ount and G&CM system have not been achieved in terms of having at least 30 percent of beneficiary households to name two means of contacting the program for the purposes of submitting a grievance or updating their information. A beneficiary s urvey completed by September 2019 indicated that the share of beneficiaries aware of two means of contacting the NSNP has been below 30 percent. DLI IN00904050 TABLE DLI 7: System for scaling up the NSNP has been adopted (Yes/No) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values No Actual values No No No No Allocated amount ($) 20,000,000.00 0.00 0.00 0.00 0.00 20,000,000.00 Disbursed amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 Comments (achievements against targets): NDEF has been established. However, the National Treasury is yet to allocate budget for its operation. Page 49 of 69 The World Bank National Safety Net Program for Results (P131305) DLI IN00904051 TABLE DLI 8: Government has adopted the consolidation strategy (Yes/No) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values No Actual values Yes Yes Yes Yes Allocated amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 Disbursed amount ($) 15,000,000.00 0.00 0.00 0.00 0.00 15,000,000.00 DLI IN00904052 TABLE DLI 9: HSNP is financed (Yes/No) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values No Actual values Yes Yes Yes Yes Allocated amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 Disbursed amount ($) 5,000,000.00 0.00 0.00 0.00 0.00 5,000,000.00 DLI IN00904053 TABLE DLI 10: NSNP beneficiaries' access to complementary social and productive services is promoted (Text) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values Not available Page 50 of 69 The World Bank National Safety Net Program for Results (P131305) Complementa Agreements ry Services signed with at module in SR least two is ready and partners formal data providing Actual values N/A sharing relevant protocols are complementar available for y services to the use of NSNP potential beneficiary partners households Allocated amount ($) 0.00 2,000,000.00 1,000,000.00 0.00 0.00 3,000,000.00 Disbursed amount ($) 0.00 2,000,000.00 0.00 0.00 0.00 2,000,000.00 Comments (achievements against targets): Complementary Services module in the MIS has been developed. However, the partnerships with at least two providers of relevant complementary services to NSNP beneficiary households has not been established yet. DLI IN00904054 TABLE DLI 11: Fiduciary systems for NSNP have been strengthened (Percentage) December December December December Baseline July 2017 Total 2017 2018 2019 2020 Original values 0.00 0.00 0.00 0.00 0.00 0.00 Actual values 0.00 30.00 70.00 100.00 100.00 Page 51 of 69 The World Bank National Safety Net Program for Results (P131305) Allocated amount ($) 0.00 0.00 0.00 0.00 0.00 0.00 Disbursed amount ($) 0.00 2,700,000.00 3,600,000.00 2,700,000.00 0.00 9,000,000.00 Comments (achievements against targets): The 100% of the critical FM Actions agreed with the Bank has been accomplished thus marking full achievement of DLI. ANNEX 1C. PROGRAM ACTION PLAN PAP_TBL Achieved Action Timing Completion Measurement (Yes/No) Revise OMs to respond to findings of the Recurrent Yearly Yes Update of the Operations Manual if/as needed. addendum to the Technical, and Integrated Fiduciary Assessments and the original ESSA Comments: Operations Manuals have been regularly updated. Improve delivery of payments to Recurrent Continuous No Regular payments to beneficiaries. beneficiaries: (i) reach agreement with Treasury on timeliness and flow of funds, (ii) regularly monitoring transfer steps against schedule, (iii) strengthening internal controls, (iv) developing long term PSP strategy Page 52 of 69 The World Bank National Safety Net Program for Results (P131305) Comments: Timeliness of payments continues to be a challenge mainly because of delayed release of funds by National Treasury. Internal controls have greatly improved with the implementation of the new payment mechanisms as well as the enhanced CCTP MIS. Implement IFAR recommendations with Recurrent Continuous Yes Timely procurements. respect to procurement Development and implementation of a Recurrent Yearly Yes Implementation of 100 percent of the agreed time-bound FM action plan to priority FM actions. comprehensively strengthen the fiduciary controls in the NSNP Finalize the EFC framework and action Due Date 30-Apr-2020 Yes Implementation of 100 percent of the agreed plan to prevent, detect, and deter EFC and critical FM actions. ensure alignment with the FM action plan Comments: EFC actions have been integrated within the revised OM and the Consolidated Action Plan, which includes agreed critical FM actions (see above). Also the new payment mechanism as well as the enhanced CCTP MIS has taken into account EFC actions. Finalization of a harmonized targeting tool Due Date 01-Apr-2020 Yes Updated HTM using new KIHBS data and update of the PMT, as and when new KIHBS data become available Comments: HTM has been finalized. The new regression based targeting formula has been developed based on most recent KHIBS data. The HTM is approved as part of the ESR strategy. NDMA uses it for re-registration in its four counties. Ensure financial allocations to the NSNP Recurrent Yearly No Adequate funding for implementation and according to the program MTEF are Page 53 of 69 The World Bank National Safety Net Program for Results (P131305) included in the sectoral MTEF and annual achievement of results. budgets Comments: Some of the Program results have not been achieved or not fully achieved because of inadequate budget allocation. Ensure collaboration with the NRB to Recurrent Continuous No Discussions held with NRB. enable issuance of ID cards to eligible beneficiaries who do not have ID cards Comments: Discussions with NRB have been carried out. It appeared that NRB`s field work to issue ID on case-by-case basis requires excessive effort that cannot be covered from NSNP budget. This inhibits effective collaboration for NSNP expansion. Ensure adequate implementation capacity Due Date 01-Apr-2020 No Implementation plan completed and expansion in nationwide and more specifically in the NEDI concluded. NEDI counties by (i) developing an implementation plan and (ii) implementation of the same. Comments: Implementation of NSNP expansion in NEDI counties has been interrupted due to lack of funding. Page 54 of 69 The World Bank National Safety Net Program for Results (P131305) ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION A. TASK TEAM MEMBERS Name Role Preparation Will Wiseman Senior Economist (TTL) Michael Munavu Social Protection Specialist Sarah Coll-Black Social Protection Specialist Cornelia Tesliuc Senior Social Protection Specialist Antonia Koleva Operations Officer Helen Craig Lead Human Capital Specialist Joyce Cheruto Bett Program Assistant Henry Amuguni Senior Financial Management Specialist Winston Onipede Cole Senior Financial Management Specialist Joel Munyori Senior Procurement Specialist Nyambura Githagui Senior Social Development Specialist Nightingale Rukuba-Ngaiza Senior Counsel Wolfgang Chadad Senior Finance Officer Pascal Tegwa Senior Procurement Specialist Supervision/ICR Yuliya Smolyar, Muhammad Iftikhar Malik Task Team Leader(s) Aleme Worku Tachbele, Joel Buku Munyori, Procurement Specialist(s) Mulugeta Dinka Henry Amena Amuguni Financial Management Specialist Nancy Makungu Gamusa Team Member Judith Sandford Team Member Vanessa Sigrid Tilstone Social Specialist Ben Okindo Ayako Miranga Environmental Specialist Kevin Sanya Heraniah Team Member Naseer Uddin Khan Team Member Page 55 of 69 The World Bank National Safety Net Program for Results (P131305) 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 0 205.00 FY13 22.454 516,376.69 FY14 2.819 24,977.14 Total 25.27 541,558.83 Supervision/ICR FY14 33.758 174,998.25 FY15 34.761 159,842.96 FY16 10.298 67,296.03 FY17 20.555 111,927.61 FY18 19.266 131,409.56 FY19 21.573 147,600.56 FY20 22.813 136,739.41 Total 163.02 929,814.38 Page 56 of 69 The World Bank National Safety Net Program for Results (P131305) ANNEX 3. PROGRAM EXPENDITURE SUMMARY Source of Estimates Actual Expenditures (Disbursement) Program Type of Co- at Financing Financing Percentage Percentage Appraisal Actual (Million US$) of Appraisal of Actual World Bank 250.0 195.5 78% 10.2% (IDA-52870) World Bank 50.0 35.8 72% 1.9% (IDA-59810) Borrower 411.2 1,501.1 365% 78.7% FCDO (DfID) Parallel 235.4 167.8 71% 8.8% UNICEF Parallel 7.8 8.2 105% 0.4% Total 954.4 1,908.4 200% 100% Source: SDSP and NDMA admin data, FCDO report, UNICEF report, PAD, Operations Portal. Note: Exchange rate of May 20, 2021 is applied to express the funding in US$ equivalent Page 57 of 69 The World Bank National Safety Net Program for Results (P131305) ANNEX 4. BORROWER’S COMMENTS The Borrower, represented by the State Department of Social Protection, has found the Implementation Competition and Results Report comprehensive and factual. The feedback noted that the ICR had drawn extensively on the Government`s Completion Report, which contains more information on the Program implementation thus complementing the presented analysis and conclusions. In the comments, the counterparts noted that that the expansion of OPCT program towards beneficiaries of 70 years and above has not been discussed in sufficient detail and called to complement the report with the numbers on the beneficiaries enrolled in the program in July 2017 and the budget allocated for their cash transfers. The ICR has been amended accordingly while respecting the limitations of the document`s length. Page 58 of 69 The World Bank National Safety Net Program for Results (P131305) ANNEX 5. SUPPORTING DOCUMENTS 1. Bahri, S. and Merttens, F. (2018) Evaluation of the Kenya Safety Net Programme Phase 2: Cost- efficiency analysis final report, Oxford Policy Management. 2. Gardner et al. (2017) “Evaluation of the Kenya Hunger Safety Net Programme Phase 2: The legacy of HSNP Phase 2: systems, practices and lessons learned, Oxford Policy Management. 3. International Development Association, Kenya Social and Economic Inclusion Project, Project Appraisal Document, Report No. PAD2911 dated November 2, 2018. 4. International Development Association, National Safety Net Program for Results, Project Appraisal Document, Report No. 78294-KE, dated June 26, 2013. 5. International Development Association, National Safety Net Program for Results, Program Paper on a Proposed Additional Credit, Report No. 111509 dated April 5, 2017. 6. International Development Association, National Safety Net Program for Results, Implementation Status and Results Report, Mission Aide Memoires, Restructuring Papers from 2012 through 2020. 7. Merttens et al (2017) Evaluation of the Kenya Hunger Safety Net Programme Phase 2, Impact Evaluation Final Report, Oxford Policy Management. 8. Ministry of East African Community, Labour and Social Protection, Inua Jamii Programme, Consolidation Strategy and Action Plan, August 21, 2016. 9. Ministry of Labour and East African Community Affairs, Programme Implementation and Beneficiary Satisfaction (PIBS) Survey for the Kenya National Safety Net Programme (undated). 10. Ministry of Labour and East African Community Affairs, Programme Implementation and Beneficiary Satisfaction (PIBS) Survey for the Kenya National Safety Net Programme (undated), PIBS II Final Report, October 2018. 11. Ministry of Labour and East African Affairs, Inua Jamii, Towards a more effective National Safety Net for Kenya, various Progress Reports. 12. Ministry of Labour and Social Protection, State Department for Social Protection, Directorate of Social Assistance, NSNP Completion and Assessment Report, March 2021. 13. National Safety Nets Program (NSNP), Operational Monitoring Reports, various cycles and dates. 14. Oxford Policy Management, Functional Review of Kenya National Safety Net Program (NSNP) Cash Transfer Programs, Main Report Part 1, October 2014. 15. Republic of Kenya, Recertification Strategy for the National Safety Net Program, May 2019. 16. World Bank, Implementation Completion and Results Report, Cash Transfer for Orphans and Vulnerable Children Project, Report No. ICR4501 dated June 19, 2019. 17. World Bank, Kenya Social Protection and Jobs Programs Public Expenditure Review, draft. 18. World Bank Group, Kenya Monthly Economic Data Update, March 2021 (internal). 19. World Bank Group, Realizing the Full Potential of Social Safety Nets in Africa, 2018. Page 59 of 69 The World Bank National Safety Net Program for Results (P131305) ANNEX 6: LINK OF OUTCOME AND INTERMEDIATE OUTCOME INDICATORS WITH DLIS, AS APPROVED AND AS RESTRUCTURED PDO Indicator Revised PDO Intermediate Revised DLI/DLR Revised DLI/DLR Indicator Result Indicator Intermediate Result (US$ millions) (US$, millions) Indicator Net change in beneficiary household monthly per adult equivalent consumption expenditures Intermediate Results Area 1: Expanding cash transfers to promote more comprehensive and equitable coverage Coverage Number of Number of new Number of additional 1c(i) Government takes over the NSNP households households enrolled in the financing of 15,000 CT-OVC beneficiaries, enrolled in the NSNP according to the beneficiary households by sex NSNP according expansion plan (in (US$3) to agreed comparison with the July expansion plan 2013 baseline) (US$100) 1d and paid for by Modified NSNP Expansion Plan, GoK including NEDI expansion, adopted (US$3 AF) 1c(ii) Government takes over the financing of 25,000 CT-OVC beneficiary households (US$5 AF) 1e(i) 30,000 additional households in the NEDI counties enrolled in the NSNP and paid for through the Government’s budget (US$11.57 AF) Page 60 of 69 The World Bank National Safety Net Program for Results (P131305) PDO Indicator Revised PDO Intermediate Revised DLI/DLR Revised DLI/DLR Indicator Result Indicator Intermediate Result (US$ millions) (US$, millions) Indicator 1e(ii) 40,000 additional households in the NEDI counties enrolled in the NSNP and paid for through the Government’s budget (US$15.43 AF) Number of Deleted households on the NSNP Payroll Equity % of Program System for DLI 2. Percent of program New DLI to replace DLI 2 Beneficiaries recertification beneficiaries who conform who conform of NSNP to the targeting criteria for 2b to the beneficiaries is the program in which they Plan adopted for how all four NSNP Targeting in place are enrolled (US$20) programs will maintain an up-to- Criteria Indicator Removed date registry of beneficiaries through recertification (US$5 AF) 2c Completion of the rollout of the recertification plan in at least 20 sub- counties (US$10 AF) Intermediate Results Area 2: Strengthening program systems to ensure good governance Timeliness % of all DLI 4. Percent of NSNP payments payments made disbursed to electronically using two- payment factor authorization service (US$15) providers on time Page 61 of 69 The World Bank National Safety Net Program for Results (P131305) PDO Indicator Revised PDO Intermediate Revised DLI/DLR Revised DLI/DLR Indicator Result Indicator Intermediate Result (US$ millions) (US$, millions) Indicator DLI 5. Percent of payments DLI 5. Percentage of payments made disbursed to Payment on time to Payment Service Service Providers on time Providers for transfer to Program (US$15) Beneficiaries Outreach and % of Governance beneficiaries who know program objectives and entitlements % of program beneficiaries who can name two means of making a complaint Grievance and DLI 6(1). DLI 6. Grievance and Case Case Functional complaint and Management mechanisms for the Management grievance mechanism Cash Transfer Programs are mechanisms for (US$15) functional the Cash Transfer Program are DLI 6(2). Percent of 6b(i) functional program beneficiaries who Beneficiary outreach strategy can name two means of developed (US$3) making a complaint (US$20) 6a(ii) Grievance and Case Indicator Removed and Management is functional at all included under new DLI levels for all four cash transfer 6b(iii) programs (US$10) 6b(ii) Beneficiary outreach strategy implemented in at least 20 sub- counties (US$10) Page 62 of 69 The World Bank National Safety Net Program for Results (P131305) PDO Indicator Revised PDO Intermediate Revised DLI/DLR Revised DLI/DLR Indicator Result Indicator Intermediate Result (US$ millions) (US$, millions) Indicator 6b(iii) Beneficiary awareness of transfer amount and G&CM system (US$7) % of complaints Deleted actioned M&E Single registry DLI 3. Single registry is fully functional with operational with program, program MISs MISs using agreed using agreed standards for internal standard for payroll controls (US$25) internal payroll controls Intermediate Results Area 3: Harmonizing cash transfer programs to increase the coherence of the safety net sector Finance % of NSNP DLI 9. Government spending which finances the HSNP in line is government with budget and policy financed commitments (US$5) Amount spent on Deleted NSNP by GoK as % of GDP Harmonizatio DLI 8. Strategy for n/Efficiency consolidating cash transfer programs (US$15) Number of DLI 7. System for scaling up households per the NSNP as part of the year paid national drought risk through the management system NSNP out of (US$20) National Drought and Disaster Contingency Fund (NDDCF) Page 63 of 69 The World Bank National Safety Net Program for Results (P131305) PDO Indicator Revised PDO Intermediate Revised DLI/DLR Revised DLI/DLR Indicator Result Indicator Intermediate Result (US$ millions) (US$, millions) Indicator NSNP beneficiaries’ DLI 10. NSNP beneficiaries’ access to access to complementary social and complementary productive services is promoted social and productive services is promoted 10a Complementary services module in SR is ready and formal data sharing protocols are available for use of potential partners (US$2 AF) 10b Agreements signed with at least two partners providing relevant complementary services to NSNP beneficiary households (US$1 AF) System Fiduciary systems for DLI 11. Fiduciary systems for NSNP Strengthening NSNP has been have been strengthened strengthened 11a Implementation of 30 % of agreed critical FM actions (US$2.7) 11b Implementation of 70% of agreed critical FM actions (US$3.6) 11c Implementation of 100% of agreed critical FM actions Page 64 of 69 The World Bank National Safety Net Program for Results (P131305) PDO Indicator Revised PDO Intermediate Revised DLI/DLR Revised DLI/DLR Indicator Result Indicator Intermediate Result (US$ millions) (US$, millions) Indicator (US$2.7) Page 65 of 69 The World Bank National Safety Net Program for Results (P131305) ANNEX 7: KEY OUTCOME INDICATORS WITH DLIS IN RELATION TO THE PROGRAM’S THEORY OF CHANGEa Improved Welfare of Program Beneficiaries Outcomes PDO: To support the Republic of Kenya’s efforts to establish and effective national safety net program for poor and vulnerable households • PDO Indicator: Net change in beneficiary household monthly per adult equivalent consumption expenditures • Number of NSNP • Increased % of program beneficiaries who • Increased % of NSNP spending which is government beneficiaries increased conform to the targeting criteria financed • • Intermediate Outcomes Increased % of all payments disbursed to Increased number of households per year paid payment service providers on time through the NSNP out of NDDCF • Increased % of beneficiaries who know program • Increased number of NSNP beneficiaries with access objectives and entitlements to complementary social and productive services • Increased % of beneficiaries for whom payments are made electronically through payment services providers using two factor authorization • Increased % of beneficiaries who can name two means of making a complaint • Fiduciary systems strengthened • New households • System for recertification of beneficiaries in • Strategy for consolidating cash transfer programs Outputs/Activities enrolled in the NSNP place • System for scaling up the NSNP as part of the according to the • Grievance and Case Management mechanisms national drought risk management system expansion plan for the cash transfer program are functional • Government finances the HSNP in line with budget • Number of households • Single Registry functional with program MISs and policy commitments on the NSNP Payroll using agreed standard for internal payroll • NSNP beneficiaries’ access to complementary social controls and productive services promoted • Expanding cash transfer • • Strengthening program systems to ensure good Harmonizing cash transfer programs to increase the Results Areas programs to promote governance coherence of the safety net sector more comprehensive and equitable coverage Note: a. This table includes only outputs, activities, and intermediate outcomes reflected as key indicators or DLIs. It does not include other activities that may have contributed to intermediate outcomes and outcomes (for example, communication campaigns, systems development, and so on) Bold: Reflected as both Key Indicator (PDO and/or Intermediate) and DLI. Italic: Introduced at restructuring. Underlined: Reflected in DLI only. Page 66 of 69 The World Bank National Safety Net Program for Results (P131305) ANNEX 8: PDO AND IR INDICATORS, AND DLIs by Program Results Area Results Area 1: Expanding cash transfer programs to promote more comprehensive and equitable coverage Box 4. Results Chain for Results Area 1 PDO 2: Number of NSNP beneficiaries, by sex Intermediate Result 1: New households enrolled in the NSNP according to agreed expansion plan and paid for by GoK Intermediate Result 2: Number of households on the NSNP payroll (deleted) DLI 1: Number of additional households enrolled in the NSNP according to the expansion plan (in comparison with the July 2013 baseline) DLI 1c(i): Government takes over the financing of 15,000 CT-OVC beneficiary households DLI 1d: Modified NSNP Expansion Plan adopted DLI 1c(ii): Government takes over the financing of 25,000 CT-OVC beneficiary households DLI 1e(i): 30,000 additional households in the NEDI countries enrolled in the NSNP and paid for through the Government’s budget DLI 1e(ii): 40,000 additional households in the NEDI countries enrolled in the NSNP and paid for through the Government’s budget Results Area 2: Strengthening program system to ensure good governance Box 5. Results Chain for Results Area 2 PDO 3: % of Program beneficiaries who conform to targeting criteria PDO 3 (revised): System for recertification of NSNP beneficiaries is in place PDO 4: % of all payments disbursed to payment service providers on time Intermediate Result 3: % of beneficiaries who know program objectives and entitlements Intermediate Result 4: % of program beneficiaries who can name two means of making a complaint Page 67 of 69 The World Bank National Safety Net Program for Results (P131305) Intermediate Result 5: Grievance and Case Management mechanisms for Cash Transfer Program are functional Intermediate Result 6: % of complaints actioned (deleted) Intermediate Result 7: Single registry functional with program MISs using agreed standard for internal payroll controls Intermediate Result 12: Fiduciary system for NSNP have been strengthened DLI 2: Percent of program beneficiaries who conform to the targeting criteria for the program in which they are enrolled DLI 2 (revised): System for recertification of NSNP beneficiaries in place DLI 2(i): Establishment of baseline (of performance to the targeting criteria for their program) DLI 2(ii): Plan adopted for how all four NSNP programs will maintain an up-to-date registry of beneficiaries through recertification DLI 2(iii): Completion of the rollout of the recertification plan in at least 20 sub-counties DLI 3: Single registry if fully operational with program, MISs using agreed standards for internal payroll controls DLI 3(i): MISs operational DLI 3(ii): Single Registry operational DLI 4: Percent of NSNP payments made electronically using two-factor authorization DLI 5: Percent of payments disbursed to Payment Service Providers on time DLI 5 (revised): Percentage of payments made on time to Payment Service Providers for transfer to Program Beneficiaries DLI 6(1): Functional complaint and grievance mechanism DLI 6(2): Percent of program beneficiaries who can name two means of making a complaint DLI 6 (revised): Grievance and Case Management mechanisms for the Cash Transfer Programs are functional DLI 6a(i): Beneficiary outreach strategy developed DLI 6a(ii) Grievance and Case Management is functional at all levels for all four cash transfer programs DLI 6b(ii): Beneficiary outreach strategy implemented in at least 20 sub-counties DLI 6b(iii): Beneficiary awareness of transfer amount and G&CM system DLI 11 (new): Fiduciary systems for NSNP have been strengthened DLI 11a (new): Implementation of 30% of agreed critical FM actions DLI 11b (new): Implementation of 70% of agreed critical FM actions DLI 11c (new): Implementation of 100% of agreed critical FM actions Page 68 of 69 The World Bank National Safety Net Program for Results (P131305) Results Area 3: Harmonizing cash transfer programs to increase the coherence of the safety net sector Box 6. Results Chain for Results Area 3 Intermediate Result 8: % of NSNP spending which is government financed Intermediate Result 9: Amount spent on NSNP by GoK as % of GDP (deleted) Intermediate Result 10: Number of households per year paid through the NSNP out of NDEF Intermediate Result 11: NSNP beneficiaries’ access to complementary social and productive services is promoted DLI 7: System for scaling up the NSNP as part of the national drought risk management system DLI 8: Strategy for consolidating cash transfer program DLI 9: Government finances the HSNP in line with budget and policy commitments DLI 10 (new): NSNP beneficiaries’ access to complementary social and productive services is promoted DLI 10a (new): Complementary services module in SR is ready and formal data sharing protocols are available for use of potential partners DLI 10b (new): Agreements signed with at least two partners providing relevant complementary services to NSNP beneficiary households Page 69 of 69