Performance-Based Fiscal Transfers for Urban Local Governments Results and Lessons from Two Decades of World Bank Financing Hyunji Lee, Sohaib Athar, Jesper Steffensen, Roland White, Ayah Mahgoub © 2022 The World Bank Group 1818 H Street NW, Washington, DC 20433, USA Telephone: 202-473-1000; Internet: www.worldbank.org All rights reserved. This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@ worldbank.org Citation Please cite the report as follows: Hyunji Lee, Sohaib Athar, Jesper Steffensen, Roland White and Ayah Mahgoub. 2022. Performance-Based Fiscal Transfers for Urban Local Governments: Results and Lessons from Two Decades of World Bank Financing. Washington, D.C.: World Bank. Performance-Based Fiscal Transfers for Urban Local Governments 3 Results and Lessons from Two Decades of World Bank Financing Contents Acknowledgements 6 4.4. Importance of capacity building to support program implementation and sustainability 58 Executive summary 7 4.5. Summary of lessons and good practices for urban Glossary of key terms used 16 performance grant programs 61 Abbreviations 17 5. The Way Forward and Emerging Issues for Urban Performance Grant Programs 62 Part I 5.1. Ensuring the sustainability of UPG programs within government systems 62 Key findings and lessons for Performance 5.2. From institutional processes to service delivery outcomes: The next generation of programs 66 Grant Programs 18 5.3. Promoting climate change action at the local level through UPGs 71 1. Introduction: Why Urban Performance Grants? 19 1.1. Context and motivation for this report 20 References 72 1.2. Objectives, methodology, and structure of this report 22 2. Role and Evolution of Urban Performance Grants in Improving Local Government Outcomes 23 Part II 2.1. Key features of urban performance grants: Case Studies from Selected Countries 73 How they work 23 6. Ethiopia: Second Urban Local Government 2.2. Evolution of urban performance grant programs 27 Development Program and Urban Institutional and 3. Key Findings from Review of World Bank Financing for Infrastructure Program 75 Urban Performance Grant Programs 31 7. Ghana: Local Government Capacity Support Project and 3.1. Performance of UPG portfolio relative to other urban Ghana Secondary Cities Support Program 79 sector projects 31 8. India: Institutional Strengthening of Gram Panchayats 3.2. Results of nine case-study UPG programs across Project Phase II 83 key performance areas 33 9. Kenya: Kenya Devolution Support Program 87 4. Lessons and Best Practices for Urban Performance Grant Programs 43 10. Tanzania: Urban Local Government 4.1. Three core components of a successful Strengthening Program 91 UPG program 43 11. Tunisia: Urban Development and Local 4.2. Lessons and good practices on key design factors Governance Program 95 for urban performance grants 44 4.3. Lessons on implementation of urban 12. Uganda: Uganda Support to Municipal Infrastructure performance grants 52 Development Program 100 Performance-Based Fiscal Transfers for Urban Local Governments 4 Results and Lessons from Two Decades of World Bank Financing Figures 2.1. Typology of Performance-based 3.7. Annual Execution Rate of Planned 4.10. Summary of Lessons and Based Grant Programs along Two Infrastructure Investments at the Good Practices for Effective Dimensions 23 Local Level 40 and Credible Performance Assessments 57 2.2. Stylized Structure of Urban 4.1. Mutually Strengthening Core Performance Grant Programs 24 Components of a UPG Program 43 4.11. Recommendations for Effective Capacity-Building Support 60 2.3. Example of How UPGs Work and 4.2. Four Key Elements of the Urban Align with Government Systems 25 Performance Grant Design 45 5.1. Increasing and Sustaining Impact of UPG Programs 66 2.4. World Bank’s Cumulative 4.3 Share of UPGs in Total Fiscal Financing for UPG Programs since Transfers to Participating Local 2000 28 Governments 46 2.5. Evolution of UPG Programs in 4.4. Relationship between Average Each Country 29 Sizes of UPGs per Capita and Capital Investment Projects 47 3.1. Share of UPGs Out of Urban Portfolio 31 4.5. Typical Linear Relationship between Local Government 3.2. Cost Effectiveness of UPG Performance and Grant Allocation 49 Programs 32 4.6. Summary of Lessons and Good 3.3. Disbursement Trends of UPG Practices on Key Design Factors and non-UPG Programs Over a for Urban Performance Grants 51 Standardized Project Timeline 32 4.7. Share of the Cost of Conducting 3.4. Common Types of Minimum APAs out of Total UPG Program Conditions across the UPG Financing 52 Programs Reviewed 34 4.8. Spectrum of Options to Improve 3.5. Increase in Average APA Score Credibility of the Performance over Program Period for All Local Assessment 54 Governments in Program 35 4.9. Typical APA Cycle during a Fiscal 3.6. Percentage Increase in Own- Year Spanning January 1 to Source Revenue Over Program December 31 56 Period 37 Performance-Based Fiscal Transfers for Urban Local Governments 5 Results and Lessons from Two Decades of World Bank Financing Boxes Tables 1.1. Scope of World Bank Financing 1.1. Overview of the Reviewed 10.1. Factsheet for the Case UPG Programs Overseen by GPURL 20 UPG Programs 21 Program in Tanzania 91 2.1. Typical Minimum Conditions 3.1. Increase in Minimum Condition 10.2. Key Performances of the Case and Performance Measures in Compliance Rate across UPG Program in Tanzania 92 UPG Programs 26 UPG Programs 33 11.1. Factsheet for the Case UPG 3.2. Employment Opportunities Program in Tunisia 95 2.2. World Bank Financing for UPG Created via Labor-Intensive Public 11.2. Key Performances of the Case Projects and Programs Using Works in Selected UPG Programs 41 UPG Program in Tunisia 96 Various Financing Instruments 30 5.1. Share of Government 12.1. Factsheet for the Case UPG 3.1. Use of DLIs to Provide Incentive Co-Financing of UPG Programs in for Institutional Actions in Programs in Uganda 100 Selected Countries 64 UPG Programs 36 12.2. Key Performances of the Case 6.1. Factsheet for the Case UPG UPG Programs in Uganda 101 3.2. Impact of UPG on Own- Programs in Ethiopia 75 Source Revenue Collection in 6.2. Key Performances of the Case Mozambique 38 UPG Programs in Ethiopia 76 3.3. Value for Money Audit in Uganda 42 7.1. Factsheet for the Case UPG Programs in Ghana 79 Map 4.1. Lessons on Improving the 7.2. Key Performances of the Case Equity of Performance Grant 2.1. Geographical Coverage of World UPG Programs in Ghana 80 Bank-Financed UPGs Overseen Allocations among Local Governments 48 8.1. Factsheet for the Case UPG by GPURL 28 Program in India 83 4.2. An Example of Exponential 8.2. Key Performances of the Case Grant Allocation Design in Kenya 50 UPG Program in India 84 4.3. Impacts of a Strategic Capacity- 9.1. Factsheet for the Case UPG Building Approach on Local Program in Kenya 87 Governments in West Bengal State, India 59 9.2. Key Performances of the Case UPG Program in Kenya 88 5.1. Building Sustainability through a Long-Term Engagement in Bangladesh 63 5.2. Recent Sector-Specific UPGs in Kampala, Uganda, and Kerala, India 70 Performance-Based Fiscal Transfers for Urban Local Governments 6 Results and Lessons from Two Decades of World Bank Financing Acknowledgements This report was prepared by a team comprising Hyunji Lee (Urban Specialist), Sohaib Athar (Senior Urban Specialist), Jesper Steffensen (Consultant and Senior Partner, Dege Consult), Roland White (Global Lead for City Management, Governance and Finance) and Ayah Mahgoub (Senior Urban Specialist). Yousra Abdelrahman (Program Analyst) carried out support activities for data collection and analyses. Administrative support was provided by Adelaide Barra (Program Assistant). The report was edited by Lisa Ferraro Parmelee and cover and production design was done by WeAreZephyr. The work was conducted under the overall guidance of Sameh Wahba (former Global Director, Global Practice for Urban, Disaster Risk Management, Resilience, and Land Global Practice) and Maitreyi Das (Practice Manager, Global Programs unit). The team is immensely grateful to the following colleagues for their cooperation and time in providing information for the various country cases presented in this report as well as their expertise and comments over several rounds of discussion: Abdu Muwonge, Berhanu Legesse Ayane, Davison Muchadenyika, Dinkneh Tefera, Gottfried Roelcke, Hannah Kim, Lamia Zaki, Martin Onyach-Olaa, Manvinder Mamak, Muratha Kinuthia, Riddhiman Saha, Salim Rouhana, Shenhua Wang, Smile Kwawukume, Stephen John Ajalu, Su Jung Song, Uri Raich, Wahid Kraiem and Vasudha Thawakar. The team is also grateful to the following peer reviewers for their valuable guidance and comments, which enriched this report: Catalina Marulanda (Practice Manager, Urban, DRM, Resilience and Land, East & Southern Africa), Yoonhee Kim (Lead Urban Specialist), and Timothy Stephen Williamson (Senior Public Sector Specialist). Performance-Based Fiscal Transfers for Urban Local Governments 7 Results and Lessons from Two Decades of World Bank Financing Executive Summary Setting the Stage: An Introduction to Urban Performance- Figure ES.1. Cumulative World Bank Financing for UPG Based Grants Programs under GPURL since 2000 The World Bank has an extensive portfolio of programs that provide performance-based grants (PBGs) to local and other subnational governments as an incentive to improve institutional performance 9 and service delivery outcomes. The Bank’s Global Practice for Urban, Disaster Risk Management, Resilience, and Land (GPURL) has led the 8 way in supporting these programs over the past two decades, overseeing 7 over US$8 billion in financing for them in this period across several regions (Figure ES.1).1 In the last ten years, a second generation of such 6 Billions of US dollars programs has substantially expanded them, both in financing volume 5 and global geographical scope. The PBG programs overseen by GPURL Launch of PforR are broadly referred to as urban performance grant (UPG) programs. 4 financing instrument UPG programs generally have the following key design features: 3 • UPGs use performance-based grant financing to provide incentives to 2 local governments to improve their institutional performance and service delivery outcomes in their jurisdictions. 1 • The programs rely on a robust, credible, and transparent common 0 annual performance assessment (APA) system to evaluate local 2000 2010 2015 2020 governments based on established criteria. The results of the assessments form the basis for the fiscal transfer (grant) allocations First generation Second generation to these local governments, which receive grants from a central or regional agency, such as the Ministry of Finance, on the achievement of predefined result targets. Source: Team analysis based on data from World Bank’s internal Operations Portal standard report and Business Warehouse database on June 30, 2021. • Grant funding is generally used for capital investments (infrastructure Note: The scope of the Urban Performance Grant programs included in this figure covers a total and service delivery). The investments are measured and reported on, of 33 financing operations approved by the World Bank Group Board of Executive Directors but the participating local governments have discretion over them, and overseen by the Global Practice of Urban, Disaster Risk Management, Resilience, and Land between fiscal years 2000 and 2020. It includes the first and second generations of UPG financing within limits, in line with local plans and priorities. programs, comprising over US$8 billion in approved financing. 1 Other global practices of the World Bank also support PBG programs, including the Governance Global Practice and the Social Sustainability and Inclusion Global Practice. The scope of this report is limited to programs supported by GPURL. Performance-Based Fiscal Transfers for Urban Local Governments 8 Results and Lessons from Two Decades of World Bank Financing • Capacity-building support is provided to different tiers of government. higher than the average Bank financing for non-UPG urban sector Both the central government agency or ministry administering the projects. While UPG programs tend to cost the Bank more to prepare program and the recipient local governments receive this support, and provide implementation support for—with total preparation and which is sometimes delivered to the latter through earmarked annual implementation support costing 24 percent and 34 percent more capacity-building grants and is typically closely linked with the on average respectively —they are much more cost effective than non- results of the performance assessment system. UPG urban sector programs in terms of per-million-dollar financing. • UPG programs generally cover multiple cities, municipalities, and Most local governments in the UPG programs reviewed in this report urban local governments. In some cases, however, a single large city or metropolitan area may be targeted for specific purposes. Urban as successfully complied with eligibility criteria (minimum conditions, well as rural local governments may be grant recipients, depending or MCs) over the lives of their respective programs. In almost half on program objectives and the specific political, administrative, and of the reviewed programs, all the participating local governments institutional context. continuously met the MCs throughout the entire program periods, In the more than two decades they have been in use, UPGs have while the rest started with lower rates that substantially increased been acknowledged as contributing to improved local government over time. The former included countries with previous experiences performance in several areas. It is important, then, to systematically in UPG systems and preparatory activities carried out before launch take stock of the implementation experience, results, and performances of the programs, which prepared them sufficiently to comply with the of these programs in achieving intended outcomes and to identify key conditions. lessons for the design and implementation of future programs that The overall performance of participating local governments in areas support this agenda. In addition to pursuing these objectives, this report targeted for institutional strengthening significantly improved across provides key metrics and high-level analysis of the UPG portfolio and all UPG programs, as measured in rigorous APAs. One key metric is presents in-depth case studies of selected UPG programs, focusing on how the average performance score for the local governments within their achievements, challenges, and implementation experiences. each program changed during its duration. Across all the UPG programs Key Findings from Review of World Bank Financing for reviewed, the average performance score for local governments— UPG Programs comprising specific performance metrics (PMs) pertaining to key institutional and capital investment processes—increased by as much as UPG programs represent a sizeable share of the World Bank’s urban 50 percent (Figure ES.2 shows this increase for each program reviewed). portfolio2 and are substantially larger in financing size and more The UPG programs enhanced urban systems mainly by improving the cost-effective than non-UPG urban sector projects financed by the generation of own-source revenues (OSR); public financial management Bank. Between Fiscal Year (FY) 13 and FY20, they account for 41 percent (PFM); transparency, accountability, and citizen engagement; capital of the urban portfolio in financing volume, amounting to US$5.2 billion investment design, execution, and maintenance; asset management; and in total, and 56 percent of total International Development Association human resources (HR) management and local staffing. The key enablers (IDA) financing in the portfolio (US$3.5 billion). The average Bank of such improvements combined strong financial incentives with targeted financing for a UPG program is US$275 million, which is 2.3 times capacity-building support, both before and during the implementation. 2 See Box 1.1 for a description of the urban portfolio and other financing programs overseen by GPURL. Performance-Based Fiscal Transfers for Urban Local Governments 9 Results and Lessons from Two Decades of World Bank Financing Figure ES.2. Increase in Average Score of APA Score over Figure ES.3. Percentage Increase in Own-Source Revenue Program Period for All Local Governments in Program over Program Period 163 Ethiopia (ULGDP II) 160 160 148 140 140 Ghana (LGCSPI) 120 120 Uganda (USMID) Increase in OSR (%) Increase in OCR (%) 100 100 Tanzania (ULGSP) 83 80 80 Tunisia (PDUGL) 67 66 60 60 Kenya (KDSP) 48 48 40 40 40 50 60 70 80 90 100 20 20 APA score (0-100 points) 0 0 Baseline Recent score Ethiopia Ghana Uganda Tanzania Tunisia Kenya Ethio (ULGDP II) (LGCSP) (USMID) (ULGSP) (PDUGL) (KDSP) (ULGD Source: Team analysis based on relevant program documents and data provided by World Bank Source: Team analysis based on relevant program documents and data provided by World Bank task teams for respective UPG programs. task teams for respective UPG programs. Note: Of nine reviewed UPG programs, three are not included because of data unavailability (i.e., Note: Percentage increase in OSR in nominal figures in local currency. Two UPG programs in it was too early to measure the improvement or too hard to track the trend due to changes to the Ghana (GSCSP) and Tunisia (PDUGL) were not included because of data unavailability. measurement system). OSR increased significantly in seven out of nine UPG cases, for example. UPG programs have also financed the rehabilitation or development of Six showed an absolute increase in OSR collection, including Ethiopia substantial capital infrastructure investments and assets for improved (ULGDP), Ghana, India-West Bengal, Kenya, and Uganda (Figure ES.3), service delivery to citizens and businesses. Figure ES.4 summarizes the while in Ethiopia (UIIDP) the share of participating local governments improved infrastructure and services outputs financed by the reviewed with 10 percent annual increases in OSR rose from 77 percent in the first UPG programs. year to 85 percent in the second. The positive impact of UPGs on OSR regeneration has not been limited to the reviewed programs but has also been found in other cases, as shown in recent empirical research on a UPG program in Mozambique. Performance-Based Fiscal Transfers for Urban Local Governments 10 Results and Lessons from Two Decades of World Bank Financing Figure ES.4. Snapshot of Results Achieved: Improved program, each of which needs to be carefully designed and implemented Infrastructure and Services Outputs Financed by UPG to achieve the development objective. Lessons and best practices on these Programs three components are summarized below. In addition, at an overarching level, UPG programs should ideally be embedded into existing IGFT systems to leverage these systems for fund transfer and oversight. 150m >1m 1) Lessons and good practices on the design of UPGs beneficiaries employment opportunities created, Ensuring UPGs are substantial in size—in both relative and absolute in person-years terms—is important to provide local governments with incentive to achieve performance targets and finance investments with impact. UPGs should constitute a meaningful share of all capital (or total) fiscal 7,000km 350ha 2,000 transfers to the recipient local governments, preferably comprising roads; of parks; classrooms; 160km of bicycle/ 90 local markets; 28 community more than 20 percent of the annual capital grants allocated to them. pedestrian ways; 15 social facilities health compounds; Grant size should also be substantial on a per capita basis to enable the over 1,270km of (community centers, etc.); 14 health clinics drainage 30 security facilities governments to deliver meaningful services to citizens and beneficiaries, preferably amounting to not less than US$5–10 per capita annually for 8,700 290 typical local governments in charge of core urban services. Finally, streetlights; waste collection points; UPGs should enable local governments to undertake larger capital 28 transit facilities 1,300 trashcans; (bus terminals, etc.) 3 dumpsites investments, generally for infrastructure, of at least US$150,000–200,000 per project, especially in urban areas. This report discusses these general Source: Compiled by authors using data from various UPG programs recommendations, as well as design proposals for contexts where the recommended levels are not feasible, especially when the policy preference is for an equitable distribution of grants for a much larger set Lessons and best practices for Urban Performance of local governments, as well as substantial geographical variation among Grant programs the governments across urban and rural areas. Figure ES.5 summarizes This review provides evidence and lessons learned on key design features key lessons on design factors for successful UPG programs, while noting and implementation drivers of UPG programs, which are applicable the interaction and mutual influence among the various factors. to other PBG programs for local and other subnational governments. The enablers of successful UPG programs are a combination of strong financial incentives provided through adequate grant design parameters; grant operationalization and implementation modalities (particularly an APA system) that are rigorous, credible, and transparent and fully integrated with the government’s budget cycle and systems; and the coordinated delivery of capacity-building support and technical assistance to local governments to achieve improved performance. These are the three mutually reinforcing core components of a successful UPG Performance-Based Fiscal Transfers for Urban Local Governments 11 Results and Lessons from Two Decades of World Bank Financing Figure ES.5. Summary of Lessons and Good Practices on Key Design Factors for Urban Performance Grants Share of UPGs relative to other funding 1 2 Monetary size of grant on per capita basis • Although it depends on local context, • Nominal grant size needs be at least UPG should not be <20% of total US$5-10 per capita for a typical development grants to urban local Key Design urban local government. governments. Factors for • Nominal size may be less important UPGs than UPG share relative to total budget. Grant allocation formulas can mitigate funding size risks 4 3 Average size of investment projects • Allocation should be proportional to • Medium to larger investment projects performance. (over US$200,000 per project) • When total available funding is are encouraged to make meaningful small (on a per capita basis) relative impacts, especially in urban areas. to total fiscal transfers, mitigating • Grant size to be determined by local design measures work. context, size of local governments, and whether they are urban or rural. Performance-Based Fiscal Transfers for Urban Local Governments 12 Results and Lessons from Two Decades of World Bank Financing 2) Lessons on operationalization and implementation of UPGs The Way Forward: Issues for the Next Generation of In addition to a well-designed program, it is important to develop a UPG Programs performance measurement and assessment process that is rigorous, The next generation of UPG programs will need to persist with lessons credible, and objective enough to build trust among stakeholders and learned from the earlier programs and innovate further as new issues provide adequate performance incentives to local governments. An emerge. Such programs are being required to tackle several emerging effective APA is designed to be objective by being contracted out to issues that affect their design and implementation: an entity external to the program, credible by having a multilayered • How to sustain the impacts of the programs beyond the program verification and quality assurance system, and rigorous through the period and embed them into intergovernmental fiscal frameworks selection of adequate and targeted performance indicators. Local governments often require technical assistance to implement such an • How to provide more direct incentives for improved service delivery outcomes, in addition to improved institutional performance at the APA system, but its use helps provide them with strong performance local level incentives. It is also essential to ensure predictability and timeliness in the allocation and disbursement of UPGs to local governments by • How to promote local climate change action by providing financial incentives to local governments. aligning the APA timelines with the governments’ annual budget and intergovernmental fiscal transfers cycles. Figure ES.6 summarizes 1) Ensuring sustainability of UPG programs within key lessons for effective and credible performance assessments and government systems program operationalization. There is some evidence emerging of the sustainability of UPG interventions within government systems after program closure, but with caveats. In 3) Importance of capacity building to support program most of the cases reviewed, the governments retained (or replicated implementation and sustainability or even scaled up) several features of their UPG programs even after Providing capacity-building support to local governments is important the programs closed, and aspects of UPGs became an element of the to strengthen their institutional performance and improve outcomes in IGFT systems in these countries. In addition, UPG programs in several infrastructure and service delivery. Well-designed capacity-building countries strengthened local institutions and contributed to policy support targets specific areas that pose institutional barriers to dialogues on fiscal decentralization, with governments continuing or improving performance and is especially helpful to low-performing local incorporating these improvements into their institutional systems. governments. Most of the UPG programs reviewed adopted a balanced The strong results have also had clear demonstration effects, leading to capacity-building strategy combining supply- and demand-driven considerable geographical expansion of such programs within countries activities, often financed through a separate funding window: a small and indicating strong government interest in their outcomes. These capacity-building grant. Figure ES.7 summarizes recommendations for experiences show that Bank support to UPG-based intergovernmental effective capacity-building support under UPG programs. fiscal transfers should be thought of as long-term partnerships with governments with time horizons of 15 years or more, with consistent and increasingly sophisticated support and incentives to local governments over this period. This will also provide governments with fiscal space for evolving and incrementally improving UPG systems. Performance-Based Fiscal Transfers for Urban Local Governments 13 Results and Lessons from Two Decades of World Bank Financing Figure ES.6. Summary of Lessons and Good Practices for Effective and Credible Performance Assessments Well-designed performance indicators • Align APA thematic areas with objectives of UPG program • Select indicators that are under local government control • Balance between comprehensiveness and impacts Neutral and objective APA protocols • Conduct through third party • Can be seen as similar with systems of external audit Robust verification and quality assurance system • External, “two-layer” quality assurance system is recommended for credibility Clear and transparent reconciliation system • Systematically reconcile any discrepancies found in APA results • Close collaboration between World Bank and government Value for Money (VfM) Audits • Utilize VfM audit findings into APA indicators where feasible • VfM costs need to be affordable relative to grant size (more relevant for larger urban projects) Launch of APA Reflections of APA results Synchronize program launch with • Disseminate & publicize APA results for transparency & accountability local government budget cycle • Use APA results as input to capacity building support focusing on weak performance areas • Incorporate APA results into overall program monitoring system Performance-Based Fiscal Transfers for Urban Local Governments 14 Results and Lessons from Two Decades of World Bank Financing Figure ES.7. Recommendations for Effective Capacity-Building Support Combine the typical Combine demand-and-supply- capacity-building support driven approaches with system development • Provide a specific capacity-building • Combine capacity-building grant allocation or ceiling in UPGs thematic areas with system Effective development such as audit, IT, Capacity-Building PFM, and M&E to ensure Support a conducive framework Capacity-building result Equitable support monitoring • Reflect APA results in capacity- • Monitor results (e.g., through building planning disbursement-linked Indicators) • Support weakest local governments and links between intermediate to climb up the performance ladder and results indicators and compete on equal terms Performance-Based Fiscal Transfers for Urban Local Governments 15 Results and Lessons from Two Decades of World Bank Financing 2) From institutional processes to service delivery outcomes: contexts where local governments have built strong processes and The next generation of UPG programs systems, UPG programs can also aim to move up the value chain with Most UPG programs aim to resolve institutional bottlenecks for service more sophisticated programs or target institutional performance and delivery and provide incentives to improve local governments’ core output indicators for specific municipal services. Entire programs can institutional processes and systems, as these are the basic building blocks be designed with a sector-specific focus, targeting performance factors of their performance, accountability, and effectiveness in delivering that are closer to the point of service delivery and alleviating critical services. The higher-level goal of these programs—improved service constraints to delivery outcomes. Some promising recent examples, delivery—is pursued through an instrument based on a theory of including improved solid waste management services and metropolitan change that is cognizant of the low-capacity and governance-challenged coordination and management for better service delivery, are reviewed institutional environments in which the programs operate. Evidence is in the report. emerging that among the important drivers for better service delivery 3) Promoting climate change action at local level is better institutional processes for public investment management, Attention is increasing on promoting local climate actions through accomplished by improving the efficiency and quality of capital financial incentives under UPG systems. In many cases, the United investment expenditure. As a result, stakeholders have continued to Nations Capital Development Fund (UNCDF) has been taking the lead demand direct incentives for and measurement of improvements in in this area through its Local Climate Adaptive Living Facility (LoCAL), service delivery at the local level and the tying of UPG funding for which provides performance-based climate resilience grants in several local governments to these outcomes. It is important to acknowledge countries.3 More generally, PBG programs are increasingly being utilized the challenges in doing so, however, especially for multisectoral UPG to focus more on climate change mitigation and adaptation at the local programs, where the local governments generally have discretion level, either through standalone grants or funding integrated into to spend the funds across multiple sectors within their remits existing multisectoral grants. These initiatives are linked with a targeted and expenditure assignments, which makes it difficult to compare performance assessment system and capacity-building support for local performance across sectors. Other issues include equity and attribution governments. The Bank has also started piloting climate-focused UPG of results, lack of flexibility and discretion for local governments, programs in some countries, including Ethiopia and Kenya. While most and time lags between inputs and results. Providing incentives and such programs have focused on adaptation, incentive is also needed for measuring outcomes may, therefore, be much more suitable for sector- local actions toward mitigation. A key tool for this may be the menu specific grant programs, where the sectors to which use of the grant of eligible investments available to local governments, which should funds is tied are predefined. include investments to reduce local greenhouse gas emissions, such One useful approach to overcoming these challenges—especially in as funding for public and nonmotorized transportation, solid waste multisectoral grants where local governments have a broad mandate to management to reduce methane emissions, and green public spaces, expend funds across sectors—is the use of proxy-type indicators to provide among others. incentive for service delivery outcomes. Such indicators, which might include results from Value for Money audits and capital expenditure execution rates (for example, measuring the efficiency, effectiveness, and quality of expenditure from UPG funds), among others, were generally effective in the UPG programs reviewed in this report. In 3 See the report for more information on this. Performance-Based Fiscal Transfers for Urban Local Governments 16 Results and Lessons from Two Decades of World Bank Financing Glossary of key terms used Annual performance assessment (APA) is a commonly used performance assessment protocol in UPG programs. APAs are conducted annually, in some cases by an independent verification agent (IVA), to Performance-based grants (PBGs) are a specific type of ensure the credibility of performance results on which—complemented intergovernmental financing mechanism, whereby local and other by various quality assurance systems—the size of the grant disbursed subnational governments receive fiscal transfers (grants) from higher to local governments is based. Chapter 4 provides details on the APA tiers of government, conditioned on their assessed performance in process. predetermined areas. (“Local and other subnational governments” Infrastructure investments as referenced in this report are specific can refer to provinces, states, governorates, counties, districts, investments in municipal infrastructure and services that are funded municpalities, corporations, and various other types of urban and and supported by UPG programs. Typically, over 80 percent of the total rural local governments.) PBGs are generally integrated into national funding available under such programs is used to cover the costs of intergovernmental fiscal transfer (IGFT) systems. They are typically executing the planned infrastructure and service delivery. designed to improve the institutional performance and service delivery Minimum conditions (MCs) are basic requirements with which all of the targeted local governments through a set of financial incentives, participating local governments should comply to become eligible for often in support of decentralization and devolution objectives of grants tied to the achievement of performance measures (see below). the national government, but they may be focused on a wider range The role of MCs is to ensure fiduciary comfort and that funds will be of performance areas. PBGs can be categorized along two basic absorbed with due safeguards in place. See chapter 1 for more details. dimensions: the type of performance the grant is trying to leverage, Performance measures (PMs) represent a set of predefined targets particularly in terms of generic institutional performance or sector- local governments should achieve, covering various thematic areas, specific performance; and the use of funds—that is, either discretionary with more refined and qualitative indicators than MCs. PMs determine or earmarked for certain types of uses or sectors.4 the size of grants allocated to local governments, and they also inform Urban performance grants (UPGs or UPG programs) are a subset capacity-building needs and overall monitoring and evaluation systems. of performance-based grants that promote enhancement of the See chapter 1 for more details. institutional performance and service delivery of an identified range of Investment Project Financing (IPF), a World Bank financing instrument, local governments in developing countries, with the support of the World helps countries develop their physical and social infrastructures and Bank’s Global Practice for Urban, Disaster Risk Management, Resilience service delivery in the medium to long terms (five to ten years). It also and Land (GPURL). They mostly cover multiple cities, but in some cases works as a vehicle for global knowledge sharing and technical assistance. a single large city can be targeted for specific purposes—for instance, to support the local economic development of a certain metropolitan area. Program-for-Results (PforR), another financing instrument offered Furthermore, while they generally apply to urban local governments by the World Bank, links the disbursement of funds directly to the (such as municipalities, metropolitan governments, corporations, and so achievement of predefined performance results using a country’s own on), both urban and rural local governments may be included, depending institutions and processes. Its aims are to build capacity within client on program objectives and the country’s institutional context. countries, enhance the effectiveness and efficiency of investments, and lead to the achievement of scalable, sustainable impacts. See chapter 1 for more details on the World Bank’s financing instruments. 4 See UNCDF 2010 for PBG systems concepts and experience. Performance-Based Fiscal Transfers for Urban Local Governments 17 Results and Lessons from Two Decades of World Bank Financing Abbreviations O&M Operation and maintenance APA Annual performance assessment OPAMS Online Performance Assessment and CFC Central Financial Commission (in India) Monitoring System CPSCL Municipal Development Fund (in Tunisia) OSR Own-source revenue DLI Disbursement-linked indicator PBG Performance-based grant DPAT District Assembly Performance Assessment Tool PDUGL Urban Development and Local Governance (in Ghana) (in Tunisia) DRM Disaster risk management PforR Program-for-Results E&S Environmental and social systems PFM Public financial management FOAT Functional and Organisational Assessment Tool PIM Public investment management (in Ghana) PIMA Public Investment Management Assessment GoB Government of Bangladesh PM Performance measure GPURL Global Practice for Urban, Disaster Risk SFC State Fiscal Commission (in India) Management, Resilience and Land SWM Solid waste management GSCSP Ghana Secondary Cities Support Program UIIDP Urban Institutional and Infrastructure Development HR Human resource Program (in Ethiopia) IDA International Development Association ULGDP Urban Local Government Development Program IGFT Intergovernmental fiscal transfer (in Ethiopia) IMF International Monetary Fund ULGSP Urban Local Government Support Program IPF Investment Project Financing (in Tanzania) ISGPP Institutional Strengthening of Gram Panchayats UNCDF United Nations Capital Development Fund Project (in India) UPG Urban performance grant IVA Independent verification agent USMID Uganda Support to Municipal Infrastructure KDSP Kenya Devolution Support Program Development LoCAL Local Climate Adaptive Living Facility USMID AF Uganda Support to Municipal Infrastructure LGCSP Local Government Capacity Support Project Development Additional Financing Program (in Ghana) VfM Value for Money audit MC Minimum condition Performance-Based Fiscal Transfers for Urban Local Governments 18 Part I Results and Lessons from Two Decades of World Bank Financing Part I Key findings and lessons for Performance Grant Programs Performance-Based Fiscal Transfers for Urban Local Governments 19 Part I Results and Lessons from Two Decades of World Bank Financing 1. Introduction: Why Urban UPG programs have generally been acknowledged to have contributed to improving local government performance in several areas.7 These Performance Grants? include but are not limited to the following: • Supporting core administrative functions and compliance with basic 1.1. Context and motivation for this report statutory requirements Performance-based grants (PBGs) are a specific type of • Catalyzing public investment management and PFM by local intergovernmental fiscal transfer (IGFT) mechanism that provides governments (for example, high-quality planning processes, incentives to local and other subnational governments to improve compliance with procurement regulations, timely accounting, audit processes, outcomes, and responses) institutional performance and service delivery outcomes. Fiscal transfers from a higher level of government (for example, the • Enhancing local government transparency and accountability, including downward accountability (the interface among local central government) to local and other subnational governments are governments, citizens, and firms), upward accountability (incentives conditioned on their performance in predetermined areas.5 PBG for local governments to comply with national laws and regulations), programs are generally embedded within countries’ IGFT systems and and horizontal accountability (between local civil servants and support the decentralization objectives of national governments. elected officials) The World Bank has established an extensive portfolio of such programs • Improving local infrastructure service delivery performance and supporting governments across several regions. The Bank’s Global expanding the stock of municipal infrastructure Practice for Urban, Disaster Risk Management, Resilience, and Land • Spotlighting cross-cutting issues, such as gender, social inclusion, (GPURL) has led the way in supporting these programs, overseeing over poverty targeting, and the environment US$8 billion in Bank financing for them. Currently, eighteen such Bank- After more than two decades in which UPG programs have been in use, financed programs across four regions of the world are being overseen by this report is intended to take stock of the design and implementation GPURL. These tend to target urban local governments and are broadly experiences of such programs and their results in achieving intended referred to as urban performance grant (UPG) programs.6 Other Global outcomes. The objective is to document these experiences and Practices of the Bank, including the Governance Global Practice and the emerging findings on performance-based grant financing that supports Social Sustainability and Inclusion Global Practice, also support several institutional and service delivery improvements for local governments. PBG programs. The scope of this report is limited to those supported by The report identifies and analyzes the key achievements and challenges GPURL, but the lessons learned and recommended best practices are of GPURL’s UPG financing portfolio to draw key lessons pertaining applicable to the design and implementation of all PBG programs. to the design and the implementation of these programs and provide recommendations for future programs. 5 Also see the Glossary for a definition of PBGs (and other terms used in this report) and the various types of local and other subnational governments that are generally recipients of PBGs. 6 The terms PBG and UPG are generally interchangeable and are used interchangeably in this report, with the difference that UPGs tend primarily to target urban local governments, while PBGs cover any type of subnational or local governments (see also the Glossary for definitions). Some UPGs, however, include both urban and rural local governments as recipients, depending on the local institutional and program context. 7 This paragraph has been derived from UNCDF 2010. Performance-Based Fiscal Transfers for Urban Local Governments 20 Part I Results and Lessons from Two Decades of World Bank Financing 1.2. Objectives, methodology, and structure of this report Box 1.1. Scope of World Bank Financing Programs Overseen This report provides key metrics and high-level analysis of the UPG by GPURL portfolio of the World Bank, specifically such programs overseen by The World Bank’s Global Practice for Urban, Disaster Risk Management, Resilience, and Land (GPURL) finances three types of programs: urban development, disaster GPURL; presents in-depth case studies of selected UPG programs, risk management and resilience, and land administration and management. Each focusing on their achievement of results, the challenges they have category includes the following thematic areas: faced, and their implementation experiences; and identifies key lessons • Urban development: Includes urban infrastructure and service delivery; services and recommendations based on these lessons for the design and and housing for the poor; public transportation; urban planning; urban water and implementation of future UPG programs. sanitation; and municipal finance (or urban performance grant programs). • Disaster risk management (DRM) and resilience: Includes disaster The report includes a combination of quantitative and qualitative preparedness and response; disaster risk reduction; post-disaster recovery and analyses. The quantitative review provides overall metrics concerning reconstruction; disaster risk information and decision-support systems; and the size, growth, and performance of the UPG portfolio, comparing DRM governance it to the rest of the GPURL’s urban development portfolio (see Box • Land administration and management: Includes geospatial services and 1.1 for a description of financing programs overseen by GPURL). The land information systems; national land policy and administration; urban land administration and management; and geospatial information frameworks reviewed portfolio includes eighteen financing operations approved In this report, the “urban portfolio” is defined as comprising all programs in the by the World Bank Group Board of Executive Directors between fiscal urban development category approved between fiscal years 2013 and 2020. It years 2013 and 2020, which represent the second generation of UPG includes a total of 81 programs, a list of which was reviewed and endorsed by GPURL financing programs. The programs comprise over US$5.2 billion in management across all regions. approved financing, cover around 4,351 local governments, and benefit For more information, see, respectively, Urban Development (World Bank 2020c), Disaster Risk Management & Resilience (World Bank 2020a), and Land (World Bank 2020b). an estimated 226 million people (see Table 1.1 for a list of the financing programs reviewed).8 Chapter 2 of this report describes the evolution of the first and second generations of these programs. Notably, most of the selected cases (seven of the nine) were in the Africa region, where the UPG experiences have been largely concentrated and The quantitative review is supplemented by a qualitative analysis have had the longest histories. Each analysis was based on desk reviews comprising in-depth case studies of nine UPG financing programs of key project documents,9 complemented by virtual internal interviews across four regions (see Table 1.1). The selection of these programs was with respective World Bank task teams.10 based on three criteria: • Regional balance 9 Key project documents reviewed included the following: (1) project appraisal documents, which • Representativeness in terms of a standard typology of a UPG program set out the World Bank staff’s appraisal and assessment of the feasibility of and justification (that is, decentralization context, second generation operations, for each financing program; (2) implementation status and results reports, which are the Bank’s main tool for internal and external reporting on the implementation performance and targeting of multiple municipalities) prospective outcomes of its financing programs; (3) annual performance assessments (typically • Programs drawing upon accumulated experiences to generate externally conducted); (4) midterm review, a tool to assess the implementation performance at the midpoint of the program implementation cycle; and (5) implementation completion meaningful lessons reports, which are among the Bank’s main instruments of self-evaluation at the close of each of its financing programs. 10 Initially, a series of field visits to the case countries was planned for rigorous reviews, entailing detailed result verifications and interviews with local stakeholders. The scope of the overall 8 This analysis is based on data retreived from the World Bank’s internal operations database as analyses was limited, however, by the COVID-19 pandemic, which necessitated virtual of June 30, 2021. modalities. Performance-Based Fiscal Transfers for Urban Local Governments 21 Part I Results and Lessons from Two Decades of World Bank Financing Table 1.1. Overview of the Reviewed UPG Programs Amount of WB Target Financing Participating local Case study No. Country UPG program funding (US$ beneficiaries, instrument governments included millions) estimate (millions) 1 Bangladesh Municipal Governance and Services Project IPF 410 26 3.4 2 Egypt Upper Egypt Local Development PforR 500 2 7.8 Second Urban Local Government Development Program 3 PforR 380 44 ✓ Ethiopia (ULGDP II) 3.6 4 Urban Institutional and Infrastructure Program (UIIP) Hybrid 600 117 ✓ 5 Local Government Capacity Support Project (LGCSP) IPF 175 46 ✓ Ghana 9.6 6 Ghana Secondary Cities Support Program (GSCSP) PforR 100 25 ✓ West Bengal Institutional Strengthening of Gram 7 India PforR 210 3229 67 ✓ Panchayats Program (ISGPP II) 8 Kenya Kenya Devolution Support Program (KDSP) PforR 200 47 54 ✓ 9 Casablanca Municipal Support Program PforR 200 Morocco 106 15 10 Municipal Performance Program PforR 300 11 Pakistan Punjab Cities Program PforR 200 16 40 12 Senegal Municipal and Agglomerations Support Program PforR 110 142 7.5 13 Tanzania Urban Local Government Strengthening Program (ULGSP) PforR 255 18 3 ✓ Urban Development and Local Governance and AF 14 Tunisia PforR 300 and 130 264 7 ✓ (PDUGL) Uganda Support to Municipal Infrastructure Development 15 Uganda PforR 150 and 310 22 3.5 ✓ Program and AF (USMID) Results-Based National Urban Development Program— 16 Vietnam PforR 250 7 0.8 Northern Mountains 17 West Bank Local Governance and Services Improvement Program PforR 18 144 3.6 18 and Gaza GZ-Third Municipal Development Project IPF 76 Total 5,224 4,351 225.8 9 Performance-Based Fiscal Transfers for Urban Local Governments 22 Part I Results and Lessons from Two Decades of World Bank Financing This report is structured in two - Chapter 5 concludes by discussing several issues that will likely parts, as follows: affect the next generation of UPG programs. It reviews the sustainability of these programs within country systems and how it can be further improved. This is followed by a discussion Part I presents key findings and lessons for the design of how the next generation of UPG programs can more directly and implementation of UPG programs. provide incentives for service delivery outcomes and how they can - Chapter 1 sets the stage by providing the context and motivation promote local action with regard to climate change. for the report and introducing the concept of performance-based financing for local and other subnational governments. It also describes the report’s objectives and methodology. Part II includes case studies of nine UPG programs across - Chapter 2 provides more details on the substance and evolution seven countries: Ethiopia, Ghana, India-West Bengal, of UPG programs, describing the key features of UPGs financed Kenya, Tanzania, Tunisia, and Uganda. by the World Bank and how they work to improve institutional performance and service delivery at the local level, followed by a description of their evolution and growth over the past two decades. - Chapter 3 reviews the results achieved by UPG programs (and participating local governments) across key performance metrics and municipal institutional variables, including OSR improvements; PFM; transparency, accountability, and citizen engagement; operation and maintenance (O&M) of capital infrastructure investments; and human resource (HR) management and local staffing. It also discusses results achieved in the provision and improvement of infrastructure and service delivery for beneficiaries. The chapter also broadly compares the performance of UPG programs with other urban sector projects financed by the Bank. - Chapter 4 discusses lessons learned and good practices for UPG programs across three components key to the successful design and implementation of future programs: grant design, grant operationalization, and capacity-building support for local governments. Performance-Based Fiscal Transfers for Urban Local Governments 23 Part I Results and Lessons from Two Decades of World Bank Financing 2. Role and Evolution of Urban • The type of local government performance that is directly measured and assessed: That is, whether the areas of local government Performance Grants in Improving performance assessed by the program to determine grant eligibility are institutional and process indicators or, rather, indicators that Local Government Outcomes aim to measure service delivery outcomes, either directly or through proxy indicators. This chapter discusses the role of UPG programs in improving institutional performance and service delivery at the local level, focusing on programs financed by the World Bank over twenty years Figure 2.1. Typology of Performance-based Based Grant across several regions. A summary of key design and implementation Programs along Two Dimensions features typical of these programs and how they work is followed by a description of their evolution and growth over this period. Multi-sector usage and Sector-specific usage and Type of Performance to Access Funding Service delivery outputs/ service delivery focus service delivery focus 2.1. Key features of urban performance grants: How they Relatively rare, though some Recent UPG programs work UPG programs have proxy (focused on climate resilience, outcomes UPG programs are typically designed to utilize performance-based indicators for services delivery solid waste, metropolitan grants to provide incentive for improvements in the institutional outcomes (e.g., Value for management) + Results- Money audits) based financing programs performance and the infrastructure and service delivery of an identified in health & education. More range of urban local governments. In general, UPG programs have two common in advanced contexts objectives. First, they aim to improve the institutional performance of urban local governments, focusing on strengthening their capacity and organizational functionality and increasing their accountability in core Multi-sector usage and Sector-specific usage and Institutional functionality areas such as service delivery planning and execution, PFM, strategic institutional focus institutional focus development planning, and citizen engagement among others; and, Most Bank-financed UPG Some recent UPG programs programs are in this category. focused on institutional second, they expand the stock and coverage of municipal infrastructure LGs have discretion to invest in improvements in specific and services within the mandates of local governments, including various sectors in their remit sectors (e.g., climate roads and public transportation, water and sanitation, municipal waste resilience, solid waste management) management, drainage, and public spaces, among others. PBG programs can generally be categorized along the following two dimensions: Multi-sector Sector-specific • The eligible uses of grant funds for local governments: That is, whether grant funds are meant to be expended only for specific, predefined sectors (such as education, health, water and sanitation, Use of Grant Funds and so on), or whether the local governments have the flexibility and discretion to expend them across multiple sectors based on their own Source: Adapted and updated by authors from UNCDF 2010. priorities and expenditure assignments. Performance-Based Fiscal Transfers for Urban Local Governments 24 Part I Results and Lessons from Two Decades of World Bank Financing Most UPG programs reviewed in this report are located in the bottom- Figure 2.2. Stylized Structure of Urban Performance left quadrant of the typology shown in Figure 2.1. Local governments Grant Programs participating in such programs have a broad mandate for using grant funds across their wide sets of expenditure assignments but typically Central/Higher-level government restrict them to capital investments (such as infrastructure and services) rather than recurrent expenses. The performance assessments also generally focus mostly on institutional indicators. The participating local governments receive their grants from a central or regional agency, such as the Ministry of Finance, upon Demonstrate Demonstrate the achievement of predefined result targets. Figure 2.2 shows performance performance an illustrative design of a UPG program, in which the result areas Capacity Grants building mostly focus on institutional performances as “drivers of change” for efficient and effective service delivery and infrastructure expansion. Local Governments The targeted results of such programs, on which local government performance is to be measured, can be calibrated from measures of institutional performance to more advanced measures such as service delivery outcomes, depending on the existing institutional environment and program objectives. The performance results of local governments are generally measured through a robust, externally executed, and transparent annual performance assessment (APA) Demand Subproject Subproject Subproject Demand system, generally comprising a set of minimum conditions (MCs) and performance performance performance measures (PMs). The MCs ensure and safeguard local Citizens governments’ capacities to handle funds (and function as eligibility criteria for local governments to receive funding from the program), while the PMs are more advanced and reflect qualitative aspects of local governments’ performances. The findings of the APA, which documents the achievement of the performance metrics by participating local Source: Authors governments, are shared with the higher tiers of government responsible for managing the PBG program and form the basis for the allocation Upon receipt of grants, local governments have discretion over of grant funds to the local governments. The APA cycle is supposed to the design and implementation of grant-funded infrastructure be synchronized with the intergovernmental annual budget planning investments in line with their local development plans and priorities. cycle to ensure the allocations are reflected in the governments’ annual Most UPG programs provide multisectoral development grants that budgets on time (see Box 2.1 for more details). support a wide range of municipal infrastructure and services, including education and health-related facilities, roads, public spaces, drainage, and solid waste management (SWM). A few programs that are increasing Performance-Based Fiscal Transfers for Urban Local Governments 25 Part I Results and Lessons from Two Decades of World Bank Financing in number are sector- or theme-specific, focusing, for example, on UPG programs also mostly cover multiple cities, but in some cases SWM, climate change, and city competitiveness.11 Citizens can play an a single large city or metropolitan area may be targeted for specific important role in setting up investment priorities and monitoring the purposes—for example, to support local economic development.12 performances of local governments in the planning and execution of In some cases, urban as well as rural local governments are grant infrastructure investments (Figure 2.3). recipients, depending on program objectives and the specific political, administrative, and institutional context. Figure 2.3. Example of How UPGs Work and Align with Government Systems YES/NO PERFORMANCE Eligibility criteria / Spending on local Performance Measures Minimum conditions Local Govt Annual Budget government mandates (Examples) (examples) (Infrastructure & services) Fiduciary systems (PFM, Audits) Local own-source Procurement Planning revenue generation FINANCING & Budgeting Environment & Social Capital investment risk management planning & execution Staffing & HR Central management Government Accountability and Others citizen engagement World Bank Source: Authors 11 Recent examples at the World Bank include a US$105 million project supporting the Indian state of Kerala in strengthening urban SWM services; a US$120 million project on cities and 12 Examples include the Competitive and Livable City of Karachi Project, the Casablanca climate change in Mozambique; and a US$500 million program in Egypt promoting the local Municipal Support Program, and the Greater Kampala Metropolitan Area Urban Development business environment and development in a lagging region. Program. Performance-Based Fiscal Transfers for Urban Local Governments 26 Part I Results and Lessons from Two Decades of World Bank Financing A key feature of UPG programs is the capacity-building support that is provided to local governments to help them achieve the institutional Box 2.1. Typical Minimum Conditions and Performance Measures performance objectives for which they are being offered incentives. in UPG Programs Typically, the programs incorporate a set of capacity-building activities Minimum conditions (MCs)a are a set of basic requirements with which all participating at different levels of government, with local governments receiving local governments should comply to become eligible to receive grant funding from the program. The role of MCs is generally to ensure fiduciary comfort, and that funds will training, mentoring, and guidelines from higher tiers. Capacity-building be absorbed with due safeguards in place. Most UPG programs have a similar set of support is sometimes delivered to local governments through specific MCs, comprising a series of binary conditions (Yes/No indicators) without qualitative earmarked grants dedicated to their capacity-building needs, with a variations in performance. small set of conditions for gaining access to the funding (see chapter 3 Performance measures (PMs) represent a set of predefined targets local governments for more details on capacity-building support). The grants earmarked should achieve, the results of which determine the success of the UPG program. PMs, which cover various thematic areas, comprise more refined and qualitative indicators for capacity-building activities support a demand-driven approach that than MCs. The selection of the indicators varies across UPG programs, ranging from very helps local governments take full responsibility for and ownership of simple designs with complex ones with over 60 indicators. The results of PMs are also their performances. used as references for capacity building and overall monitoring and evaluation systems. The generic design of UPGs reflects the goals and impacts of the program in the institutional contexts in which they are introduced. Table B2.1.1. Illustrative Examples of MCs and PMs for Common Broadly speaking, UPG programs are focused on cross-cutting Thematic Areas institutional strengthening and reform because the local governments they target are characterized by deep institutional weaknesses, and Minimum Conditions Performance measures more effectively functioning institutions are needed before service- Agreement to Participation agreements signed with local delivery improvements can be made (in other words, a sequencing participate in the n/a governments for duration of imperative exists). The higher-level goal of these programs—improved program program service delivery—is pursued through an instrument based on a theory of Plans and budgets approved Planning and Quality of the development plan change that is cognizant of the low-capacity and governance-challenged budgeting by local council in timely and links to budget manner institutional environments in the countries in which the programs operate. If one were trying to improve the service delivery of a highly Timely preparation and Book of accounts, asset approval of annual budgets capable local government in a higher-income country through such a management up to date and with and end-of-year financial PFM & Auditing standard quality program, an explicit focus on institutional strengthening might not be statements & budgets Clean audit and follow-up on necessary, as the baseline performance might already be strong. In such Financial Audits without audit findings adverse opinions/disclaimers cases, the grant can focus on providing direct incentives for service delivery improvements, as PBGs in developed countries have often done. Capital budget Adherence to eligible Rate of execution of planned execution and expenditures in previous capital increase in OSR implementation fiscal year investment budget Value for Money (VfM) audit, Quality of capital sustainability. Completed n/a Investments infrastructure rated meeting quality & efficiency standards Performance-Based Fiscal Transfers for Urban Local Governments 27 Part I Results and Lessons from Two Decades of World Bank Financing In summary, UPGs generally have the following key design features: Minimum Conditions Performance measures • Performance assessment system: The program relies on a robust, Increased OSR collection year- externally executed, and transparent common APA system that Increase in OSR, but used in on-year which informs the grant allocation and funding to local governments OSR performance rare cases Improved OSR management in strict accordance with established conditions. systems • Measured results focus on improvements in institutional Participatory budgeting & investment prioritization performance. Local governments receive funds upon achievement Accountability, Sharing of information of pre-defined performance indicators. The targeted results of UPG through public disclosure Involvement of citizens in budget Transparency programs, on which local government performance is to be measured, planning, monitoring and reviews & Citizen Functional local government Engagement websites etc. Functional Grievance redress can be calibrated from measures of institutional performance to more mechanisms and their advanced measures such as service delivery outcomes, depending performance on the existing institutional environment and program objectives. Timely preparation & disclosure However, doing so also introduces challenges and complexities, as Presence of procurement of Procurement plans Public discussed in Chapter 5 of this report. plans and basic systems in Procurement Procurement performance and place contract management • Grant funding is generally used for capital investments (infrastructure and service delivery) within the mandates of local Development of Asset registers Asset / inventory, asset management governments, which are measured and reported on. Within limits, management plans participating local governments have discretion over the grant- n/a and O&M of infrastructure Improved O&M planning & funded capital investments in line with local plans and priorities. execution • Capacity-building support is provided, both to the central Environmental & government agency or ministry administering the program and to Basic E&S systems or staff Social (E&S) risk Performance of E&S systems capacity in place the grant recipient local governments, with delivery to the latter management sometimes through earmarked capacity-building grants and typically HR management Minimum staffing and Overall staff functionality, etc. closely linked with the results of the performance assessment system. and staffing positions in place Mainly environmental and Gender, local economic 2.2. Evolution of urban performance grant programs Cross-cutting development, disaster risk social safeguards and The UPG financing portfolio overseen by GPURL has grown areas management and resilience, local complaints handling system climate change action etc. substantially over the past two decades, reaching over US$8 billion in financing volume (Figure 2.4). The Bank-financed UPG programs Sources: Authors, based on review of various UPG programs; UNCDF 2010. over this period have also evolved substantially in design and scope, a. Also referred to as minimum criteria, minimum access conditions, and minimum mandatory conditions in various programs. and they generally constitute two broad generations of programs. The first generation began around 2000, with programs in East Africa (Tanzania and Uganda) and South Asia (Bangladesh). These programs often built upon and scaled up some of the pilot PBGs initiated by other development partners, such as the United Nations Capital Development Fund (UNCDF). In Uganda, for example, the concept of PBGs, which Performance-Based Fiscal Transfers for Urban Local Governments 28 Part I Results and Lessons from Two Decades of World Bank Financing was first piloted in five districts in 1997 with support from UNCDF, The second generation of UPG programs has seen them expanded was subsequently scaled up with Bank support under multiple UPG substantially, both in financing volume and geographical scope (Map programs, totaling US$600 million in cumulative Bank financing to date 2.1 and Figure 2.5). While UPGs, like general PBGs, can be implemented (UNCDF 2021). in any country or region around the globe,13 over two-thirds of those overseen by GPURL at the World Bank are concentrated in Africa and South Asia. Several World Bank–financed UPG programs have utilized Figure 2.4. Cumulative World Bank Financing for UPG Program-for-Results (PforR), a new financing instrument introduced by Programs under GPURL since 2000 the Bank in 2012 (see Box 2.2 for a discussion of Bank financing for UPGs 9 using various financing 9 instruments). In a number of countries, the first phase of the second-generation programs is now evolving into a second 8 phase, with the number 8 of local governments covered often expanded under follow-on programs and with refined approaches. Generally, both 7 7 generations of UPG programs can be broadly divided into initial and 6 6 with the initial phase in most countries followed by subsequent phases, Billions of US dollars Billions of US dollars scaling up and expansion to cover additional local governments. 5 5 Launch of PforR Launch of PforR 4 financing instrument Map 2.1. Geographical 4 financing Coverage instrument of World Bank-Financed UPGs Overseen by GPURL 3 3 2 2 West Bank & Gaza Morrocco Tunisia 1 1 Pakistan Bangladesh Egypt 0 Senegal 0 India Vietnam Ethiopia 2000 2010 2015 2020 2000 2010 2015 2020 Ghana Uganda Indonesia Kenya First generation Second generation First generation Tanzania Second generation First generation Second generation Source: Team analysis based on data from World Bank’s internal Operations Portal standard report and Business Warehouse database on June 30, 2021. Note: The scope of the Urban Performance Grant programs included in this figure covers a total of 33 financing operations approved by the World Bank Group Board of Executive Directors and overseen by the Global Practice of Urban, Disaster Risk Management, Resilience and Land Source: Authors between fiscal years 2000 and 2020. It includes the first and second generations of UPG financing programs comprising over US$8 billion in approved financing. 13 Overall, performance-based grants have expanded across all regions. Such programs began in Africa and subsequently were implemented in Asia (for example, in Bangladesh, India, Lao, Mongolia, and Nepal), Europe (such as in Kosovo and Serbia), and the Pacific Islands (Solomon Islands and Tuvalu). See more examples in UNCDF 2021. Performance-Based Fiscal Transfers for Urban Local Governments 29 Part I Results and Lessons from Two Decades of World Bank Financing Figure 2.5. Evolution of UPG Programs in Each Country REGION COUNTRY 2000~ 2015 2010 2015 2020 2022 Uganda (UGA) UGA 1 & 2 UGA 3 UGA 4 and AF Tanzania (TZA) TZA 1 TZA 2 Senegal (SEN) SEN 1 SEN 2 Sub-Saharan ETH 1 ETH 2 Ethiopia (ETH) Africa ETH 3 Ghana (GHA) GHA 1 GHA 2 KEN 1 Kenya (KEN) KEN 2 Bangladesh BGD 1 BGD 4 (BGD) BGD 2 BGD 3 BGD 5 South Asia IND 1 IND 3 India (IND) IND 2 Pakistan (PAK) PAK 1 PAK 2 West Bank and GAZ 1 GAZ 3 Gaza (GAZ) GAZ 2 Middle-East Tunisia (TUN) TUN 1 & North Morocco "MAR 1" Africa (MAR) "MAR 2" Egypt (EGY) EGY 1 East Asia Indonesia (IDN) IDN 1 Pacific Vietnam (VNM) VNM 1 Source: Team analysis based on data from World Bank’s internal Operations Portal standard report and Business Warehouse database on June 30, 2021. Note: The orange bars represent the first generation of UPG programs, with blue bars for the second generation. Each urban performance grant program is labeled with a respective country code and the ordinal number of a given program in the country (for example, the first UPG program in Uganda is shown as UGA 1). Performance-Based Fiscal Transfers for Urban Local Governments 30 Part I Results and Lessons from Two Decades of World Bank Financing Box 2.2. World Bank Financing for UPG Projects and Programs Using Various Financing Instruments In 2012, a new World Bank financing instrument, Performance for Results (PforR), was launched to support directly part of an existing government program that targets specific results. Funds under PforR are disbursed upon the achievement of disbursement-linked indicators (DLIs) that reflect the results of government programs. One of the benefits of the PforR modality is that it encourages governments to take ownership of the design and implementation of the investments. Previously, most UPG programs used other Bank investment financing instruments, such as Investment Project Financing (IPF). The World Bank has increasingly used its PforR instrument to support UPG programs because it naturally serves the objective of UPGs to support the improvement of intergovernmental fiscal transfer systems in a country, as well as the improvement of infrastructure and service delivery upon achievement of the predefined results. Using this instrument, the Bank disburses funds to the central government based on the verified performance of local governments, which, in turn, receive funds from central government transfers. The first PforR-type UPG was introduced in Tanzania, followed by Uganda. As Figure B2.3.1 shows, as of FY22, GPURL ranked as the fifth largest PforR user among all fifteen global practices (sectors) at the World Bank. Figure B2.3.1. Use of PforR by World Bank Global Practices Education 17.5% Energy and extractives 12.2% Governance 11.8% Health, nutrition and population 0.9.9% Urban, resilience and land 9.1% Water 9.1% Social protection & jobs 7.2% Agriculture and food 6.6% Finance, competitiveness, and innovation 6.5% Others 10.1% 0 0.05 0.1 0.15 0.2 PforR share of the World Bank operation portfolio per sector in financing volumn (%) Source: Team’s analysis based on the data from the World Bank’s internal Operations Portal standard report database as of June 30, 2021. Performance-Based Fiscal Transfers for Urban Local Governments 31 Part I Results and Lessons from Two Decades of World Bank Financing 3. Key Findings from Review of Figure 3.1. Share of UPGs Out of Urban Portfolio World Bank Financing for Urban Performance Grant Programs Chapter 3 reviews how UPG programs perform relative to other urban sector projects overseen by GPURL, referred to as the “urban portfolio” in this report. The chapter also provides the results attained by UPG 41% 56% programs across three key performance areas—overall performance of local governments, improvements in key urban and municipal institutional variables, and infrastructure, service delivery, and beneficiaries—based on findings from in-depth case studies of nine UPG programs. Share of UPGs of urban Share of UPGs of urban 3.1. Performance of UPG portfolio relative to other urban portfolio in financing volume portfolio in IDA financing sector projects US$5.2 billion invested in UPGs US$3.5 billion invested in UPGs between FY13 and FY20 out of US$6.3 billion between FY13 UPG programs represent a sizeable share of the urban portfolio. They and FY20 account for 41 percent of the portfolio in financing volume, amounting to US$5.2 billion in total (see Figure 3.1). The UPG programs also take 56 percent of total International Development Association (IDA) Source: Data retrieved from the World Bank’s internal Operations Portal standard report and financing14 (US$3.5 billion). This significant share of UPG programs Business Warehouse database on June 30, 2021. Projects overseen by Bank’s GPURL. shows a strong demand from IDA client countries for performance- Note: Scope includes Urban Operations and second-generation UPGs approved between FY2013 based grants that focus on addressing urban development challenges and 2020, with one exception (Ghana LGCSP) that was approved before FY2013 but was an important part of the second-generation UPGs. and reiterates the important role of UPGs in the overall success of the urban portfolio. The Bank invests substantially more in UPG programs, on average, implementation support for—with 24 percent higher total preparation than in non-UPG urban-sector financing projects, and they are costs and 34 percent higher annual implementation support costs, more cost effective. The average Bank financing for a UPG program is on average—they are much more cost effective in per-million-dollar US$275 million, which is 2.3 times higher than for non-UPG projects. financing terms (see Figure 3.2). The average project preparation cost for While UPG programs tend to cost more to prepare and provide UPG programs per million-dollar financing is about $1,800, compared to over $3,000 for non-UPG urban-sector projects, while the average 14 The IDA is the part of the World Bank Group that provides concessional funds, including zero- annual implementation support cost is also substantially lower in per- to low-interest loans (called “credits”) and grants, to the world’s poorest countries for programs million-dollar financing terms. that boost economic growth, reduce inequalities, and improve people’s living conditions. See https://ida.worldbank.org/ for more details. Performance-Based Fiscal Transfers for Urban Local Governments 32 Part I Results and Lessons from Two Decades of World Bank Financing Figure 3.2. Cost Effectiveness of UPG Programs Figure 3.3. Disbursement Trends of UPG and non-UPG Programs Over a Standardized Project Timeline US$3,238 100 90 US$ 2,647 80 UPG 70 Disbursement rate (%) US$1,788 Non-UPG 60 US$1,090 50 US$911 US$650 40 30 20 Average project preparation cost Average annual project implementation 10 per US$1 million in Bank financing support cost per US$1 million in Bank financing 0 Approval E ectiveness Closing UPG Non-UPG Urban portfolio Project timeline Source: Data were retrieved from World Bank’s internal Operations Portal standard report and Source: Data were retrieved from World Bank’s internal Operations Portal standard report and Business Warehouse database on June 30, 2021. Projects overseen by Bank’s GPURL. Business Warehouse database on June 30, 2021. Note: The X-axis refers to standardized values (0–1) of the project timeline: 0 = just approved projects; 0.5 around midterm review; and 1 for closed projects. The Y-axis refers to the percentage of program costs utilized out of allocated funding. The red dots represent UPG programs, with blue dots for non-UPG programs out of GPURL’s Urban Portfolio. Moreover, UPG programs disburse Bank funds faster than non-UPG On a scale of 1 to 5, with 5 being highest, UPG programs have an average urban-sector projects over their lifetimes. While one may expect the rating of 4.53 on their achievement of development objectives, which is disbursement rates of UPG programs to be lowered by the challenging slightly higher than the average rating of 4.37 for non-UPG urban -sector conditions of their designs—such as the need for timely and transparent projects in the comparative period. A similar pattern was observed assessment of all participating local governments and a demanding list of in average reported ratings for implementation progress, with UPG performance metrics and disbursement-linked indicators (DLIs)—they programs scoring 4.29, compared to 3.96 for the non-UPG projects.15 disburse Bank funds well relative to non-UPG urban-sector projects over the course of their timelines (see Figure 3.3). 15 Ratings are for active or ongoing programs or projects and are based on publicly available UPG programs also perform as well as non-UPG urban-sector projects implementation status and results reports for each, prepared by Bank teams and approved by in achieving program development objectives and implementation Bank management. The rating scale of 1 to 5 is defined as follows: satisfactory = 5; moderately satisfactory = 4; moderately unsatisfactory = 3; unsatisfactory = 2; NA = 1. Data were retrieved progress, as reported in biannual program ratings by Bank task teams. from the World Bank’s internal Operations Portal standard report and the Business Warehouse database on June 30, 2021. Performance-Based Fiscal Transfers for Urban Local Governments 33 Part I Results and Lessons from Two Decades of World Bank Financing 3.2. Results of nine case-study UPG programs across key Table 3.1. Increase in Minimum Condition Compliance Rate performance areas across UPG Programs This section discusses results achieved by UPG programs in nine % of urban local governments case-study countries. These cases show overall performance with MC compliance Years taken % increase improvements achieved by UPGs in the following results areas: Most recent for increase Baseline result • Overall performance of local governments, in terms of how those participating fared in complying with UPG program MCs, and their Ethiopia (ULGDP II) 100 100 0 4 improvements in institutional performance over time Ethiopia (UIIDP) 100 100 0 1 • Improvements in key urban and municipal institutional Ghana (GSCSP) 100 100 0 1 variables, including OSR, operation and maintenance (O&M) of Ghana (LGCSP) 65 89 24 3 capital investments, PFM, transparency, accountability, and citizen India-West Bengal 52 75 23 3 engagement, and human resource (HR) management (IGSPP-II) • Infrastructure, service delivery, and beneficiaries, covering Kenya (KDSP) 28 81 53 3 physical infrastructure improved, number of beneficiaries served, Tanzania (ULGSP) 83 100 17 5 and jobs created Tunisia (PDUGL) 89 97 8 3 • Improvements to environmental and social (E&S) systems Uganda (USMID) 100 100 0 4 3.2.1 Overall performance of urban local governments Uganda (USMID 78 100 22 2 AF)a Most local governments participating in the UPG programs reviewed Source: Team analysis based on relevant program documents and data provided by World Bank successfully complied with the eligibility criteria (minimum task teams for respective UPG programs. conditions, or MCs) over the lifetimes of their respective programs. a. USMID AF is shown separately from USMID, as their performance measurement designs were In almost half the programs, all the participating local governments very different. consistently met MCs throughout the entire program periods, while A prominent example was Ethiopia, where a rich learning experience the rest started with lower rates of compliance that substantially about UPG systems took place at all levels of government. The central increased over time (Table 3.1). The former included UPG programs in government successfully executed and rolled out performance grant Ethiopia (ULGDP II and UIIDP), Ghana (GSCSP), and Uganda (USMID). financing across the two UPG programs in the country. Several regional The experiences of these countries in UPG systems and preparatory governments adopted and tested a new performance measurement activities before the launch of their programs (such as capacity-building model that was developed under the first UPG program in the country, activities and awareness building among local governments) sufficiently and local governments participating in the first and second UPG prepared the local governments to comply with the MCs and gain access programs actively shared lessons learned with nonparticipating local successfully to UPG funding. In the cases where they could not comply governments. In addition to attaining a high level of understanding of in the beginning, their response to the demands of the MCs improved UPG systems, 73 local governments that had newly joined the third rapidly. UPG program (UIIDP)16 in the country received technical assistance 16 The third UPG program (UIIDP) scaled up its geographical coverage from 44 local governments under the previous UPG (ULGDP II) to 117 local governments. Performance-Based Fiscal Transfers for Urban Local Governments 34 Part I Results and Lessons from Two Decades of World Bank Financing and awareness-raising support before the program launch, enabling low, it increased rapidly over the implementation period. This indicates them to comply with the MCs from the outset. that even noncomplying local governments generally take only a few years to come up to speed with all minimum requirements and improve The remaining UPG programs showed annual increases in MC their systems to ensure that they do. compliance rates of approximately 9 percent, on average. A common pattern observed was that more than 80 percent of Thus, MCs seem to be a strong instrument for providing incentive to participating local governments managed to comply with the improve core systems. It is important, however, to ensure they are MCs after three years of implementation. Even if a UPG started not too easy to achieve (which could explain 100 percent compliance with lower MC compliance rates, the local governments tended to at baseline), and that they are properly selected within the range of respond quickly to financial incentives. Kenya’s first UPG program, commonly used indicators (see Figure 3.4). for example—the Kenya Devolution Support—introduced two sets of MCs for different funding windows, one for capacity-building support grants with fewer access conditions and the other for development grants with more demanding conditions. At the beginning, only 28 percent of the participating local governments met the MCs for the Figure 3.4. Common Types of Minimum Conditions across the development grants. The compliance rate dramatically increased, UPG Programs Reviewed however, to 81 percent within three years. This improvement Timely approval of plans/budgets indicates that the financial incentives in the MCs were adequately Financial statements produced on time designed, and, complemented by targeted capacity-building support, Acceptable audit reports they produced an effective response. Figure 3.4 lists common MC Adherence with eligible expenditures indicators across the reviewed UPG programs. Procurement plans/systems HR minimum sta ng and positions These trends merit a few observations that are relevant for the design E&S safeguards management of future programs. First, while all participating local governments Participation/grant agreement complied fully with the MCs right from the start of some programs Complaints handling system (that is, 100 percent compliance at baseline for Ethiopia and Ghana), Transparency and accountability these were programs or contexts that had strong previous experience Program management with UPG systems. They also had strong preparatory activities, Co-funding requirements managed by the central government, prior to the launch of the Compliance with other national conditions programs and the assessments. Through them the local governments Progress and financing reporting OSR increase were familiarized with program requirements, increasing their chances of achieving the MCs. These experiences indicate that 0 1 2 3 4 5 6 7 8 9 conducting such preparatory activities is an important way to prepare Number of UPG programs local governments for program implementation, familiarize them Note: Shows the list of MCs for UPGs. There are often separate MCs for capacity-building grants with the program system, and mobilize the needed fiduciary and under UPG programs, as in Ghana, Kenya, and Uganda, which are not included in the graph. These MCs typically include a requirement for preparing a capacity-building plan and adherence other support. Second, even in cases where baseline compliance was with eligible expenditure categories. Performance-Based Fiscal Transfers for Urban Local Governments 35 Part I Results and Lessons from Two Decades of World Bank Financing The overall performance of participating local governments in areas In addition to indicators used for the APAs, Bank-financed UPGs targeted for institutional strengthening also significantly improved using the PforR financing instrument included disbursement-linked across all UPG programs, as measured in annual performance indicators (DLIs) to provide incentive for higher tiers of government assessments (APAs). In most cases, average performance measure (PM) to support local governments in institutional strengthening. scores for local governments started from around 60–70 percent and Here, the disbursement of Bank financing was tied to the successful reached more than 80–90 percent over the reviewed program periods, accomplishment of actions by higher tiers of government, as well as to increasing by an average of 49 percent (Figure 3.5). It is also noteworthy institutional improvements by participating local governments. See Box that the UPG programs provided targeted support to weaker local 3.1 for a discussion of how DLIs have been used to provide incentive for governments so they could catch up with the top-performing ones. these actions and results across different levels of government. In Tanzania, for example, the standard deviation of PM scores among participating local governments decreased by 40 percent over the program period. In some cases, the performance trend was hard to analyze if substantial changes had been made to the PM indicators during implementation; typically, the bar was raised over time in some places. Figure 3.5. Increase in Average Score of APA Score over Program Period for All Local Governments in Program Ethiopia (ULGDP II) Ghana (LGCSPI) Uganda (USMID) Tanzania (ULGSP) Tunisia (PDUGL) Kenya (KDSP) 40 50 60 70 80 90 100 APA score (0-100 points) Baseline Recent score Source: Team analysis based on relevant program documents and data provided by World Bank task teams for respective UPG programs. Note: Of nine reviewed UPG programs, three are not included because of data unavailability (i.e., it was too early to measure the improvement or too hard to track the trend due to changes to the measurement system). Performance-Based Fiscal Transfers for Urban Local Governments 36 Part I Results and Lessons from Two Decades of World Bank Financing Box 3.1. Use of DLIs to Provide Incentive for Institutional Actions in UPG Programs Figure B3.1.1. Percentage Achievement of Disbursement-Linked Indicators in Various UPG programs Bank-financed UPG programs generally make use of six to ten disbursement- linked indicators (DLIs) related to the key performance areas of the APA and aligned with the programs’ development objectives. These are indicators that 100 100 100 have, in past experience, proved challenging for urban local governments and/ 95 or higher tiers of government to deliver on and, therefore, need a boost in terms 87.5 82.5 of incentive. DLIs among the reviewed UPGs commonly included requiring compliance with MCs; the achievement of annual PM targets; and the delivery 63 of capacity-building support. Other key results areas incorporated into DLIs include timely audits; OSR collection; HR management; accountability; and timely execution of infrastructure and/or VfM results. In some cases, DLIs are used to resolve specific program management challenges, such as delays in APAs and/or audit reports. Three out of seven reviewed UPGs using the PforR financing instrument included a DLI on the timely completion of the APA by the government, even though delays persisted in this, at times due to strong external factors outside the scope of the program or its coordinating Ethiopia Ethiopia Ghana India Kenya Tunisia Uganda agency. Some programs, such as in Ethiopia (ULGDP II and UIIDP), had DLIs (UIIDP) (ULGSP II) (GSCSP) (ISGPP II) (KDSP) (PDUGL (USMID) and AF) formulated around defined capacity-building support to be provided by a higher tier of government (regional or central) and/or for various types of Achievement of program DLIs (%) institutional systems for participating local governments. The Bank-financed UPG programs reviewed that used the PforR financing instrument achieved Source: Team analysis based on latest data provided by World Bank task teams for respective UPG an average of 90 percent of their DLIs, ranging from 63 to 100 percent (see programs in FY21. Figure B3.1.1). Note: This is the share of DLIs achieved against targets out of a total number of DLIs under each UPG program, according to the latest results available. Performance-Based Fiscal Transfers for Urban Local Governments 37 Part I Results and Lessons from Two Decades of World Bank Financing 3.2.2.Key urban and municipal institutional variables Figure 3.6. Percentage Increase in Own-Source Revenue over UPG programs have helped enhance urban systems, mainly by Program Period improving OSR performance, PFM, transparency, accountability and citizen engagement, operation and maintenance (O&M) of capital 163 160 160 infrastructure investments, and human resource (HR) management 148 and local staffing (see part II for detailed results of each UPG program). 140 140 Key enablers of such improvements have been a combination of strong 120 120 Increase in OSR (%) Increase in OCR (%) financial incentives and targeted capacity-building support, both before 100 100 and during the implementation. 83 80 80 OSR significantly increased in the seven out of the nine reviewed 67 66 60 60 UPG cases for which these data were available17. Six cases showed an 48 48 absolute increase in OSR collection, including Ethiopia (ULGDP II), 40 40 Ghana (LGCSP), India-West Bengal (IGSPP II), Kenya (KDSP), Tanzania 20 20 (ULGSP), and Uganda (USMID) (Figure 3.6). In the seventh case (Ethiopia, 0 0 UIIDP), the share of participating local governments with 10 percent Ethiopia Ghana Uganda Tanzania Tunisia Kenya Ethio annual increases in OSR rose from 77 percent in the first year to 85 (ULGDP II) (LGCSP) (USMID) (ULGSP) (PDUGL) (KDSP) (ULGD percent in the second. The positive impact of urban performance grants on OSR regeneration has not been limited to the reviewed programs but Source: Team analysis based on relevant program documents and data provided by World Bank has applied to other UPG cases, as well, as shown by recent empirical task teams for respective UPG programs. research on a UPG program in Mozambique (see Box 3.2). These are Note: Percentage increase in OSR in nominal figures in local currency. Two UPG programs in Ghana (GSCSP) and Tunisia (PDUGL) were not included because of data unavailability. remarkable results, seen in the context of various challenges in most of these countries with OSR collection systems, risks of crowding out of OSR improvement when fiscal transfers are increased, and previous OSR trends. In Uganda, for example, OSRs were decreasing or stagnant in many years prior to the UPG program, and increasing fiscal transfers In the reviewed programs, the UPGs contributed to improving O&M was among the challenges mentioned.18 of capital investments in urban infrastructure and services. In all but one program (in India-West Bengal), where O&M was not monitored during implementation, performance indicators related to O&M showed 17 It should be noted that an increase in fiscal transfers due to UPGs can at times create a progress. The share of local governments with the planned budget crowding out effect on OSR in the same period, as pointed out in the literature on municipal finance. In this case, however, OSRs have increased in absolute terms in the same period as allocated to O&M, for instance, increased by 94, 79, 45, and 35 percentage fiscal transfers via UPGs. A counterfactual analysis has not been done on this subject for this points, respectively, in Tanzania (ULGSP), Uganda (USMID), Ethiopia report. 18 See, for instance, Mahler (2005), who argues that the substantial increase in fiscal transfers (UIIDP), and Tunisia (PDUGL). In Ghana, the average PM score on asset undermined local government incentives to collect OSR. Researchers from Nigeria, such as management plans and strategies increased by 15 percent under the Kayode (2020), confirm this, although others—for example, Masaki (2018)—show it depends on the context, and that this may not always be the case, depending on the design of the fiscal first UPG in the country (LGCSP) and specific urban infrastructure and transfers. Performance-Based Fiscal Transfers for Urban Local Governments 38 Part I Results and Lessons from Two Decades of World Bank Financing Box 3.2. Impact of UPG on Own-Source Revenue Collection in Mozambique Figure B3.2.1. Impacts of UPGs on ORS Collection in Mozambique A recent study by Erman, Uehara, and Beaudet (2021) evaluated the impact on local revenue collection in Mozambique of an urban performance grant 30 financed under the Cities and Climate Change program. Between 2013 and Share of own revenue in total revenue (%) 2018, the program provided half of the country’s municipalities (that is, cities and towns) with annual transfers of over US$40,000 and technical assistance in such areas such as fiscal management, urban development, and investment 25.9 planning. 25 23.9 According to the study, the UPGs provided by the program were successful in 23.4 providing adequate incentives to municipalities and increasing local revenue 22.6 collection. Local revenue grew by 114 percent, and its share in total revenue 21.3 increased from 19 percent in 2013 to 24 percent in 2018, which was more than 20.1 20 in nonparticipating municipalities during the same period (see Figure B3.2.1). 18.8 19.4 The positive effect of the UPG was time lagged, which indicated the important 18.2 role of UPGs in building enabling environments for revenue collection, leading 17.4 17.5 to a gradual increase in the revenue in the following years. 15.4 15 Another important finding was the significant impact of UPGs on revenue collection when combined with technical assistance, implying that capacity- 2013 2014 2015 2016 2017 2018 building support and technical assistance are key to the success of a Received the grant Did not receive the grant performance-based grant program, especially in a low-capacity country. When combined with technical assistance, local revenue increased by 10–11 meticais (Mozambique currency) per capita in cities and 24–60 meticais per capita in towns for every grant of 10 meticais per capita. The larger effect of the UPGs on local revenue collection in towns than in cities may have been due to the stronger incentive provided by the grants to the towns, where they Source: Erman, Uehara, and Beaudet 2021. constituted a larger share of the total budgets than in the cities. Performance-Based Fiscal Transfers for Urban Local Governments 39 Part I Results and Lessons from Two Decades of World Bank Financing service maintenance indicators also showed improvement under the periods. The transparency and accountability indicators included subsequent UPG program (GSCSP)—for example, the pedestrian network public disclosing, communication framework and engagement plans in maintenance score increased by 48 percent during the program period. place, and budget forums held. PFM was enhanced substantially in all of the nine UPG programs. HR management and local staffing improved in seven out of the PFM was a key performance measurement area across most of the nine UPGs. HR management was often measured as an MC. Examples programs. In Uganda (USMID), Tanzania (ULGSP), Ghana (LGCSP), were in Ethiopia (both ULGDP II and UIIDP) and Uganda (USMID), and Ethiopia (ULGDP II), the APA scores on PFM increased by 67, where MCs on HR management were consistently met, although some 50, 47, and 46 percent, respectively, during the program periods. The of these measures were demanding, such as core staff in place with indicators measured included clean audit statements, timely reporting, required skills. Where HR was measured among key PM indicators, the commitment controls, cash flow, and internal control, among others. UPGs mostly showed improvements in preparing and implementing Among other examples were the following: organizational structures and staffing plans, job descriptions, • Ethiopia: The number of local governments with clean audits specifications and competency framework, staff appraisals, and increased from 25 to 56 out of the 117 participating in UIIDP during performance management, among others. The average performance the first two years of its implementation. All of the participating score on HR in Kenya (KDSP), for example, increased by 17 percent over local governments conducted timely audits, which was a major the program period. Among the common challenges reported related to achievement, considering previous challenges in this area. HR management were delays in filling in core staffing positions, mostly • Ghana: The share of local governments submitting financial accounts associated with changes in the governance system (for example, delays in a timely manner for audits increased from 13 percent to 100 percent in the adoption of new decrees in Tunisia); inadequate incentives to under LGCSP. central or provincial governments; and frequent transfers of staff. • India-West Bengal: The share of local governments without an In sum, UPG programs generally tended to provide incentive for adverse or disclaimed audit was more than 98 percent throughout improvements to core municipal and institutional performance the program. systems by measurably improving OSR generation, O&M of capital • Kenya: The APA score on monthly financial reporting improved by infrastructure investments, PFM, transparency, accountability and 13 percent, and performance on the internal audit trended upward citizen engagement, and HR management and local staffing. These by 26 percent. improvements tended to continue (for example, in Ethiopia, India- Transparency, accountability, and citizen engagement also improved West Bengal, and Uganda) when the programs were expanded and with in all nine UPG programs. UPGs reportedly contributed to an increase subsequent phases and rollouts. in the use of citizen charters and public disclosing, as well as to the development of innovative information-sharing tools. In Kenya (KDSP), the number of local governments that incorporated citizen inputs into budgeting increased by 52 percent through the public disclosure of previously unpublished documents. In Ethiopia (ULGDP II), Uganda (USMID), and Tanzania (ULGSP), transparency and accountability were monitored as part of key performance areas, and the resulting scores increased by 82, 56, and 25 percent, respectively, over the project Performance-Based Fiscal Transfers for Urban Local Governments 40 Part I Results and Lessons from Two Decades of World Bank Financing 3.2.3. Infrastructure and service delivery performance and Figure 3.7. Annual Execution Rate of Planned Infrastructure beneficiaries Investments at the Local Level Investments in urban infrastructure were well executed across the UPGs reviewed. The average annual execution rate of planned infrastructure investments was 88 percent, and it steadily improved 100 over program periods across all UPG programs (Figure 3.7). Most UPG expenditures were also effectively used for infrastructure investments.19 80 Evidence from Uganda also indicated that the number of participating 60 local governments that completed investment projects financed by UPGs gradually rose over the life of the program, from 42 in 2017/18 40 to 125 (out of around 150) in 2021/22, showing a marked increase in the ability of participating local governments to spend UPG funds in a timely 20 manner and pointing to gradual improvement in public investment management. The share of infrastructure investment costs compared 0 to total program expenditures was 83 percent on average, ranging from Ethiopia Ethiopia Ghana India Kenya Tanzania Tunisia Uganda (UIIDP) (ULGDP II) (LGCSP) (ISGPP II) (KDSP) (ULGSP) (PDUGL (USMID) 64 percent (in Kenya’s KDSP, which was the first UPG program in the and AF) country with a more comprehensive focus on capacity building) to 95 percent (in Ethiopia’s ULGDP II). Source: Team analysis based on relevant program documents and data provided by World Bank task teams for respective UPG programs. The UPG programs reviewed financed the rehabilitation or Note: The Y-axis represents the average share of executed capital infrastructure subprojects out of development of substantial capital infrastructure investments and planned infrastructure subprojects (in number) per annual local infrastructure development plan over program periods. One UPG in Ghana (GSCSP) was not included due to data unavailability. assets for improved service delivery for citizens and businesses. The improvements included, but were not limited to, the following: • Mobility and transportation: 7,000 km of roads; 160 km of bicycle or pedestrian ways; 28 transit facilities (such as bus terminals) The improved access to infrastructure and services tended to lead • Security and public safety: 8,700 streetlights; 30 security facilities to broader impacts, including improved mobility, resilience, and • Public or green spaces: 350 ha of parks; 90 local markets; 15 social public health. In some cases, for example, improved hygiene through facilities (such as community centers) new drainage and roads was found to reduce waterborne diseases • Solid waste management (SWM) and drainage: 290 waste collection and the number of disease-carrying mosquitoes. The construction of points; over 1,270 km of drainage; 3 dumpsites; 1,300 trashcans urban roads was also associated with increases in land values in areas • Education and health facilities: 2,000 classrooms; 28 community surrounding investment sites, as seen in Tanzania. health planning and service compounds; 14 health clinics; and more, especially after COVID-19 19 Note that use of funds for eligible investments was often an MC, as well. Performance-Based Fiscal Transfers for Urban Local Governments 41 Part I Results and Lessons from Two Decades of World Bank Financing Moreover, these UPG-funded infrastructure projects and services contributed to the creation of an estimated 1 million (mostly Table 3.2. Employment Opportunities Created via Labor- temporary) employment opportunities, primarily on labor-intensive Intensive Public Works in Selected UPG Programs capital infrastructure works (Table 3.2). While direct and/or indirect Estimated employment opportunities created across UPG programs were not person-years Projects Notes systematically tracked, some programs tracked the number of people of employment employed (in person-years of work), mainly on labor-intensive works, added such as road construction. The program led to a high level of job creation, in the range of about 160,000 jobs per year, The nine reviewed UPG programs collectively reached almost 150 Ethiopia (ULGDP II) 651,583 through such labor-intensive works as million target beneficiaries. Of these, over 80 percent benefited from cobblestone roads. two programs that targeted all local governments in a state or a country; The program led to a high level of job creation, 67 million were under India-West Bengal (ISGPP II) and 54 million under Ethiopia (UIIDP) 306,184 especially for unemployed youth. Kenya (KDSP). Other UPG programs targeted specific geographical At least 52,402 jobs were created during areas and served fewer beneficiaries, ranging from 3 million in Tanzania construction and through small businesses Tanzania (ULGSP) 52,402 to 9.6 million in Ghana. The reviewed UPGs benefited over 4,300 local along upgraded roads, markets, and bus terminals financed under the program. governments, accounting for 150 million people. In addition to indirect jobs created, 11,778 Value for Money (VfM) audits conducted for selected programs persons (79 percent males) were directly found UPG-funded capital infrastructure investments to be cost Ghana (LGCSP) 11,778 engaged for the provision of the subprojects, effective. In Uganda, the audits showed that 91 percent of infrastructure and approximately US$23,804,360 was paid to workers as wages. investment projects in the first generation of UPG programs were of high quality, generally functional, and being utilized as intended, Source: Data provided by World Bank task teams for respective UPG programs. pointing to improved efficiency and quality of capital expenditure. Note: Not intended as an exhaustive list of employment opportunities created through the The central government had expanded the use of the VfM audit tool reviewed programs. outside the program as well. Ongoing investments under the second- generation program (USMID) recorded higher VfM scores and were more cost efficient than similar non-UPG funded investments across Thus, despite some limitations, the UPG programs reviewed generally the country (Box 3.3). In India-West Bengal (IGSPP II), 98 percent of succeeded in improving institutional performance and infrastructure road construction projects were found to be satisfactory. In general, and service delivery outcomes in targeted local governments. Most UPG program funds were used efficiently, and the quality of the of the participating local governments successfully complied with infrastructure was satisfactory. This was supported in some cases by the minimum conditions for getting access to grants and improving use of VfM results in the APA process to enhance local governments’ institutional performances, as evidenced in the APA results and incentives to deliver these outputs. various reviews. The programs also had positive impacts on improving urban institutional systems, mainly by improving OSR generation, HR management, O&M, PFM, and accountability, and they disbursed funds Performance-Based Fiscal Transfers for Urban Local Governments 42 Part I Results and Lessons from Two Decades of World Bank Financing effectively relative to non-UPG programs. Key challenges commonly Box 3.3. Value for Money Audit in Uganda observed among the UPG programs were delays in APA processes A Value for Money (VfM) audit is a specific type of audit to measure the quality and and procurement and, in some places, timely deployment of sufficient actual execution (as compared to targets) of investments in urban infrastructure capacity-building support and technical assistance. and services. An increasing focus on the quality of infrastructure investments led to factoring the results of VfMs into APA scores and/or disbursement-linked indicators 3.2.4. Improving environmental and social standards and (DLIs). In Uganda (USMID), overall results of the VfMs were incorporated into a systems at the local level specific DLI on the actual delivery of infrastructure against targets; this measure The rigorous APA system seems to be an effective tool to cover cross- showed a stable increase in the VfM score, even during the expansion from 14 to 18 urban local governments. cutting issues, especially—based on the review of UPG programs for this These VfMs were designed to measure three “Es”: report—with regard to improving environmental and social standards • Economy: The unit cost of delivery of the infrastructure, as compared to works of and systems among participating local governments and reducing E&S similar quality and quantity. risks. In most cases so far, E&S system improvement and risk mitigation • Efficiency: The level of implementation of the works, as compared to the agreed- have been addressed, first, by incorporating E&S-focused indicators into upon contract-approved work programs and outputs. Indicators included physical MCs or PMs (or both, in some cases) to be measured under the APA; and, progress lag, timeliness in payments, percentage of quality progress reports, and second, by providing targeted and comprehensive capacity-building percentage of approved equipment onsite, among others. support to participating local governments in these areas (for example, • Effectiveness: The quality of the works being implemented and the utilization through expanded training modules on E&S issues and by assisting local of infrastructure. Indicators included presence of material test results on file, percentage conformance of site works to design drawings and specifications, governments in developing necessary plans). Incorporating E&S-focused percentage of conformance of audit test results to specifications, and whether indicators into the performance assessment framework has emerged as the infrastructure was in use by the intended beneficiaries. a good practice because it gives local governments incentive to focus The VfM results demonstrated that the USMID-funded infrastructure investments fully on understanding and implementing E&S requirements under were more economical, efficient, and effective than non-USMID investments. The country laws and program systems and creates awareness about these USMID-funded projects significantly minimized disparities in unit costs, and they filed detailed qualitative progress reports and minutes of site meetings conducted pertinent issues. Capacity gaps in addressing E&S issues at the local for all the participating local governments. Materials test results were also on file level, however, persist as a challenge, with inadequate staffing and skills for all USMID infrastructure investments. Key challenges remaining were delays in commonly experienced by local governments in most UPG programs. contracts and quality control of the infrastructure construction process, especially As the next chapter, shows, the provision of financial incentives through for non-USMID projects. grants needs to be supplemented by centrally driven capacity-building support and technical assistance in these areas. Source: Government of Uganda 2019. Performance-Based Fiscal Transfers for Urban Local Governments 43 Part I Results and Lessons from Two Decades of World Bank Financing 4. Lessons and Best Practices for Figure 4.1. Mutually Strengthening Core Components of a UPG Program Urban Performance Grant Programs This chapter draws upon findings from case studies to present lessons learned and best practices on the three key components that inform • Performance-based award successful design and implementation of a UPG program: grant design, • Clear formula-based grant operationalization, and capacity-building support for the local governments participating in the program. 1 distribution of divisible pool of program funds Grant Design • Significant local discretion on use of funds 4.1. Three core components of a successful UPG program • Significant relative + A successful UPG program has three mutually reinforcing core absolute size of grant components, each of which needs to be carefully designed and implemented over the course of the program to achieve the desired development impact: grant design, grant operationalization, and capacity-building support for the local governments participating in the program (Figure 4.1). Even before reviewing the lessons and best practices for these three components, however, it is important to note that the first step in scoping and planning a UPG program is to review 2 3 whether contextual and external factors in the country (including the prevailing institutional environment) are conducive to ensuring such a program can be functional and effective. This initial understanding of Grant Assessment Capacity- Building the local context then informs the grant design, where such key factors & Management Support as per capita grant size, scale of grant-funded infrastructure projects, and grant allocation formulas are determined. Subsequently, during the operationalization and implementation of the grants, the performance of participating local governments is measured, assessed, and verified, • Assessment system • Combination of demand- the results of which are used as a basis for grant disbursement to the with clear performance driven and supply-side indicators support to local govts participating local governments. Finally, targeted capacity-building • Annual assessment • Provides targeted support is crucial for all levels of governments to maximize and sustain process with robust support for weaker areas the impacts of UPGs. The subsequent sections of this chapter discuss quality assurance system • More easily accessible lessons and best practices for each of these components. • Program coordination than capital grant Performance-Based Fiscal Transfers for Urban Local Governments 44 Part I Results and Lessons from Two Decades of World Bank Financing 4.1.1 The critical role of contextual and external factors 4.2. Lessons and good practices on key design factors for The presence of an enabling institutional environment, such as a urban performance grants decentralization agenda and related initiatives, is central to the Adequate design of a UPG program is important to ensure strong success of UPG programs. Among the major nonconducive conditions incentives for local governments and to enable the implementation that impede their effective design and implementation are centralizing of meaningful investments and expenditures for improved service policies that limit the mandates and functions of local governments delivery. According to best practice, the design of UPGs should consider (including and especially pertaining to OSRs), fragmentation in local the following four key elements (see Figure 4.2): government systems, and human resource constraints, as well as 1. The share of performance-based grants relative to other funding insufficient support from central to local levels in capacity building, streams and budgetary resources of local governments: UPG development of guidelines, and other areas. A recent example is funding needs to constitute a substantial enough share of the total provided by Tanzania, where property tax collection was centralized in funding available to local governments to provide a sufficient 2016 and, thus, all property tax–related indicators had to be taken out of incentive for them to undertake actions to gain access to these funds. the performance measurement metrics for the UPG program (ULGSP), 2. The nominal monetary size of performance grants for which with the result that the general level of discretionary formula–based local governments are eligible once the relative size is sufficiently development funds decreased. substantial: Once the size of the grant established is substantial Unexpected external shocks can also negatively affect the design and relative to the total budgets of the local governments, a nominally implementation of UPG programs. Most of the programs reviewed higher grant size will provide them with stronger incentive to improve reported that the COVID-19 pandemic changed the policy priorities their performance to gain access to the funding, all else being equal. This is especially true for urban local governments catering to larger of client countries, and program preparation and implementation populations and/or service delivery needs. processes were significantly delayed. The design of some UPGs was also 3. The average size of the grant-funded capital infrastructure altered to support the emergency responses of the affected countries and investments to be executed by the local governments: Larger to focus more on improving health aspects. A UPG program in Tunisia investment projects executed by local governments lead to more (PDUGL), for example, was restructured in response to COVID-19 efficient utilization of grant funds and reduced likelihood of diffused to adapt activities and targets to sanitary conditions and support the impact. economic recovery of local governments in the medium term. 4. The grant allocation formulas—how the divisible pool of grants is allocated across local governments—and how the performance of local governments affects grant allocation: These formulas also contribute to establishing the incentive structure, as an improvement in performance by a participating local government should lead to a corresponding increase in access to funding; otherwise, the incentive and impact of the PBG scheme are diluted. Performance-Based Fiscal Transfers for Urban Local Governments 45 Part I Results and Lessons from Two Decades of World Bank Financing Figure 4.2. Four Key Elements of the Urban Performance Grant Design Share of performance grants relative to other funding streams 1 2 Actual nominal per capita grant size Key Elements of the UPG Design Grant allocation formulas: 4 3 Average size of grant-funded How divisible pool is allocated investment projects and costs of across local governments and how meaningful investments performance impacts allocation In addition to these elements, at an overarching level UPG programs the likelihood of stronger coordination and timely flow of funds. It also should ideally be embedded into existing intergovernmental fiscal helps develop the capacity of local governments and central government transfer (IGFT) systems to leverage these systems for fund transfer entities to manage these types of fiscal incentives and reinforces local and oversight. This often entails a flow of funds from a central agency20 governments’ accountability to higher tiers of government.21 The UPG to participating local governments. With the program embedded in it, programs reviewed in Ghana, for example (LGCSP and GCSCP), were oversight of the IGFT system would typically be entrusted to a national or closely linked with the existing national performance grant system (the provincial agency in charge of finance, local governments, or planning. District Development Fund, or DDF). The Bank-supported program Utilizing IGFTs also avoids incurring transaction costs associated with specifically used the results of the DDF as the minimum criteria for the development and management of parallel systems and increases grant eligibility. 20 Such agencies may include a ministry of finance or of local government or urban affairs or a provincial agency. 21 For further discussion on these points, also see UNCDF 2010. Performance-Based Fiscal Transfers for Urban Local Governments 46 Part I Results and Lessons from Two Decades of World Bank Financing 4.2.1. Share of PBG relative to other funding streams and Figure 4.3. Share of UPGs in Total Fiscal Transfers to budgetary resources of local governments Participating Local Governments UPG funding needs to constitute a substantial enough share of total funding available to local governments to provide a sufficient incentive for them to undertake actions to gain access to these funds. 80 In six out of the nine programs reviewed, UPG funding accounted for 70 70 Share of UPGs in total transfers (%) a sizeable share (ranging from 25 percent to as much as 70 percent) of the total intergovernmental fiscal transfers to local governments (Figure 60 4.3). UPG funding in these programs also constituted an overwhelming 50 majority of total development and capital grants to local governments 38 40 39 (that is, total fiscal transfers less recurrent grants), ranging from around 40 70 to 90 percent across these six programs. It was also observed that the 30 25 two cases with the lowest UPG shares had the lowest rates of compliance 20 with MCs, as shown in chapter 3 (Table 3.1). Although the wide variation in the share, depending on local context, makes it impossible to provide 10 9 5 categorical advice on this issue, the experience suggests UPGs need to 0 constitute at least 20 percent of the total discretionary development Ethopia Ghana India Kenya Tanzania Tunisia Uganda and capital grant budget available to local governments to have their intended incentive effects. Source: Team analysis based on data provided by World Bank task teams for respective UPG programs. The presence of other substantial fiscal transfer streams in parallel Note: The bars represent the share of UPGs in total fiscal transfers, including capital and recurrent to a UPG scheme can also reduce the financial incentives for local transfers to participating local governments, in the UPG programs reviewed in each country. Data governments to benefit fully from UPGs. This was found to be the case are from latest FY figures for most cases or figures at the program design stage. with the UPG program in India-West Bengal. The share of UPG funding in total fiscal transfers to participating local governments in the state fell from around 20 percent at the program design stage to only 9 percent by The presence of other substantial fiscal transfer streams also affects 2020, due to an increase in the share of non-performance-based grants the absorptive capacity of the receiving local governments to spend from the central government through Central Financial Commission all these fiscal resources in an appropriate, timely, and efficient (CFC) funds.22 This substantial shift toward non-performance-based manner. Absorptive capacity refers to the ability (and capacity) of grants changed the financial incentives for local governments and local governments to plan for and spend the financing available to reduced the effectiveness of the qualitative performance measures them for capital investment needs. Weak absorptive capacity hinders applied by the program. higher infrastructure investment overall, regardless of the source of the funds. This is a common issue in several contexts where such programs operate, including with respect to local governments and contexts that are relatively advanced. A recent Bank report, for example, has shown this to be case even for local governments in large and economically 22 Only 10 percent of the CFC funds were in the form of performance-based grants. Performance-Based Fiscal Transfers for Urban Local Governments 47 Part I Results and Lessons from Two Decades of World Bank Financing prosperous cities in India. These governments have been receiving Figure 4.4. Relationship between Average Sizes of UPGs per increasing levels of fiscal transfers for capital investment in recent years, Capita and Capital Investment Projects but constraints on absorptive and project design and implementation capacity have prevented them from utilizing these increased resources fully (see Athar, White, and Goyal 2022). 2 4.2.2. Monetary size of performance grants and capital Larger/regional Uganda investment infrastructure investments financed by these grants (USMID AF) (1 UPG) Average investment project size Best practice shows the importance of ensuring annual UPGs are 1.5 substantial in size in per capita terms and of allowing local governments to undertake large capital investments so they have adequate incentive (US$ million) to achieve performance targets and make effective service delivery 1 improvements. In practice, large variations are apparent in the size of the grants on a per capita basis and in the average infrastructure investments Ghana undertaken by the local governments utilizing them, depending on the (GSCSP) Medium size 0.5 (5 UPGs) context. As Figure 4.4 shows, for example, annual grants in the cases Ethiopia reviewed ranged from US$1 per capita (for programs in India-West Kenya Tunisia (ULGDP II &UIIDP) Bengal and Kenya) to US$45 per capita in one case (Uganda), while the India Small size Ghana (LGCSP) (2 UPGs) average project size for grant-funded capital infrastructure investments 0 ranged from US$60,000 (for programs in India-West Bengal and Ghana) 0 10 20 30 40 50 to as much as US$1.8 million (in Uganda). The case studies indicated that Average per capita UPG size (US$) a sufficiently large grant tends to enable local governments to undertake capital investments in strategic infrastructure and services at a larger Source: Based on data provided by World Bank task teams for respective UPG programs. scale in complex urban settings. UPG programs with an average annual Note: The X-axis represents average UPG size per capita. The Y-axis refers to average monetary size of investment subprojects under each UPG program. One UPG program in Tanzania is not grant of more than US$5 per capita, for instance, tended to invest in included due to data unavailability. larger investment projects (between US$200,000 and 1 million) than those with grants of less than US$5 per capita. When the per capita size of grants to typical local governments was less than US$5–10, especially Broadly speaking, the experiences reviewed for this report suggest in urban areas (the figure might have been lower for smaller rural that annual performance grants for recipient urban local governments local governments such as village councils), the Bank staff providing should be larger than US$5–10 per capita and should finance capital implementation support to these programs generally perceived this investment projects of a meaningful size (at least US$200,000 per to be too low to provide substantial improvements in service delivery, project, on average). This observation needs to be qualified, however, and the investment projects in such programs were smaller on average in that absolute amounts obviously need to be considered relative to (under US$200,000). broader local government budgets and to the institutional size and functional mandates (expenditure assignments) of the participating local governments. A grant of US$1–2 per capita to a small rural Gram Performance-Based Fiscal Transfers for Urban Local Governments 48 Part I Results and Lessons from Two Decades of World Bank Financing Panchayat (village council) in West Bengal state in India, responsible for limited functions and with a population of around 20,000, means Box 4.1. Lessons on Improving the Equity of Performance Grant something quite different than a grant in the same amount to a Allocations among Local Governments medium-size city in Uganda responsible for delivering networked urban Generally speaking, UPG programs are not designed for the equitable distribution infrastructure systems. of funds among local governments. Equitable distribution is generally covered by equalization factors in the non-performance-based aspects of intergovernmental In Ghana, the first UPG program in the country (LGCSP) was fiscal transfers before the impact of performance is considered. Performance metrics designed to provide relatively small grants to local governments (less under UPG programs tend to focus more on local government “institutional efforts.” In the reviewed UPG programs, fiscal incentives were often stronger for smaller local than US$5 per capita) and to spread them across multiple sectors governments and those with relatively weak fiscal positions, who were often able to over four hundred small-scale investment projects, including to compete successfully with the larger local governments for access to funds. The health clinics, schools, toilet facilities, security facilities, and roads. UPG programs were designed with the objective that performance measures would The small (in per capita terms) grant size and the large number of provide an equal playing field. In some cases (for example, in Ethiopia and India-West Bengal), some phasing-in of more advanced conditions was done for newly enrolled diffused infrastructure projects caused fragmentation of investments local governments. Some performance indicators can be designed to measure and reduced the efficiency of procurement processes. As a result, the progressive improvements in performance regardless of the size or capacity of the program gradually moved toward fewer, medium-size investments. local governments. These may include, for example, measuring percentage increases Reflecting the lessons learned from it, Ghana’s second UPG program in OSR annually instead of in nominal amounts, and measuring the execution rates of (GSCSP) was designed with higher annual grants for local governments local governments’ respective capital investment budgets. in per capita terms (up to US$15 per capita) and was focused on larger It is generally recommended that UPG programs focus on performance aspects and not attempt to change the prevailing basic distribution of fiscal transfers to capital investments of over US$500,000. local governments, except when it comes to their performance. The impact of The prevailing institutional context often determines the average performance on fiscal allocations can be weighted with the basic allocation to ensure adequate incentives for all types of local governments. In addition, it is important to grant size per capita in such programs, with several countries choosing ensure that performance measures reflect the efforts of these governments and to extend access to performance grants to all local governments. This are fully under their attribution and scope, giving them the authority to achieve necessarily reduces the grant size by diffusing the total volume of these measures. Local governments that are unfamiliar with such programs and are program funding across more local governments. Examples include being introduced to them in a phased manner can have transitional arrangements, especially in the first year of the program, to acquaint them with the performance the UPG programs reviewed in India-West Bengal, Kenya, and Tunisia. process. Finally, it is important to ensure the coordinated provision by the central Within a limited amount of total available financing, prioritizing wider government of substantial capacity-building support, especially for the weaker local geographical coverage will likely reduce the average grant size per local governments, to enable them to perform better. government and, if reduced too much, may distort financial incentives Generally, there will be a tradeoff between designing the program with fewer local for those participating. Hence, the implications of reducing the average governments to ensure the provision of strong financial incentives focused on grant size should be clearly communicated to all stakeholders at the performance versus spreading the incentives over more local governments with consequently smaller grant amounts for each. design stage (see also Box 4.1 for a discussion on improving the equity of grant allocations among local governments). Performance-Based Fiscal Transfers for Urban Local Governments 49 Part I Results and Lessons from Two Decades of World Bank Financing 4.2.3. Grant allocation formulas and how local government When the overall size of a performance grant—that is, on a per capita performance affects grant allocation basis; on the basis of the average capital investment size to be financed The grant allocation formulas—how grants are allocated across by these grants; and/or relative to the total fiscal transfers received by local governments—and how the performance of local governments the local governments—is considered insufficient to create meaningful affects grant allocation also contribute to establishing the incentive impacts and incentives for these governments, innovative mitigating structure, as an improvement in performance by a participating local design measures may to be adopted. One possible solution currently government should generally lead to a corresponding increase in access being used in some countries is to design the grant allocation formula to to funding; otherwise, the incentive and impact of the PBG scheme are increase exponentially the impact of local governments’ performances on diluted. A typical allocation formula has a linear relationship between the actual monetary allocation to provide them with stronger incentives. performance and grant allocations (Figure 4.5 shows a typical model) The UPG program reviewed in Kenya (KDSP) is a representative case; until a grant amount reaches a monetary ceiling for a particular local there, the average annual grant for local governments was found to be government, as determined by the program. less than US$2 per capita and to constitute only 5 percent of the total fiscal transfers for local governments, but an “exponential” allocation formula was applied to maximize the impact of the UPG program and improve incentives (see Box 4.2 for details). This led to substantial Figure 4.5. Typical Linear Relationship between Local performance improvements across participating local governments. Government Performance and Grant Allocation 4.2.4. Summary of key design factors for urban performance grants Grant allocation in nominal currency The key lessons on design factors for successful UPG programs based 25 on lessons learned are summarized in Figure 4.6. It should be noted, however, that the various factors interact and are mutually influencing. 20 15 10 5 0 No Yes 10 20 30 40 50 60 70 80 90 100 MC Compliance Performance Score Performance-Based Fiscal Transfers for Urban Local Governments 50 Part I Results and Lessons from Two Decades of World Bank Financing Box 4.2. An Example of Exponential Grant Allocation Design in Kenya Figure B4.2.1. Difference in Grant Size for a Local Government In general, the monetary size of a UPG provided to a local government is between Linear and Exponential Models linearly proportional to that government’s measured performance, but the Kenya Devolution Support Program (KDSP) introduced a new allocation 3 approach to achieve institutional performance improvements by increasing the UPGs’ incentive effect. Given the much smaller average annual grant per capita relative to other UPG programs, the grant size was designed to vary Capital grant size (US$ millions) exponentially with performance scores. This means a marginal improvement in the performance score of a given local government significantly increased its 2 eligible grant amount, providing a stronger incentive effect. Linear As Figure B4.2.1 illustrates, the variation in grant size for a local government allocation under such an “exponential” model can be much larger in either direction (that is, higher or lower) than under the “linear” model. According to the simulation Exponential presented in the figure, the exponential approach would provide the best- 1 allocation performing local governments with grants up to 300 percent (or four times) more than those for the lowest performers, while under the linear model this difference would be only about 50 percent. The grant allocation formula is based on key factors in addition to the performance scores, such as population, poverty, equal shares, land area, and others. 0 In Kenya, the resulting grant amounts in absolute monetary terms under 60 65 70 75 80 85 90 this program ranged from US$400,000 to US$600,000 for very small, Indicative performance score (out of 100) low-performing local governments to US$6 million to US$8 million for well- performing ones. This led to consistently improved institutional performances across the participating governments during the implementation stage. The way the formula was developed, as well as the publicity and awareness surrounding it, generated incentives for them to rally to the top level. More broadly, this approach can be used in cases where the total funding envelope (divisible pool) is small relative to needs or other fiscal resources, or where the program covers a very large number of local governments (which in turn reduces the funding per government). This is because it has the advantage of amplifying the performance of each local government (through the nonlinear allocation) and thus can mitigate against the smaller program envelope that would otherwise lead to diluted financial incentives. The approach also has a clear tradeoff, however. Local governments that are on the lower end in terms of performance will receive proportionally many fewer grants relative to a linear allocation model, which may also in theory have a disincentive effect and may not provide them with sufficient fiscal resources. This will certainly affect the overall efficacy of the program, with some local governments having access to far more funds than others. Performance-Based Fiscal Transfers for Urban Local Governments 51 Part I Results and Lessons from Two Decades of World Bank Financing Figure 4.6. Summary of Lessons and Good Practices on Key Design Factors for Urban Performance Grants Share of UPGs relative to other funding 1 2 Monetary size of grant on per capita basis • Although it depends on local context, • Nominal grant size needs be at least UPG should not be <20% of total US$5-10 per capita for a typical development grants to urban local Key Design urban local government. governments. Factors for • Nominal size may be less important UPGs than UPG share relative to total budget. Grant allocation formulas can mitigate funding size risks 4 3 Average size of investment projects • Allocation should be proportional to • Medium to larger investment projects performance. (over US$200,000 per project) • When total available funding is are encouraged to make meaningful small (on a per capita basis) relative impacts, especially in urban areas. to total fiscal transfers, mitigating • Grant size to be determined by local design measures work. context, size of local governments, and whether they are urban or rural. Performance-Based Fiscal Transfers for Urban Local Governments 52 Part I Results and Lessons from Two Decades of World Bank Financing 4.3. Lessons on implementation of urban performance Figure 4.7. Share of the Cost of Conducting APAs grants out of Total UPG Program Financing As with grant design, it is important to develop a performance measurement and assessment process that is rigorous, credible, and Estimated cost of APA over life of UPG program, as % share of total program financing, for selected UPG programs objective enough to build trust among stakeholders and provide adequate incentives to local governments. This generally includes India (ISGPP II) 0.5 the process of collecting and reporting accurate data on the results of participating local governments relative to the established targets and Uganda (USMID) 0.8 milestones and the allocation of UPG funding to the governments eligible Tunisia (PDUGL) 0.8 based on those results. UPG programs make use of an annual performance assessment (APA) system to measure and verify the results of participating Kenya (KDSP) 1 local governments on the performance indicators (generally the MCs and Ethiopia (ULGDP II) 1 PMs). This APA process has been frequently utilized and substantially strengthened over the evolution of UPGs over the past two decades. Ethiopia (UIIDP) 1.5 4.3.1. Importance of reliable APAs to ensure credibility and Ghana (LGCSP) 2 transparency Ghana (GSCSP) 2 Implementation experience, especially from the UPG programs reviewed here, has shown that the APA process is best conducted by an independent external party to make it more objective and neutral. Note: Cost estimated by World Bank task teams. This approach has tended to reduce conflicts of interest and increase the quality of the assessment. Such a party is usually hired and paid for by the national or provincial ministry or agency managing the program. The An independent APA is the recommended approach to assessing key is to try to keep the assessment process separate from the rest of the performance-based fiscal programs. Other approaches can have grant administration machinery, especially from the agents in charge of inherent conflicts of interest and have, therefore, not been highly the program and the delivery of capacity-building support. Contracting successful, especially when such programs are novel in a specific a third party assessor who is not involved in the UPG disbursement context. These alternatives may be considered in more sophisticated process is, therefore, fundamental to ensuring the independence of settings where the programs have an implementation history, subject the APA process, enhancing the division of responsibilities, providing to adequate measures for quality assurance and efforts to reduce adequate incentive to conduct it in a timely and efficient manner.23 conflicts of interest. In such cases, an independent party should still be Although government often considers it costly, the benefits of a robust hired to perform sample quality assurance checks and/or audits of the APA process outweigh the costs of managing and implementing it, assessment. which, for all the cases reviewed, proved to be generally less than 2 One alternative is to have the APA conducted by a dedicated percent of the total UPG program cost (see Figure 4.7). central government entity; this might contribute to strengthening the performance management system of the government entity involved, 23 For further discussion on this aspect, see also UNCDF 2010 and Salhab et al. 2022. Performance-Based Fiscal Transfers for Urban Local Governments 53 Part I Results and Lessons from Two Decades of World Bank Financing which can promote the long-term adoption of such systems. Such an thorough quality assurance review of the assessment on a sample basis. entity may, however, lack the necessary capacity for the desired rigor, Discrepancies identified from this review are discussed and resolved which may reduce the effectiveness of the assessment, and it may be with the government agency before the assessment is finalized and subject to political incentives that reward the disbursement of funds, forms the basis of results achievement and fiscal allocation. This review which can dilute the overall reliability of the assessment process. During provides all stakeholders with a high degree of confidence in the reported the first generation of UPG programs supported by the Bank, some results achieved and thus functions as an effective risk management APAs were conducted internally by government entities and/or without measure. Figure 4.8 shows there is a set of choices available to program adequate quality assurance processes, resulting in reduced robustness of designers (and the Bank and other development partners supporting the APA process over time. In Tanzania and Uganda, for example, where these programs) along a spectrum, which will determine the credibility central local government agencies conducted APAs in some earlier of the performance assessment. phases of the UPG programs, the programs faced important challenges, such as outside observers’ skepticism of assessment objectivity, limited checks and balances, greater instances of conflicts of interest, and lack of standardization.24 Such issues ultimately led to deteriorating quality in these systems, which were addressed in their second generation. Indeed, most second-generation UPG programs put into place two key processes to improve the credibility of the APA system: • They contracted out the assessments to external entities, who also shared the findings with stakeholders, including higher-tier governments (as discussed above). • They instituted robust quality assurance systems, whereby the results of the APAs would be thoroughly reviewed both by an external consultant contracted by the program unit or government and by the World Bank quality assurance team, often consisting of Bank staff and a hired specialized consultant, who would conduct either sample-based field reviews or desk-based reviews. Implementation experience has shown the importance of putting into place a quality assurance and review system for verifying the accuracy of the assessment results submitted by the government agency managing a UPG program. Since the program rests on the independent and credible conduct of an APA of the participating local governments— usually done by a contracted third party—central government agencies and World Bank implementation support teams often undertake a 24 For further discussion on this aspect, see also UNCDF 2010. Performance-Based Fiscal Transfers for Urban Local Governments 54 Part I Results and Lessons from Two Decades of World Bank Financing Figure 4.8. Spectrum of Options to Improve Credibility of the Performance Assessment APA by an external party* * It can be a private firm or a specialized government agency (e.g., in Tunisia) NO YES Quality Assurance System in place NO YES External consultants hired both by govt. and the World Bank NO YES APA conducted internally Despite APA conducted • Quality check by government • Two-layer system or without an arm’s-length independently, no systematic or the Bank • Critical to building robust and distance from program way to find and reconcile • Can be either formalized reliable process management responsibilities, discrepancies reviews by external • Government and the Bank with weak or nonexistent quality consultant or general reviews can form a join team (e.g., assurance system by teams (e.g., Kenya) Ethiopia and Uganda) Increasing level of robustness and credibility of grant operationalization process Performance-Based Fiscal Transfers for Urban Local Governments 55 Part I Results and Lessons from Two Decades of World Bank Financing The Bank has provided substantial technical assistance to • Publicly sharing results: Making results publicly available implementing agencies for institutionalizing such a robust system promotes transparency, builds the credibility of the system, raises for UPGs. In Uganda, some phases of the countrywide PBG program overall awareness of the UPG program, allows participating local relied heavily on a countrywide APA system, wherein the central governments to compare their performances with those of their government inspection staff engaged some external consultants only peers, and bolsters accountability. for quality assurance reviews. The second-generation UPG program • Ensuring timely and prompt processes for results sharing and (USMID), though, successfully strengthened its APA system by dispute resolution: Timely sharing and prompt dispute resolution contracting out APAs to an external firm, supplemented by the World allows the program to comply with established schedules, which is essential for participating local governments to plan and execute the Bank’s quality assurance team and a reconciliation system between the grant funding allocations. firm and the implementing entity (and development partners) in case of discrepancies. In India-West Bengal (IGSPP II), the government hired 4.3.2. Importance of timely completion of APAs and alignment external consultants only to conduct quality assurance reviews during with government budgetary process the first year of implementation. From the second year, however, the Synchronizing the APA cycle with the intergovernmental annual program followed a “two-layer” quality assurance approach, and the Bank budget planning cycle is critically important to ensure grant funding team hired specialized consultants as well to help improve the process. allocations are reflected in the participating local governments’ annual Similarly, in Ethiopia, the quality assurance system was strengthened budgets on time. In a typical UPG program, each APA reviews the from the first-generation (ULGDP I) to the second-generation (ULGDP performance of the participating local governments at the end of the II and UIIDP). fiscal year and determines the UPG allocations for the subsequent fiscal year. Thus, to be effective, it is important that the APA timeline is The completion of the APA ideally is followed by a process in which well synchronized with the annual budget cycle and is launched at an the performance results are certified and reported. In most such appropriate time during the fiscal year, and the external entity (often programs, assessment results are certified by the central government a private firm) that will conduct the APA needs to be procured in due ministry or agency in charge of disbursing resources or the program’s time to launch activities. steering committee (when applicable). To promote transparency, recommended good practices include the following: A challenge commonly reported among the UPG programs reviewed was repeated delays in completing APAs. Seven of the nine UPGs • Preparing the participating local governments for the APA: The necessary resources need to be invested in the APA process and experienced critical delays in APA processes, citing several primary in awareness raising for the local governments, with prior internal reasons: assessments and learning conducted to establish a baseline of • Prolonged government procurement processes performance. • Mismatched timing of a program launch (that is, the start of the • Defining a process by which parties can attest to the results: program in the middle of the fiscal year) and approval steps, including Even with a robust, objective, and neutral assessment process, negotiations and approvals by parliaments stakeholders—especially participating local governments—should be • Logistical challenges in assessing the results of a large number of provided an opportunity to dispute or question results if they believe local governments their performance has been incorrectly assessed (if, for example, mistakes have been made in the assessment process). • Time-consuming reconciliation processes Performance-Based Fiscal Transfers for Urban Local Governments 56 Part I Results and Lessons from Two Decades of World Bank Financing Ideally, an APA cycle should be completed before the start of the new participate in the process. This will also reduce procurement-related fiscal year so the APA results can correctly determine the size of the administrative transaction costs for the government coordinating grants for each local government, the resulting funds can be disbursed agency. on time, and the local governments can develop their investment plans • New technology and data tools can be substituted for several tasks accordingly (see Figure 4.9 for a typical APA cycle during a fiscal year). that previously required in-person field verification by the assessment entity. This can reduce the time and cost of the assessment. While the traditional APA processes are not costly relative to the total program Figure 4.9. Typical APA Cycle during a Fiscal Year financing (as seen in the previous section), innovative solutions, Spanning January 1 to December 31 such as remote implementation support tools and online reporting systems, can help further improve the efficiency of APA processes. • As noted earlier, a careful start of project launch and linkages with APA conducted in Quality assurance Approval of APA results the local government budget cycle should be ensured. shortly after close of review and verification by Government FY, to inform grant of APA results Once a set of performance indicators is designed, programs should allocations for next FY consider keeping these metrics relatively stable, avoiding frequent and/or large-scale changes over the implementation period. If a change to the indicators is inevitable, it should be communicated Local govt budgets Local govts do budget Announcement of to participating local governments in time for them to respond to it. finalized & approved, planning for next FY, results and indicative including UPG grant based on announced grant allocations for A solution to potential negative effects of measurement changes is to allocation amount grant allocations next FY conduct a pilot assessment, using the APA tool with a sample of local governments before any alterations to the indicators are in effect. In India-West Bengal (ISGPP II), performance indicators were significantly Grant disbursed to Capacity building adjusted in the middle of implementation to improve fiscal incentives local govts, at start of support to all local for local governments, timing that, in part, made it difficult for them new FY govts across the year to improve performance to respond to the revised assessment system. Reportedly, more than 3,000 participating local governments were insufficiently aware of the changes made and were thus unable to keep their performance levels up to where they were prior to the change. The conduct of the APA process and determination of grant funding under 4.3.3. Summary of lessons for credible performance a UPG program may be completed in a timely manner by considering the assessments and program operationalization The key lessons on recommendations for effective and credible following recommendations based on implementation lessons: performance assessments and program operationalization are • The external entity (often a private firm) that will conduct the APA summarized in Figure 4.10. needs to be procured in due time to launch activities. Multiyear contracting for APA firms is encouraged to raise interest from high-quality vendors and provide them with stronger incentive to Performance-Based Fiscal Transfers for Urban Local Governments 57 Part I Results and Lessons from Two Decades of World Bank Financing Figure 4.10. Summary of Lessons and Good Practices for Effective and Credible Performance Assessments Well-designed performance indicators • Align APA thematic areas with objectives of UPG program • Select indicators that are under local government control • Balance between comprehensiveness and impacts Neutral and objective APA protocols • Conduct through third party • Can be seen as similar with systems of external audit Robust verification and quality assurance system • External, “two-layer” quality assurance system is recommended for credibility Clear and transparent reconciliation system • Systematically reconcile any discrepancies found in APA results • Close collaboration between World Bank and government Value for Money (VfM) Audits • Utilize VfM audit findings into APA indicators where feasible • VfM costs need to be affordable relative to grant size (more relevant for larger urban projects) Launch of APA Reflections of APA results Synchronize program launch with • Disseminate & publicize APA results for transparency & accountability local government budget cycle • Use APA results as input to capacity building support focusing on weak performance areas • Incorporate APA results into overall program monitoring system Performance-Based Fiscal Transfers for Urban Local Governments 58 Part I Results and Lessons from Two Decades of World Bank Financing 4.4. Importance of capacity building to support program on addressing bottlenecks identified from the APAs. On the other hand, implementation and sustainability a UPG program in Tanzania (ULGSP) included a DLI on completion of The review of the case-study UPG programs demonstrated that annual capacity-building activities executed by the central government, capacity-building support plays a central role in strengthening the but the plans the central government prepared did not sufficiently institutional performance of local governments and improving urban respond to local capacity-building needs, and zonal teams were unable infrastructure and service delivery. Over the course of rolling out to find qualified key staff for the activities. As a result, the program and expanding the UPG programs in Ethiopia, India-West Bengal, and faced challenges in implementing the capacity-building plans, and a Uganda, the substantial capacity-building support that was provided disbursement under a DLI on completion of annual capacity-building enabled the local governments new to these programs to catch up activities by the central government was dropped and reallocated. quickly with those that had participated in a precedent program (see Demand-driven capacity-building activities have often been Box 4.3 for how this was rolled out in West Bengal state in India). A supported through a separate and smaller funding allocation to local recent empirical study on a UPG project in Mozambique also showed governments, often called a capacity-building grant. In general, UPGs that technical assistance and capacity-building support provided by require the achievement both of more elaborate MCs to get access to the central government had a significant impact on OSR improvements the grants and PMs that determine their size. Capacity-building grants, across program local governments—in fact, the effect of the OSR however, offer local governments easier access conditions than typical improvement was stronger for those governments that benefited from UPGs. Their disbursement can be triggered when local governments capacity-building support (see Erman, Uehara, and Beaudet 2021). Well- comply with the MCs that focus only on the preparation, planning, designed capacity-building support also targets specific performance and execution of capacity-building activities. Programs in Kenya and areas that present institutional barriers for local governments working Uganda, for example, had capacity-building grants. In Ethiopia, Ghana, to improve their performances. Most such support provided to the UPG and Tanzania, a slightly different model was applied. Instead of creating programs reviewed addressed the institutional weaknesses that were a separate funding window for capacity-building activities, the programs identified from APA results and baseline assessments. This targeted allowed participating local governments to spend up to 5 percent support effectively helped prepare the local governments to gain of UPG funding on them. Typical demand-driven capacity-building access to grants by achieving planned APA results and to improve their activities were based on multiyear and/or annual plans prepared by the performances, especially during the early years of implementation. local governments and included formal, classroom-style training from Most of the UPG programs reviewed adopted a balanced capacity- prequalified local institutions; consultancy support; and IT systems and building strategy by combining supply- and demand-driven activities. equipment. Supply-side capacity-building activities included support from central and/or regional teams to local governments, such as mentoring, training, and system development. Well-designed incentives for the central and regional governments proved central to the successful delivery of the planned activities. In Ethiopia, for example, where disbursement- linked indicators provided incentive for team mobilization and regional support to the local governments, the result was the effective provision of customized technical assistance and capacity-building support focused Performance-Based Fiscal Transfers for Urban Local Governments 59 Part I Results and Lessons from Two Decades of World Bank Financing Box 4.3. Impacts of a Strategic Capacity-Building Approach on Local Governments in West Bengal State, India Figure B4.3.1. Positive Relationship between In West Bengal state in India, UPG programs have been scaled up across two Mentoring Support and Local Government Institutional generations. A first-generation program, called the Institutional Strengthening of Gram Performances in India-West Bengal Panchayats Project (Phase I, 2010–16), targeted 1,000 village-level local governments in the state; this was expanded to include all village-level local governments in the state (3,342 in total) under the second-generation program (Phase II). 70 The deployment of mentors to local governments as part of the capacity-building support provided to these programs was recognized as one of the most successful aspects of the first phase, and it was further expanded under the second. Over 400 Average APA Scores specialized mentors were deployed to all local governments with access to regular, 65 on-the-job backstopping and advice. The midterm review of Phase II observed that the intensity of the mentoring support had a positive relationship with the institutional performances of the local governments, as shown in Figure B4.3.1. 60 An interesting feature of the Phase II capacity-building support was the employment of different strategies for two local government groups: the 1,000 Phase I local governments, who had previously received program support, and the 2,342 Phase II local governments, who had received no support before that program phase. 55 The program gradually reduced the intensity of the mentoring for the Phase I local 110 115 120 125 130 governments while providing consistently strong support to those in Phase II. This was Mentoring support hours per year to ensure the program provided equitable support to all the local governments, given the different levels of understanding of UPG systems between the two groups. Note: Based on data provided in the midterm review of Phase II (2020). A comparison of the performances of the two local government groups made by the midterm review, however, highlighted a decline in the performance of the Phase I local governments, partially because of the reduction in the mentoring they received. The Phase II local governments, for example, increased their compliance rates on the basic minimum (level one) conditions from 2016 to 2018 by 56 percent, whereas the Phase I governments decreased their compliance by 11 percent. A possible interpretation of this result is that the focused capacity-building support provided during Phase I helped the participating local governments maintain their performance levels in the first year of the Phase II program, but the lessening of mentoring intensity gradually resulted in a decline in the number of qualified Phase I local governments in the following years. The capacity-building support strategy was adjusted to address this gap between the two groups, recognizing the importance of continued capacity-building support to all local governments to achieve the long-term objectives of the UPGs and the need to incorporate this support into the financial incentives for local governments. Performance-Based Fiscal Transfers for Urban Local Governments 60 Part I Results and Lessons from Two Decades of World Bank Financing Figure 4.11. Recommendations for Effective Capacity-Building Support Combine the typical Combine demand-and-supply- capacity-building support driven approaches with system development • Provide a specific capacity-building • Combine capacity-building grant allocation or ceiling in UPGs thematic areas with system Effective development such as audit, IT, Capacity-Building PFM, and M&E to ensure Support a conducive framework Capacity-building result Equitable support monitoring • Reflect APA results in capacity- • Monitor results (e.g., through building planning disbursement-linked Indicators) • Support weakest local governments and links between intermediate to climb up the performance ladder and results indicators and compete on equal terms Performance-Based Fiscal Transfers for Urban Local Governments 61 Part I Results and Lessons from Two Decades of World Bank Financing 4.5. Summary of lessons and good practices for urban As with grant design, it is important to develop a performance performance grant programs measurement and assessment process that is rigorous, credible, and This review has provided evidence and lessons learned on key design objective enough to build trust among stakeholders and provide features and implementation drivers of UPG programs, along with adequate incentive to local governments. The APA is an effective tool for innovations to address increasingly complex urban challenges. These enhancing general institutional performances, and it contributes as well lessons are also broadly applicable to all other types of PBGs targeting to the World Bank’s corporate commitments, including environmental local and other subnational governments. Successful UPG programs and social (E&S) safeguards and climate actions. A well-designed APA combine strong financial incentives provided through adequate grant can be made objective by contracting it out to an external entity, credible design parameters with operationalization modalities (particularly by using a two-layered verification system, and rigorous by selecting APA systems) that are rigorous, credible, transparent, and fully adequate and targeted performance indicators. While such systems integrated with the government’s budget cycle and systems and with often require extensive handholding support to local governments the coordinated delivery of capacity- building support and technical participating in a UPG program, they have been effective in providing assistance to local governments to improve performance. adequate incentives to them and in leading to a gradual scaling up of the program’s impacts. To motivate the governments to complete the APA, First among the lessons learned is that ensuring UPGs are substantial the performance indicators selected need to be specific, measurable, in size is vital to provide local governments with adequate incentive achievable, realistic, and time-bound. Also essential to a good APA to achieve performance targets and make infrastructure investments is to ensure predictability and timeliness in the announcement and effective. Based on global experiences with UPGs, the per capita size disbursement of grants by synchronizing the assessment cycle with the of grants for urban local governments should typically be not less than project launch and the IGFT annual budget cycle in the country. about US$5–10;25 should enable infrastructure investments of at least US$150,000–200,000 per project; and should represent more than 20 Lastly, capacity-building support is central to strengthening percent of the total development grants allocated to the participating local institutional performances and improving urban infrastructure and governments. Among the external contexts that often caused the UPGs service delivery. Well-designed capacity-building support targets to be smaller in the reviewed countries were policies pursuing equitable specific performance areas for improvement that present institutional access to grants for all local governments, irrespective of the total amount barriers for local governments and helps low-performing ones climb of funds available. In such cases, determining the optimal grant size in the performance ladder. Most of the UPG programs reviewed adopted accordance with program objectives and local contexts requires clearly a balanced capacity-building strategy, combining supply- and demand- communicating to stakeholders the consequences of making the grants driven activities through a separate funding window. Another important small. If the grants are, indeed, considered too small to create meaningful element of effective capacity-building support is integrating the results impacts, adopting innovative mitigating design measures can help. These of these activities into the performance measurement system to enable could include, for example, the application of exponential allocation the results to form a clear basis for grant allocations and the local formulas; a stronger focus on awareness raising and publication; and governments’ capacity-building plans in subsequent years. acknowledgment of well-performing urban local governments. 25 There will be exceptions to this, however, especially where local governments are not in charge of major urban infrastructure projects, where the system is highly centralized in terms of funding, and/or where the governments have rural and less cost-intensive functions, as well. Performance-Based Fiscal Transfers for Urban Local Governments 62 Part I Results and Lessons from Two Decades of World Bank Financing 5.1. Ensuring the sustainability of UPG programs within 5. The Way Forward and Emerging government systems Issues for Urban Performance Grant A key question for UPG programs is how to sustain their impacts Programs beyond the period of financing provided by development partners and embed the grant systems into the intergovernmental fiscal framework. Sustainability is fundamental to a successful decentralization process This concluding chapter of Part I discusses several issues that will as well as to institutional improvement and the empowerment of local likely affect the design and implementation of the next generation governments in the long run. It is an especially important issue for of UPG programs. Building on the rich lessons and implementation development programs of this kind because their ultimate objective is experiences documented in previous chapters, it reviews ongoing to introduce durable changes in government systems. Thus, the real test discussions and provides recommendations on several emerging issues. of their success is whether the interventions they introduce are retained It starts by discussing the sustainability experiences of UPG programs once they close. This question can be answered by looking at three within country systems and how sustainability can be improved through interrelated dimensions on which their sustainability can be measured: long-term engagements that incrementally embed UPG programs into intergovernmental transfer (IGFT) systems. It discusses how the • Whether the UPG system implemented is retained and continues to next generation of UPG programs can gradually move from providing be financed by the higher levels of governments in some form after incentives for achieving results in institutional processes to improving the program closes service delivery outcomes, the challenges associated with this transition, • Whether the institutional improvements achieved under the and possible approaches to addressing these challenges in the context program—both at the level of individual local governments and of of multisectoral as well as sector-specific UPG programs. Finally, it the overall local government system in the country—are maintained explores how UPG programs can promote local climate change action • Whether the program-financed capital investments made by the by providing financial incentives to local governments. local governments are maintained in subsequent years Some evidence is emerging of the sustainability of UPG interventions within government systems after program close, but with caveats. Several programs financed by the Bank and other development partners have not closed yet and are continuing in their next iterations (see Figure 2.5, above, for examples of long-term engagements). In most of the cases reviewed, the governments retained, replicated, or even scaled up some features of the UPG programs after they had closed. Specific aspects of the programs, including processes for the institutional strengthening of local governments, also became part of the countries’ IGFT systems. Beyond the reviewed cases, other prominent examples exist of governments successfully sustaining performance-based financing for local governments. One is the government of Bangladesh (GoB), with which the Bank has been working for more than fifteen years to Performance-Based Fiscal Transfers for Urban Local Governments 63 Part I Results and Lessons from Two Decades of World Bank Financing implement a PBG system for the lowest tier of local governments in the country (targeting over 4,500 rural local governments), which has Box 5.1. Building Sustainability through a Long-Term Engagement in now been embedded into government systems as an intergovernmental Bangladesh fiscal transfer. The successive programs incrementally increased in The successive programs in Bangladesh consistently known since 2006 as the “Local sophistication over this period, and several reforms were formally Governance Support Project” (LGSP) targeted union parishads, the lowest tier of local governments in rural areas of the country. Part of a PBG system that the World Bank worked adopted by GoB. Grants to local governments, for example, are now being with the government of Bangladesh (GoB) for the past decade and a half to implement, the fully funded from GoB resources rather than external partner financing, programs incrementally increased in sophistication over this period. The union parishads and these allocations are on-budget and reflected in the GoB medium- would receive block grant funds based on satisfactory annual financial audit reports (often term budgetary framework. This case emphasizes the importance of undertaken by auditors hired from the private sector to supplement public sector capacity), with more PBGs provided as a top-up to the block grants upon the local governments’ a long-term engagement in sustaining the impacts of such programs, meeting additional performance criteria. The third and most recent iteration of LGSP began with decentralization and the institutional strengthening of local in 2017/18, with the objective of ensuring sustainability by institutionalizing program reforms governments requiring a long-term commitment by the government and interventions within the government systems for supporting rural local governments. and its development partners. (See Box 5.1 for more details on this long- Several of these reforms were formally adopted by GoB, including as follows: term engagement in Bangladesh.) Another example is provided by the • Block grants to the union parishads are now fully funded from the GoB budget rather than Solomon Islands, where the central government continued the PBG other sources. GoB’s share of financing for block grants has consistently increased over system for several years once development partner financing lapsed, time, from 50 percent in FY2011/12, to 60 percent in FY2016/17, to full funding now. before such financing was subsequently reintroduced. • GoB allocations to these grants are on-budget and reflected in the relevant ministries’ three-year medium-term budgetary frameworks and their annual development plans. • GoB has developed and officially adopted guidelines for “Union Parishad Development Assistance” as its framework for program management for PBGs to rural local governments, incorporating most LGSP interventions, including, among others, annual financial audits, grant allocations based on satisfactory audits, and direct disbursements to local government treasury accounts instead of disbursement through intermediary agencies. These actions have ensured that the successive set of interventions implemented over the past decade and a half under LGSP will be sustained, as they are now integrated into GoB’s programmatic support for the lowest tier of local governments. This emphasizes the importance of a long-term engagement in this area, with decentralization requiring a long-term commitment by both the government and its development partners to bring about a gradual transformation in long-held governance arrangements. Stakeholders in the country have gradually come to see LGSP reforms as regular practice, with union parishads officials now expecting grants to be disbursed from GoB based on LGSP procedures and ministry officials now expecting the union parishads will be audited annually and satisfactory audits will trigger grant allocations. One important factor for success has been the scale of the program: since inception it has been nationwide in scope, covering the entire universe of over 4,500 union parishads in the country. To that extent, LGSP has become the national “norm” for union parishad grants and grant administration. Sources: World Bank task team and World Bank IEG (2018). Performance-Based Fiscal Transfers for Urban Local Governments 64 Part I Results and Lessons from Two Decades of World Bank Financing An indicator of strong government commitment to a UPG program Table 5.1. Share of Government Co-Financing of UPG is its share of co-financing in the program’s total cost, which can hint Programs in Selected Countries at continuing government financing once the program concludes. While seven of the nine UPG programs reviewed for this report (some Country (UPG program) Government co-financing share of which are still under implementation) were co-financed by central Ethiopia (UIIDP) 30% governments, the share of this co-financing varied greatly, from over Ethiopia (UIIDP) 32% two-thirds of total program cost (in India-West Bengal and in Ghana in its Ghana (LGCSP) 0% second program) to less than 10 percent (in Uganda) to none (in Tanzania Ghana (GSCSP) 62% and the first Ghana program; see Table 5.1). Despite the high share of co- India-West Bengal state (IGSPP-II) 65% financing and the gradual expansion and scaling-up of several of these Kenya (KDSP) 34% programs, however, in some cases they were not sustained. In Tanzania, Tanzania (ULGSP) 0% for example, the government’s broader recentralization initiatives, such Tunisia (PDUGL and AF) 17% and 0% as the centralizing of property tax collection and the reduction of fiscal Uganda (USMID and USMID AF) 6% and 11% and operational support to local governments, diminished the scope of the local governments’ mandates nationwide and discouraged continued Source: Team analysis based on data provided by World Bank task teams for respective programs. efforts to improve fiscal systems at the local level. The UPG program Note: The government co-financing share represents the co-financing from the recipient country was substantially curtailed following the closure of Bank financing in government for the total cost of the World Bank–financed program over the full duration of the the country. It is worth noting that, in the future, governments in other program. In some cases (for example, in Ethiopia), the co-financing was also contributed by the participating local governments. countries may also choose not to finance UPG programs with their own resources once the externally financed program closes. Further evidence will emerge in coming years. Strong evidence has been emerging on how UPG programs have contributed to policy dialogues on fiscal decentralization and strengthened local institutions, and governments have continued or expanded these improvements to institutional systems. In Kenya, for example, the government showed consistent commitment to promoting devolution during the implementation of KDSP, the first large-scale devolution support program in the country. Kenya’s government cascaded lessons from KDSP to other World Bank–supported projects promoting devolved services, including, among others, the Kenya Urban Support Program, the Climate Smart Agriculture Project, and the Water and Sanitation Improvement Project. The ministry mandated with promoting devolution advanced this agenda among development partners, other ministries, and county governments.26 26 See World Bank 2022a. Performance-Based Fiscal Transfers for Urban Local Governments 65 Part I Results and Lessons from Two Decades of World Bank Financing In another example of a country’s continuing or expanding In sum, the evidence emerging from these examples shows that improvements to institutional systems resulting in part from UPG institutional improvements brought by UPG programs have led to programs, the central government in India is exploring the introduction the development of needed systems and capacities from which local at the national level of performance-based metrics for the provision governments have continued to benefit beyond the programs’ end. of grants to urban local governments, aligned with the principles Central governments have generally continued with and expanded of UPGs discussed in this report. The 15th Finance Commission, the improved institutional systems. That said, in some cases the APAs which the government of India established in 2017 to advise on the being implemented by UPG programs may not continue to take place devolution of taxes to states and local governments and other fiscal after those programs close, and complete evaluations in these cases matters, recommended linking the devolution of fiscal transfers to local may no longer be possible without specific performance benchmarks governments to verified improvements in OSR metrics and fiduciary and improvement measures provided at the level of individual local measures, such as the conduct and disclosure of independent audits governments. of annual accounts. The central government is considering adopting The strong results of the UPG programs reviewed have, however, these recommendations as eligibility criteria for fiscal transfers and as had a clear demonstration effect, leading to their considerable possible entry-level conditions for local governments in future flagship geographical expansion within these countries over time and urban development programs. At the state level, the UPG system indicating strong government interest in their outcomes. In Ethiopia, has been embedded in West Bengal’s IGFT scheme—the State Fiscal the ongoing UPG program (UIIDP) covered 117 local governments at the Commission (SFC) grants—and expanded to the entire state, leading time of this review, up from 18 under the first program (ULGDP I) and to institutional improvements in M&E systems, grievance handling 44 under the second (ULGDP II). The second phase of the program in mechanisms, and accounting and PFM systems. India-West Bengal (ISGPP II) covered all of the more than 3,000 local In Uganda, the UPG program (USMID) was linked with and affected governments in West Bengal state, 30 percent of which had been part the countrywide APA system for local governments and several other of the first phase since 2011. In Ghana, the first program (LGCSP) related reforms, including VfM concepts adopted by the Office of targeted 46 large urban local governments, while the second shifted its Auditor General, which are now used in multiple other programs; focus to 25 secondary cities, complemented by a general countrywide and, in Ethiopia, the UPG system developed under two consecutive performance-based grants system. programs was increasingly replicated across nonparticipating regions in The impact of UPG programs can likely also continue to be the country and linked with a countrywide rollout of financial and E&S increased and sustained by actions falling into two broad categories: audits for local governments and at the regional level. Ethiopia’s central those focusing on improving existing government systems for local government was also exploring how top-performing local governments, governments and those ensuring strong intergovernmental coordination which were showing strong evidence of sustained institutional and incentives for all stakeholders involved in the programs. Figure 5.1 improvement, could either be phased out of the program or made provides specific areas of action under these categories. subject to a more sophisticated set of performance metrics for getting access to UPG funding, while a new set of local governments that had previously not participated in the program could be brought into it. Performance-Based Fiscal Transfers for Urban Local Governments 66 Part I Results and Lessons from Two Decades of World Bank Financing Figure 5.1. Increasing and Sustaining Impact of UPG Programs • Promote interministerial comamittees to play A focus on improving existing systems a conducive role for checks and balances and reconciliation of results. • Provide support to long-lasting systems, • Ensure critical agents have incentives to such as PFM, procurement systems, M&E, promote a conducive environment for the local and grievance handling, that will persist after Sustainability governments. project closure. • Ensure wide sharing of results from APAs and • Align to the extent possible with existing other program features. IGFTs and support strengthening of this. Strong coordination and incentives In conclusion, World Bank support to UPG-based intergovernmental 5.1. From institutional processes to service delivery fiscal transfers should be considered in terms of long-term outcomes: The next generation of programs partnerships with governments, potentially providing them with the The next-generation UPG programs will need to continue taking heed fiscal space for an evolving and incrementally improving UPG system of lessons learned in the first generation and to innovate further as new over a 15- to 20-year time horizon. This extended involvement is more issues emerge that affect their design and implementation. More direct likely to move the needle toward improved institutional performance incentives are needed, for instance, for improving service delivery that enhances urban service delivery. Without continued support, outcomes, in addition to improving institutional performance, at the institutional gains in smaller and weaker local governments can local level. UPG programs also must determine how to promote local diminish, given their limited resources and capacity and, in many cases, climate change action through financial incentives to local governments. high staff turnover. Evidence shows that local government systems Sections 5.2 and 5.3, respectively, discuss these issues. seem to respond relatively quickly to incentives if those incentives are combined with capacity-building support and sustained. Such support and incentives will need to be sharpened toward service delivery outcomes, as discussed in sections 5.2 and, especially, 5.3, in more detail. Performance-Based Fiscal Transfers for Urban Local Governments 67 Part I Results and Lessons from Two Decades of World Bank Financing 5.2.1. Improved institutional processes as a channel for better Evidence is emerging that among the important drivers for better service delivery outcomes service delivery are improved institutional processes for public Most UPG programs aim to resolve institutional bottlenecks for service investment management (PIM). A recent study by the International delivery and provide incentives to improve local governments’ core Monetary Fund (IMF) has linked the strengthening of PIM institutions processes and systems, especially in contexts with very low capacity. and processes with the efficiency of public investment. It finds that These core processes have typically included capital investment planning countries with stronger PIM processes27 have more efficient public and execution, PFM, procurement, OSR, citizen engagement, municipal investments even after controlling for their per capita income levels. asset management, environmental and social systems, HR management, A comparison between the value of public investment capital (input) and so on—the basic building blocks of local government performance, and measures of infrastructure coverage and quality (output) across a accountability, and effectiveness in delivering services. The programs’ sample of 25 countries revealed an “efficiency gap” in public investment focus on improving these processes also reflects the contexts in which of around 30 percent for the average country28 and found that they have been implemented, where local governments are generally strengthening PIM practices (in this case, adopting the practices of the characterized by institutional weaknesses that need to be addressed best-performing country in the sample) could reduce this efficiency before more advanced interventions can be introduced. The approach gap by around two-thirds in the average country. The largest payoffs also aims to improve incrementally the capabilities of the targeted local occurred in countries emerging markets and low-income countries.29 governments. Section 3.2 of this report has also shown positive results of institutional improvements and service delivery outcomes occurring in the various Ultimately, the goal of UPG programs is improved service delivery, UPG country cases. VfM audits conducted in Uganda, for example, pursued through an instrument that recognizes the low-capacity and showed that UPG-funded capital infrastructure investments were governance-challenged institutional environments in which they cost effective and of higher quality than similar investments in the operate. In a highly capable municipality in a higher-income country, country that were not funded by the program. They also noted that an explicit focus on institutional strengthening to improve service most UPG-financed capital investments were executed in accordance delivery may not be necessary, as the baseline performance may already with procurement and PFM guidelines, with costs within budgets and be strong. Instead, the grant can focus on providing direct incentives minimal defects. Such findings indicate a high degree of efficiency of for improvements, as PBGs in developed countries—such as the “No public investment expenditure for local service delivery. Child Left Behind” program in the United States—have often done. But not having core systems in place while focusing solely on service delivery outputs or outcomes may result in poor sustainability of these interventions, or they may get delivered in a highly inefficient manner. In short, improving the institutional foundations of local institutions, if not a sufficient condition for improved local service delivery, is certainly 27 As measured by the IMF’s Public Investment Management Assessment, which evaluates a necessary one. processes for the planning, allocation, and implementation of public investments (IMF 2015). 28 This efficiency gap is measured as the distance between the average country and a “public investment efficiency frontier” (calculated by the authors) for a given level of public capital stock and GDP per capita (IMF 2015). 29 These findings come with two important caveats. First, the sample is relatively small (only 25 countries), introducing a degree of uncertainty as to whether the results also hold for a larger set of countries. Second, because a robust causal relationship is difficult to establish, only correlations between institutional processes and service delivery are reported. Performance-Based Fiscal Transfers for Urban Local Governments 68 Part I Results and Lessons from Two Decades of World Bank Financing Given this level of success, stakeholders have continued to demand or output of roads developed by one local government with improved direct fiscal incentives for and measurement of improvements in solid waste management (SWM) developed by another, as various local service delivery outcomes at the local level and the tying of UPG governments may have different expenditure priorities based on their funding to demonstrated service delivery improvements. Funding needs. Thus, the assessment process in such cases becomes fraught with may, for example, be tied to specific results achieved in municipal water practical difficulties. and sanitation services, transportation, or other areas. The emergence Besides such obstacles to comparability, several other challenges to of this demand has been especially notable in long-running programs direct assessment of performance on service delivery outcomes can where local governments have received UPG funding for several pose a risk of distorting incentives for local governments. They are as years and where stakeholders expect the governments’ strengthened follows: institutional performance to translate into better delivery of services. As a result, UPG programs may start focusing fiscal incentives more • Equity among local governments and control over achieving directly on service delivery outcomes, and moving in this direction results: Local governments often claim that achieving service delivery outcomes may not be under their full control but may be where possible is certainly necessary. Doing so presents important influenced by such external factors as support from line ministries or limitations, however, that merit discussion. the level of resources available. If not carefully selected, indicators 5.2.2. Challenges to providing direct fiscal incentives for local for assessment could penalize those local governments whose prior service delivery resources and capacity are lower than those of well-resourced local governments that can secure grant allocations without significant Fiscal incentives to local governments based on direct measurement effort. Some local governments or areas may also require more inputs of service delivery outcomes may be suitable for sector-specific UPG to achieve similar results due to various contextual factors, such as programs, where funds are tied to use in specific predefined sectors. geography, levels of poverty, and so on. These may include programs dedicated to water and sanitation, roads, • Lack of flexibility and discretion: In the case of multisectoral solid waste management, public health, education, or other such results- programs, performance measurement based on outcome indicators based financing; see again the typology of PBG programs in figure 2.1. may tend to distort the incentives of the program by rewarding sector- A predefined list of sectors (or eligible expenditure categories) enables specific, short-term investments rather than longer-term investments comparison of performances and results across local governments, made in a coordinated manner across the local government. This is which then allows the determination of grant allocations. The number mainly because it is difficult to incorporate different characteristics of water supply connections provided by various local governments of various sectors (in terms of costs, timeframe, and capacities or utilities can be compared, for example, if the program provides required) into one common assessment system. The difficulties can, incentive for this result. When local governments have discretion to in turn, limit the flexibility of local governments in multisectoral spend funds across multiple sectors within their remits, however (as grants in which they are meant to have discretion over the selection is generally the case in most UPG programs that, by design, provide of investments, and they end up being guided (or indirectly forced) to local governments with this discretion), and where external factors focus more on sectors where spending is easier to plan and execute. influence results, comparing performance across local governments • Time lag between input and results: Service delivery outcomes, in one particular sector over another is generally not feasible, as the especially those pertaining to investments in capital infrastructure, sectors different local governments will choose for investment is not may not be achieved in time for the annual assessment and budgeting cycle within which most PBG programs operate. Investments financed known a priori. It is not feasible, for instance, to compare the quality Performance-Based Fiscal Transfers for Urban Local Governments 69 Part I Results and Lessons from Two Decades of World Bank Financing by a local government, for example, may not be complete by the Assessing the quality and timely completion of projects is a strong next round of annual assessment simply because they are large and metric, as it ensures the local government has effectively spent funds bulky. In such cases, it is important to find alternative performance within its remit for local service delivery. It improves the investment indicators, especially for the first few years of the program, when planning and expenditure capacity of the governments, which, in turn, investments are not finished. improves their effectiveness, even beyond the UPG program. • Feasibility of results assessment: Also extremely challenging is the undertaking of annual surveys of outcome indicators across all Such indicators generally provided strong results in the UPG participating local governments, because such surveys generally programs reviewed, as noted in section 3.2, which highlights results require a complex and expensive assessment process with significant from various VfM audits that show investments in urban infrastructure fieldwork on a regular basis. This may lead to delays in the assessment were well executed across these programs. Notably, while the programs process and increase associated costs, which is already an issue in generally had a large set of institutional and process indicators on which some countries. local governments were assessed and became eligible to receive funds, • Tradeoff between broad institutional system strengthening and the funds were expended on local service delivery and infrastructure, achieving specific outputs: UPG programs have focused strongly on which was demonstrably carried out economically and efficiently, with building the longer-term capacity and institutional performance of high-quality results. local governments as entities and corporate bodies with the capacity The use of well-designed VfM audits (and similar performance to spend larger sums of public funding and provide the systems and procedures to do this efficiently. A focus on short-term, specific metrics) to inform the allocation of grants can be valuable in outputs may dilute this core objective. focusing the incentives of local governments on the timely and efficient delivery of infrastructure. These audits should measure such 5.2.3. Potential approaches going forward parameters as economy of expenditure (that is, cost, especially relative to How can better incentives be provided to improve and measure other similar investments), efficiency (such as expenditure execution rates outcomes in local service delivery? Several UPG programs are and timeliness), and effectiveness (whether the objectives were achieved overcoming this challenge—especially in multisectoral programs —by and whether the services are delivered as planned and being utilized by using proxy-type indicators to measure the efficiency and quality of end-beneficiaries); see Box 3.3, above, for a good-practice example of expenditure from grant funds. These include, for example, results from their use. Efficiency is measured in terms of the use of resources by local VfM audits of grant funds and capital expenditure execution rates of governments, with results that can be readily compared across them infrastructure investments. The use of such indicators to measure and utilized to inform future grant allocations. The audits also focus on improved service delivery outcomes is especially useful where funds are the actual quality and end-use of the infrastructure delivered. utilized for multisectoral interventions, which is common where local governments have broad mandates, because they measure the efficiency, In contexts where local governments have built strong processes, effectiveness, economies, and quality of the governments’ capital UPG programs can aim to move them up the value chain with more expenditure. In multisectoral UPG programs where local governments sophisticated metrics or by targeting process and output indicators are not limited to spending grant funding in a specific sector but instead for specific municipal services or sectors. Entire programs can be can prioritize spending across their remits, it is important to provide designed that target sector-specific institutional performance factors incentive for and measure the efficient and timely use of the funds closer to the point of service delivery and alleviate critical constraints across areas—an objective for which proxy indicators are well suited. on delivery outcomes. Several promising recent examples include Performance-Based Fiscal Transfers for Urban Local Governments 70 Part I Results and Lessons from Two Decades of World Bank Financing providing incentives for and promoting local climate change action through UPGs (see next section); a new Bank-financed UPG program Box 5.2. Recent Sector-Specific UPGs in Kampala, Uganda, and supporting metropolitan-level coordination (and increased access to Kerala, India infrastructure and services) targeting all local governments in Uganda’s Recent examples of sector-specific UPG programs include the Greater Kampala Greater Kampala metropolitan region; and another recent Bank- Metropolitan Area Urban Development Program (Uganda) and the Kerala Solid Waste Management Project (India). financed project to improve municipal SWM services by providing UPGs The goal of the Greater Kampala Metropolitan Area Urban Development Program for improved SWM-related metrics to all urban local governments in is to improve institutional capacity for metropolitan coordination and management and the Indian state of Kerala (see Box 5.2 for a summary of these examples, increase access to improved infrastructure and services. It will provide performance- which can be replicated in other contexts to achieve specific sectoral based grants to all nine urban local governments in the metropolitan region, using a objectives). mixture of performance measures (PMs, or eligibility criteria) to provide grants based on achieving targets for institutional processes, as well as on the completion of specific Finally, it is important to acknowledge that in many contexts where types of investments. Under one PM, for example, a certain share of investments made baseline capacity and institutional performance are weak, due time by all local governments in the metropolitan area in a preceding year need to have been “multijurisdictional” in nature—that is, coordinated and co-implemented by various local is needed to bring local governments up to a level where investments governments in the area—to promote the integration of metropolitan-level planning can be sustainably made. Service delivery outcomes likely will not and the delivery of infrastructure investments, thus addressing a common problem appreciably improve in the early years of such programs, and a longer- in the management of metropolitan areas. In another, a certain share of investments term commitment is needed to observe meaningful change. This is made by each local government in the metropolitan area need to be for climate- clearly an evolving topic and requires continued exploration to capture resilient infrastructure that is completed and to contribute to climate change mitigation and adaptation (for example, through the rehabilitation of stormwater drainage; the emerging trends and lessons. preparation of SWM strategies and the promotion of waste segregation and sorting that help reduce drainage clogging; and so on). The Kerala Solid Waste Management Project covers the entire universe of urban local governments in the state of Kerala, India, providing UPG financing to them upon the achievement of metrics specifically targeted toward building their institutional systems for planning, implementing, and managing SWM infrastructure and services. These metrics include key institutional results and technical readiness activities pertaining to the SWM sector, identified based on technical assessments during project preparation. A key feature of this project—representing a departure from other UPG programs—is that the eligibility criteria for grants comprise one-time activities to be completed by local governments, rather than full annual assessments. This enables the governments to confirm up front their entire grant allocation ceilings for the whole duration of the project, which, in turn, allows them to undertake multiyear investment planning to prepare citywide five-year SWM investment plans, to be gradually implemented over the project duration using project funds. The actual disbursement of grants to local governments still follows an annual cycle—in line with the government annual budgeting and planning cycle—based on the achievement of annual triggers to ensure continued compliance. Sources: World Bank 2021b; 2022b. Performance-Based Fiscal Transfers for Urban Local Governments 71 Part I Results and Lessons from Two Decades of World Bank Financing 5.3. Promoting climate change action at the local level The Bank has also started piloting climate-focused UPG programs through UPGs in some countries. Among the countries where Bank-financed UPG Attention is increasing on promoting local climate actions through programs were reviewed, Ethiopia was found to incorporate urban financial incentives under UPG systems. In many cases, the United climate resilience indicators in its annual performance assessment. More Nations Capital Development Fund (UNCDF) has been taking the lead significantly, a new PBG program in Kenya—the Financing Locally-Led in this area through its Local Climate Adaptive Living Facility (LoCAL), Climate Action Program (World Bank 2021a)—will support community- which provides performance-based climate resilience grants in several identified and locally led climate resilience projects at the county level countries.30 LoCAL increases awareness and capacities to respond to and strengthen their climate finance and governance systems, with the climate change at the local level, integrating climate change adaptation grants providing the performance incentive. Emerging experiences into government planning and budgeting systems in a participatory in such programs is showing that local governments are responding manner. A targeted cross-sectoral grant, it attaches conditions to to financial incentives in this area, pointing to a need for further use the use of its funding for climate change adaptation and has a menu of these types of programs. Climate change action will become a defining eligible investments for local governments. Climate-related substantial area of focus. performance metrics used under the program include the use of climate While most climate-focused UPG programs have generally focused information (local data on climate change, climate risk assessments, on climate change adaptation, incentives are also needed for local and vulnerability assessments); the mainstreaming of adaptation in actions for climate change mitigation. This can be an area of focus local planning, budgeting, procurement and contracting, and execution; for the next generation of UPG programs. The key tool for providing compliance with technical specifications for incorporating climate- incentive to motivate mitigation actions under such programs is the related aspects in infrastructure; and the effective use of funds. It also menu of eligible capital investments available to local governments includes a set of indicators that promote inclusive local governance to for the use of grant funds. Capital investments that reduce greenhouse strengthen community resilience through participatory local planning gas emissions should be among these. They may include, for example, processes and the implementation of climate adaptation measures. investments to encourage switching from private to public (and/or The carefully defined menu of eligible investments encourages capital nonmotorized) modes of transportation; investments in SWM to reduce investment in local climate adaptation needs.31 methane emissions, especially from landfills; and investments in green Globally, PBG programs increasingly are being utilized to sharpen public spaces, among others. the focus on climate change mitigation and adaptation at the local level, either through standalone, performance-based climate resilience grants or integrated into existing multisectoral grants as top-up with increased focus on climate. These initiatives are linked with a targeted performance assessment system and capacity-building support for local governments for climate risk resilience. 30 LoCAL has applied this concept in 17 countries already, with almost as many new country programs being planned. It is generally developed as a supplement to the existing PBGs in a country (see UNCDF 2020). 31 For discussions on LoCAL program, see UNCDF 2020 and 2022. Performance-Based Fiscal Transfers for Urban Local Governments 72 Part I Results and Lessons from Two Decades of World Bank Financing References ———. 2020. Financing Local Adaptation to Climate Change: Experiences with performance-based climate resilience grants. New York: United Nations Capital Development Fund. Athar, Sohaib, Roland White, and Harsh Goyal. 2022. Financing India’s Urban Infrastructure Needs: Constraints to Commercial Financing and ———. 2021. Mobilizing Capital for Local Development Funds: A Longitudinal Prospects for Policy Action. Washington, DC: World Bank. Case Study in Leveraging Millions to Billions. New York: United Nations Capital Development Fund. Erman, Alvina, Carla Solis Uehara, and Chloé Beaudet. 2021. “Leveling Up: Impacts of Performance-Based Grants on Municipal Revenue ———. 2022. Local Government Finance Is Development Finance. New York: Collection in Mozambique.” Policy Research Working Paper 9789, United Nations Capital Development Fund. World Bank, Washington, DC.. World Bank. 2020a. “Disaster Risk Management.” World Bank IBRD/IDA, Government of Uganda. 2019. Consolidated Value for Money Audit Report on May 2. https://www.worldbank.org/en/topic/disasterriskmanagement/ Selected Urban Infrastructure Projects Implemented by the 18 Uganda Support overview. to Municipal Infrastructure Development Program (USMID) Participating ———. 2020b. “Land.” World Bank IBRD/IDA, April 16. https://www. Municipal Councils in the FY 2017/2018. Kampala: Government of Uganda worldbank.org/en/topic/land#1. IMF (International Monetary Fund). 2015. “Making Public Investment ———. 2020c. “Urban Development.” World Bank IBRD/IDA, April 20. More Efficient.” IMF Policy Paper. https://www.worldbank.org/en/topic/urbandevelopment/overview. Kayode, Taiwo. 2020. “Intergovernmental Transfers and Own Source ———. 2021a. “New US$150 Million Program to Strengthen Kenya’s Revenues of Sub-National Governments in Nigeria.” Munich Personal Resilience to Climate Change.” Press release, October 26. https://www. RePEc Archive Papers, September 23, 1–31. worldbank.org/en/news/press-release/2021/10/26/new-us-150-million- Mahler, Water. 2005. Options for Financing Local Governments in Uganda program-to-strengthen-kenya-s-resilience-to-climate-change. Context. Durham: Duke Center for International Development, Sanford ———. 2021b. Project Appraisal Document for Kerala Solid Waste Management Institute of Public Policy, Duke University. Project. Washington, DC: World Bank. Masaki, Takaaki. 2018. “The Impact of Intergovernmental Transfers on ———. 2022a. Kenya Devolution Support Project: Implementation Completion local Revenue Generation in Sub-Saharan Africa.” World Development & Results Report. Washington, DC: World Bank. 106 (C): 173–86. ———. 2022b. Program Appraisal Document for a Greater Kampala Salhab, Jade, Diana Hristova, Sohaib Athar, and Ayah Mahgoub. 2022. Metropolitan Area Urban Development Program. Washington, DC: World Subnational Competitiveness Grants Guidebook : A Tool to Promote Jobs and Bank. Economic Transformation in Cities and Regions through Performance-Based Financing. Washington, DC: World Bank. World Bank IEG. 2018. Independent Evaluation Group’s Review of the Implementation Completion and Results Report for Bangladesh: Local UNCDF (United Nations Capital Development Fund). 2010. Performance- Governance Support Project II. Washington, DC: World Bank. Based Grant Systems: Concept and International Experience. New York: United Nations Capital Development Fund. Performance-Based Fiscal Transfers for Urban Local Governments 73 Part II Results and Lessons from Two Decades of World Bank Financing Part II Case Studies from Selected Countries Performance-Based Fiscal Transfers for Urban Local Governments 74 Part II Results and Lessons from Two Decades of World Bank Financing Part II will showcase key findings from the in-depth case studies of nine UPG programs conducted for this report. The reviewed programs cover seven countries across four regions: 6. Ethiopia: Second Urban Local Government Development Program (ULGDP II) and Urban Institutional and Infrastructure Development Program (UIIDP) 7. Ghana: Local Government Capacity Support Project (LGCSP) and Ghana Secondary Cities Support Program (GSCSP) 8. India: West Bengal Institutional Strengthening of Gram Panchayats Program (ISGPP II) 9. Kenya: Kenya Devolution Support Program (KDSP) 10. Tanzania: Urban Local Government Strengthening Program (ULGSP) 11. Tunisia: Urban Development and Local Governance and AF (PDUGL) 12. Uganda: Uganda Support to Municipal Infrastructure Development Program (USMID and USMID AF) These programs represent the standard typology of a UPG program, such as targeting multiple municipalities, conducive environments in countries (that is, decentralization initiative), and refined grant designs drawing upon accumulated experiences. Each case was analyzed based on desk reviews of key project documents32 and virtual internal interviews with respective World Bank task teams. 32 See section 1.4 for the detailed list of key project documents reviewed. Performance-Based Fiscal Transfers for Urban Local Governments 75 Part II Results and Lessons from Two Decades of World Bank Financing 6. Ethiopia: Second Urban Local Table 6.1. Factsheet for the Case UPG Programs in Ethiopia Government Development Program and Urban Institutional and The second Urban Local Urban Institutional and Program title Government Development Infrastructure Program (UIIDP) Program (ULGDP II) Infrastructure Program Financing PforR Hybrid 1. Background instrument Ethiopia is among the countries that have successfully scaled up UPG IDA: US$600 million programs across two generations. Based on lessons learned from Phase Financing IDA: US$380 million Borrower: US$248.7 million amount I of the Urban Local Government Development Program (ULGDP I),33 (Financier) Borrower: US$176 million Agence Française de the program was rolled out to 44 large urban local governments under Développement: US$10.8 million ULGDP II and, subsequently, to 117 medium and smaller ones under Duration UIIDP. The three consecutive programs (ULGDP I, ULGDP II, and (Board approval May 2014 to December 2019 May 2018 to July 2023 UIIDP) made a significant contribution to shaping the various local to closing) development systems. Throughout them, the major components and Geographical principles of the UPG system were retained, with few refinements. To 44 urban local governments 117 urban local governments coverage meet the changing demands, however, different types of investment To assist the Recipient in instruments were applied, building upon the lessons learned from each To enhance Institutional enhancing the institutional phase: ULGDP I applied Investment Project Financing (IPF); ULGDP performance of participating Project performance of the urban local governments to II featured a Program for Results (PforR) operation; and UIIDP used development participating urban local develop and sustain urban objective governments in developing and a hybrid option to strengthen capacity-building systems through a infrastructure, services, and sustaining urban infrastructure economic development focused IPF window. The thematic focus areas were expanded, as well. and services UIIDP, for example, focused on emerging cross-cutting urban agendas, such as local economic development, climate resilience, and gender. Table 6.1 summarizes key facts of the two of Ethiopia’s UPG programs reviewed for this study. 33 The Urban Local Government Development Project (ULGDP) (2008–13), implemented by the Federal Government of Ethiopia Ministry of Urban Development, Housing, and Construction, was a US$208 million program financed by the World Bank to improve the institutional capacities of 18 urban local governments to meet their infrastructure requirements through discretionary funds provided as specific-purpose grants. The program was in line with the country’s efforts to meet infrastructure requirements for basic services in city administrations while empowering the local governments to undertake capital investments, build institutional systems, and take ownership of resource mobilization. Performance-Based Fiscal Transfers for Urban Local Governments 76 Part II Results and Lessons from Two Decades of World Bank Financing 2. Key performances Table 6.2. Key Performances of the Case UPG Programs in Ethiopia The review found the achievements of the second and third UPG programs in Ethiopia (ULGDP II and UIIDP) to be good overall, consistently exceeding the planned program targets. Under ULGDP Details (changes over the program Key performance areas Improvement period) II, compliance with the minimum conditions (MCs) was continuous throughout the program period, and the average performance measure General MC Consistently 100% of local governments met MCs (PM) score of all the participating local governments increased by 30 performance compliance met under through both programs. rate (%) both percent over the four years. Despite the rapid scaling-up process, all 117 local governments under UIIDP also managed to comply with the Average APA Increased from 71 to 90 (ULGDP II); MCs from the first year of implementation, with the average PM score performance Yes for both from 83 to 95 (UIIDP). score increasing by 6 percent from the first to the second annual performance assessment (APA). The positive results indicate the local governments Key urban OSR collected increased from variables US$33.6 million to US$49.8 million were able to respond fast and effectively to the financial incentives (48%) under ULGDP II; the number provided by the UPG programs. OSR Yes for both of local governments with more than 10% OSR improvement per year Institutional performance also significantly improved in the following continuously increased, from 20% to areas (Table 6.2): 76% throughout the two programs. • OSR. The share of the 44 local governments under ULGDP II with Human HR was a minimum condition under annual improvement of more than 10 percent in OSR increased from resource Yes for both both programs and was consistently 20 percent to 84 percent. Under UIIDP, 76 percent of the 117 local management met. governments had OSR exceeding 10 percent. The nominal amount of OSR collected also significantly increased, by 48 percent, under Share of local governments executing Asset ULGDP II. over 80% of the planned O&M budget management Yes for both increased from 20% to 80% (ULGDP II) • Human resource management. Measured as an MC under both / O&M and from 23% to 68% (UIIDP). programs, HR management met the minimum requirement consistently throughout implementation. PM score on PFM increased from 9 to 13 out of 15 maximum points; • Asset management. The local governments became capable of local governments with clean audits operation and maintenance (O&M) planning according to standards, increased from 11% to 57% under executing over 80 percent of the planned budget. Yes (ULGDP II) ULGDP II. UIIDP saw a decrease from PFM 57% to 49% but mainly because in the • Clean auditing. Progress for the audit status under the two programs NA (UIIDP) first year only 44 local governments was stable. While only 25 local governments had clean audits at the were assessed, while in the third beginning of UIIDP, after three years of implementation 66 out of the year all 117 local governments were 117 local governments did, and the rest had qualified audit reports. assessed, and the score reached 14 out of a possible 15. All the participating local governments conducted timely audits, which, together with other performance improvements in PFM, was a major achievement. Performance-Based Fiscal Transfers for Urban Local Governments 77 Part II Results and Lessons from Two Decades of World Bank Financing Implementation of infrastructure and service delivery showed continuous 3. Operational design progress over the ULGDP II and UIIDP program periods. The participating Grant design. The size of the grants under the two programs was local governments were able to execute infrastructure investments as substantial, at about US$19–20 per capita for both. The UPGs also planned in a timely and effective manner. The implementation ratio of constituted the only block (specific-purpose) grant from the federal planned local infrastructure investments was 72 percent in the first year of to local levels (city administrations), accounting for 70 percent of the ULGDP II, increasing to 94 percent in the last year, and it continued to be entire funding streams to urban local governments, on average. The very high during the rollout of the program under UIIDP, at 94 percent in grants were typically used for smaller and medium-size investments the second year. Most of the funds were spent effectively on investments with an average size of US$150,000, which was appropriate, given in infrastructure and service delivery—95 percent and 80 percent under the limited functional mandates of the urban local governments. ULGDP II and UIIDP, respectively. As a result, infrastructure outputs The per capita allocation formula, however, left smaller urban local from the two programs surpassed the targets, including 1,299 km of governments with relatively lower budget allocations, while the larger cobblestone roads (against the target of 600 km), and 620 km of drainage, ones tended to receive more funding. Some of the participating urban 192 bridges, and 284 ha of parks, benefiting over 10 million people in all. local governments raised equity issues related to the allocation formula. The quality of the provided infrastructure was satisfactory, according to Value for Money (VfM) audit results. The programs also contributed Performance assessment system. The APA was designed based on to a high level of temporary employment creation, amounting to about overall lessons learned from the first-generation program (ULGDP I), field 160,000 person-years of employment under ULGDP II and 270,000 per testing of the local governments’ performance capacities, and detailed year under UIIDP. Given the pressing nature of the unemployment technical assessment covering a comprehensive range of performance problem in Ethiopia, labor-intensive techniques and job creation were areas. Performance indicators were relatively comprehensive, with a set to be a high priority in UPG programs in the country and needed to be of 50 to 60 indicators, covering the entire PFM cycle from planning to further integrated into the programs. auditing, as well as cross-cutting areas such as climate resilience, Local Economic Development, gender, and governance. It functioned well in addressing the performance of all tiers of government, from local governments and regions to the federal level, and informed targeted technical assistance and capacity-building support from the upper tiers. In addition, disbursement-linked indicators (DLIs) were embedded in the APA indicators for the improvement of institutional performances, such as PFM, OSR, safeguards, and capacity-building activities. Among the challenges at the upper levels were delays in the timely completion of APAs, but this did not significantly affect the actual budget execution at the local government level. Although the feasibility of continuing to conduct comprehensive APAs was sometimes questioned, the actual costs of the entire APA process were about only 1 percent of the total annual UPG. Performance-Based Fiscal Transfers for Urban Local Governments 78 Part II Results and Lessons from Two Decades of World Bank Financing Quality assurance. A strong system for quality enhancement review In response to a number of local governments having “plateaued” in rendered the APA results highly credible and provided adequate their improvements to the core institutional performance areas, the incentive for local governments to strive for continued high-level federal government discussed a potential exit strategy by which high- performances. Quality assurance teams, both from the government and performing local governments could opt to phase out from the UPG the World Bank, conducted field visits to a sample of about 15 percent programs. An alternative view was that those local governments could of the participating local governments to spot-check the accuracy of the go on to develop innovative ways to enhance their competitiveness APA results, and detailed desk reviews of the reports covered all the through continued and advanced support from the programs. local governments. Environmental and social management. The review found that the Capacity building. The capacity-building system was designed to target programs effectively addressed E&S issues through the incentives the identified capacity gaps of local governments, with a combination they provided in the performance measures, through the incentives of demand-driven support, incentives for support from the regions, in the DLIs on regional governments’ timely review of local safeguards and backstopping support from the federal mobile team. The ULGDP compliance, and through the incentives for central governments to II was a Program-for-Results (PforR) with one DLI on capacity-building provide targeted and comprehensive technical assistance and capacity- activities, while the UIIDP used an Investment Project Financing (IPF) building support to local governments on E&S issues. The programs also window for a selected set of these activities. While capacity-building significantly improved their grievance redressal management systems. activities at the local level were retained in the PforR modality under both programs, some of the central and regional activities were shifted from PforR (under ULGDP II) to IPF (under UIIDP) modalities to reduce the risks for undertaking federal-level actions, such as analytical works, assessments, the provision of additional teams for capacity-building activities, and others. Both programs allowed local governments to spend up to 5 percent of the UPGs on capacity-building activities. Sustainability. The review found the impacts of the UPG programs on institutional performances and system development highly likely to be sustained beyond the program period for three reasons: • Several nonparticipating regions had adopted similar UPG schemes for their own grants to local governments, suggesting the programs would have future impacts on the overall fiscal decentralization framework. • The development of and enhancements to institutions accomplished during the program were likely to be sustained after its closing. • A high level of co-funding from government indicated ownership and a willingness to continue and expand the approach. Performance-Based Fiscal Transfers for Urban Local Governments 79 Part II Results and Lessons from Two Decades of World Bank Financing 7. Ghana: Local Government Table 7.1. Factsheet for the Case UPG Programs in Ghana Capacity Support Project and Local Government Capacity Ghana Secondary Cities Ghana Secondary Cities Support Program title Support Project (LGCSP) Support Program (GSCSP) Program Financing instrument IPF PforR 1. Background Financing IDA: US$100 million In 2008, in line with the National Decentralization Policy Framework amount IDA: US$175 million (Financier) Borrower: US$161 million in Ghana, a countrywide performance-based grant system known as the District Development Fund (DDF) was started with support Duration from the government of Ghana and its development partners. When, (Board approval March 2011 to June 2018 September 2018 to June 2024 to closing) in 2011, the World Bank started its first UPG program in Ghana, the Local Government Capacity Support Project (LGCSP), it based the 25 local governments with total 46 local governments design of this IPF operation on the lessons learned from the DDF. The Geographical population of 100,000–250,000 (metropolitan and municipal coverage (secondary cities) and at least Ghana Secondary Cities Support Program (GSCSP), launched in 2018, assemblies) 60% urban population. further entrenched the UPG culture and strengthened the country’s intergovernmental fiscal transfer (IGFT) system. GSCSP, a Program-for- To strengthen the intergovernmental fiscal Results (PforR), strategically catered to the specific needs of secondary framework; to strengthen cities, with the objective of easing the development pressure on the local PFM and accountability To improve urban management primary cities. GSCSP focused more strongly than LGCSP on core Project for improved infrastructure and basic urban services development and services in urban urban performance improvement, offering substantially larger grants in participating municipal objective assemblies; and to improve assemblies. and investment projects. Table 7.1 shows key facts of the two UPG citizens’ engagement with programs reviewed here. urban assemblies and their perceptions of urban management. Performance-Based Fiscal Transfers for Urban Local Governments 80 Part II Results and Lessons from Two Decades of World Bank Financing 2. Key performances Table 7.2. Key Performances of the Case UPG Programs in Ghana Overall, both LGCSP and GSCSP were successful in achieving key performance areas. At the closing of LGCSP, the program exceeded the targets set at the design stage; the MC compliance rate increased Details (changes over the program Key performance areas Improvement period) by 37 percent and the average PM score by 19 percent over the project period. GSCSP was also on track to achieve its planned performance General MC Yes for both Increased from 65% to 89% under LGCSP performance compliance and was consistently 100% under GSCSP. targets, although its quantitative results were still limited, given that rate (%) program had been running for only two years at the time of the review. All the local governments under GSCSP were in full compliance with Average APA Yes for both Increased from 64 to 94 (LGCSP) and from MCs, and the average PM score increased by 10.7 percent from the first performance 57 to 63 (GSCSP). score to the second annual performance assessment (APA). The institutional performance significantly improved in the following key result areas Key urban OSR collected in 46 local governments variables increased from ¢75.8 billion in 2011 to ¢188 (Table 7.2): Yes for billion in 2017 (148%) under LGCSP; OSR OSR • OSR. The amount of OSR collected under LGCSP increased by 148 LGCSP collected in 25 local governments increased from ¢35.8 million to ¢43.6 million (21.8%) percent during the project period, and 91 percent of the participating between 2018 and 2020 under GSCSP. local governments had achieved annual increases of more than 5 percent in OSR collection at program closing. Share of local governments achieving social accountability performance indicator in • Asset management. Operation and maintenance were monitored as Yes for APAs increased from 76% to 100% under Accountability a key result area in the APAs; the average score on O&M increased by LGCSP LGSCP; this indicator was not monitored 14.8 percent over the project period of LGCSP and by 20 percent over under the GSCSP results framework but was embedded in the DPAT.a the two years of GSCSP implementation. • Timeliness of reporting and quality of auditing. Under LGCSP, the Average PM score on asset management increased by 15% (LGCSP); average PM share of local governments that submitted financial statements on Asset score on sustainable urban systems time and had clean audits reached 100 percent in both respects in management Yes for both (including indicators on property rate the fourth and fifth APAs. The APA score for PFM increased by 47 / O&M administration and enforcement and on percent. maintenance of road, drain, and pedestrian networks) increased by 19% (GSCSP). • Budget credibility and social accountability. Performance on budget credibility and social accountability improved significantly Average PM score on asset management under LGCSP. Almost all participating local governments disclosed increased by 15% (LGCSP); average PM score on sustainable urban systems annual fee-fixing resolutions during the project period, and all had Yes for PFM (including indicators on property rate institutionalized town hall meetings by program closure. LGCSP administration and enforcement and on maintenance of road, drain, and pedestrian While the task team believed GSCSP had also been effective in networks) increased by 19% (GSCSP). institutional performance enhancement at the time of the review, supportive quantitative evidence was as yet insufficient. This is mainly a. District Assembly Performance Assessment Tool. because, first, the program focused more on urban infrastructure Performance-Based Fiscal Transfers for Urban Local Governments 81 Part II Results and Lessons from Two Decades of World Bank Financing and service outcome indicators than on intermediate process and 3. Operational design system indicators; and, second, it was designed to avoid duplication of Grant design. To scale up the impacts of the UPGs, the grant design indicators, such as HR management, transparency and accountability, was enhanced across the two programs. The grants under LGCSP were and financial management, that were already being measured as part of relatively small, at US$4.56 per capita, on average. During the program the national performance measurement system. The program instead implementation, this was deemed inadequate for many of the investment incorporated the results of the national assessments into its MCs, which projects envisaged in annual local development plans. As a result, the requires local governments to obtain higher scores than the national- program gradually reduced the number of investment projects in its last level average performance score to access the grants. years of implementation to avoid fragmentation of the investments and While LGCSP invested in a fairly wide range of asset classes, to make the procurement processes more efficient. The lessons learned including education, health, and social facilities, the focus of GSCSP from LGCSP in this regard were reflected in the design of GSCSP, whose was narrowed down to core urban infrastructure for connectivity, grants were significantly larger, at approximately US$15 per capita. The stormwater drainage, and competitiveness. At the closing of LGCSP, grants were to be used for a selected set of urban assets and to target all the planned infrastructure had been delivered, including 120 km a larger scale of investment projects (over US$500,000). The technical of rehabilitated roads, 2,215 school classrooms, 69 markets, 4 health complexities of the larger investments, however, in addition to the clinics, 18 police stations, and 93 toilet facilities. In total, approximately impact of the COVID pandemic, led to some delays in project execution 6.6 million citizens benefited from the program, exceeding the target during the first years of implementation, limiting the infrastructure and of 2 million. The results of VfM audits showed good cost effectiveness. service delivery results found by this study. In addition, about 12,000 temporary employment apportunities were Performance assessment system. Neither program made significant created for investment projects, and approximately US$24 million was changes to its APA system, which both integrated with a national paid to workers as wages. performance assessment system called the Functional and Under the ongoing GSCSP, despite the significant delays in Organisational Assessment Tool (FOAT), later changed to the District implementation caused by the advent of the COVID-19 pandemic, 17 Assembly Performance Assessment Tool (DPAT). As LGCSP utilized km of roads, 7.49 km of pedestrian walkways, 8.94 km of storm drains, the existing FOAT as its APA tool, additional testing was not necessary 7 pieces of economic infrastructure, and 1,892 streetlights had been before the launch of the program. GSCGP also integrated DPAT constructed or rehabilitated as of February 2022, and more were under assessments into its APA system by requiring local governments seeking construction. The shares of infrastructure investment costs out of total UPGs to exceed national average performance scores from the DPAT UPG expenditures for LGCSP and GCSCP were 79 percent and 87 assessments as minimum conditions. Some of the challenges observed percent, respectively. with other UPG programs were repeated in both LGCSP and GSCSP, especially belated completion of APAs as a result of contract delays and limitations in planning and implementation capacities at the local level. Quality assurance. The APA results under the two programs were subject to robust quality assurance. Instead of hiring an independent firm, LGCSP effectively made use of the existing quality assurance system of the DDF, comprising experts from the DDF secretariat, Performance-Based Fiscal Transfers for Urban Local Governments 82 Part II Results and Lessons from Two Decades of World Bank Financing consultants, and development partners who jointly reviewed the building grant helped enable participating local governments to adopt assessment results. In particular, the DDF secretariat was responsible and implement integrated urban development plans with a better for overall quality assurance, including contracting with companies, understanding of their own urban agendas. The review found that the reviewing the results, and coordinating with a review committee. Under Capacity Support Grant had contributed to the procurement of office GSCSP, the World Bank team recognized that the quality of the APAs equipment, spatial information system development, improvement in was heavily dependent on the capacities of a hired consulting firm and revenue generation, and disaster risk management. so provided substantial handholding support and guidance to the firm at Sustainability. The impacts of UPGs in Ghana were highly likely to be the beginning of the process to ensure the credibility of the results. The sustained beyond the program period, for four reasons: contract for the consulting firm lasted two years, and the handholding efforts were found to have improved the capacity of the firm in the • The UPG programs were well-aligned with and further upgraded the second year. A system was also in place to reconciliate the firm’s results country’s existing IGFT system (the Responsiveness Factor Grant). with those of a quality assurance review team hired by the World Bank. • Regions increasingly employed UPG approaches, which was The ensured robustness of APAs generated adequate incentives for expected to have comprehensive impacts on the country’s overall participating local governments to improve their performances in the fiscal decentralization framework. targeted areas. • Ownership was strong at all levels of government, with the governments providing a high level of co-funding. Capacity building. Capacity-building support was provided by LGCSP • Core urban systems developed under the programs, such as a spatial mainly through a specific fund through which local governments could planning database and asset management tools, were likely to be apply for targeted capacity support in five PFM areas, according to retained beyond the program period. their specific needs. The fund complemented other capacity-building Environmental and social management. E&S management support channels, such as grants for the purpose from the DDF. It was generally functioned well under LGCSP. As the program was an IPF, found, however, that the fund under LGCSP (an IPF) did not provide the World Banks’s safeguards policies supplemented the existing sufficient incentive to the local governments to undertake planned national procedures (that is, the Environmental Protection Agency’s capacity-building activities because it was disbursed against inputs environmental assessment procedures), which local governments found (capacity-building proposals by the governments), instead of outputs challenging to adhere to fully. Most of the local governments, however, (actual capacity-building activities undertaken). Thus, GSCSP changed successfully managed to screen UPG-funded infrastructure projects its approach to a performance-based capacity-building grant—called early enough to comply with the E&S management rules. Also, the the Capacity Support Grant—which would make resources available provision of capacity-building support focused on E&S management, upon achievement of DLIs to strengthen the demand-driven aspects of and the development of a system to monitor E&S issues contributed to the support and the local governments’ ownership of their performance improving the overall effectiveness of the governments in addressing enhancement. One of the MCs a local government needed to meet to these issues. Among the challenges noted during the design stage were get access to the Capacity Support Grant was preparation of a three- displacement and compensation issues due to the lack of a land tenure year Urban Development Action Plan, providing a strategy for capital and property ownership database. investment to address key urban development challenges. In addition, compliance with the MCs to apply for the grant was included under GSCSP as a disbursement-linked indicator. This design of the capacity- Performance-Based Fiscal Transfers for Urban Local Governments 83 Part II Results and Lessons from Two Decades of World Bank Financing 8. India: Institutional Strengthening Table 8.1. Factsheet for the Case UPG Programs in India of Gram Panchayats Project Phase II Institutional Strengthening of Gram Panchayats Program II Program title 3. Background (ISGPP II) In support of India’s devolution process, the Institutional Strengthening Financing PforR of Gram Panchayats Project (ISGPP) was designed and implemented instrument from 2011 to 2016 as the first UPG program in the state of West Bengal. Financing IBRD: US$210 million After the successful completion of Phase I of the program directly amount targeting 1,000 selected local governments, Phase II (IGSPP II) was (Financier) India: US$384 million initiated in 2017 at the request of the state and the local governments Duration (called Gram Panchayats) to scale up the program’s initial impacts (Board approval March 2017 to December 2022 and expand the support to all the local governments in West Bengal in to closing) line with its decentralization and local development agenda. ISGPP II Geographical 3,229 local governments (all Gram Panchayats in the state of West retained and refined most of the aspects of Phase I, including alignment coverage Bengal) with a statewide performance grant system (called the State Fiscal Commission, or SFC) and the deployment of mentoring support and Project To strengthen further the institutional and financial capacities of system development components. ISGPP II was also unique in that it was development the Gram Panchayats across the state. objective designed to address the contextual differences among local governments, such as differences in demographic factors, location, overall capacity, and the level of support they had received for institutional strengthening in previous years. The case study that follows categorizes the local governments into two groups: those that had received support under the Phase I (hereafter Phase I local governments) and those that newly joined in Phase II and were less familiar with the UPG system (Phase II local governments). Table 8.1 shows key facts of the UPG program reviewed. Performance-Based Fiscal Transfers for Urban Local Governments 84 Part II Results and Lessons from Two Decades of World Bank Financing 2. Key performances Table 8.2. Key Performances of the Case UPG Programs in India In general, and despite some challenges, IGSPP II achieved satisfactory results relative to overall objectives and performance targets. In fact, performance trends were difficult to analyze because substantial Details (changes over the program Key performance areas Improvement period) changes were made to the performance measurement indicators during implementation of the program. While the MC compliance rate of General MC local governments overall improved from 52 to 75 percent, the average performance compliance Yes Increased from 52% to 75%. rate (%) PM score decreased slightly, by 31 percent. The scores did not decline because the local governments’ performance had worsened over time but Decreased from 84 to 58, but the rather because the performance bar had been raised significantly (that is, Average APA results are not comparable due to performance Not available changes to indicator targets which the requirements of the indicators had been strengthened), and a number score were increased or strengthened of external factors had had effects. The changes to the PM indicators also during implmentation resulted in mixed results in several performance areas (Table 8.2): Key urban OSR collected increased from US$28 OSR Yes • OSR increased significantly (by 68 percent) over the program period, variables million to US$47 million (68%) as a continuation of improvements from Phase I. The scores under the accountability • Transparency and accountability remained stable, as indicated Partially yes and transparency thematic area Accountability by the APA score in the accountability thematic area. The program (stable) remained largely the same over the years. also focused on developing digital platforms to improve information sharing. A statewide e-governance system, for example (called the Gram The overall PFM score of local Panchayat Management System), was developed and implemented governments improved significantly, under the program, and adequate training and handholding support in PFM Yes but planning and budgeting the use of such IT systems were provided to local governments. performance remained static over the years. • PFM showed mixed results. Although the overall PFM score (including planning, budgeting, accounting, financial reporting, and audit) No quantatitive data, but staff Human recruitment reportedly improved improved over the years, performance in the planning and budgeting resource Partially yes duirng the first three years and then thematic areas remained static. On the other hand, local government management was put on hold due to state policy. performance in financial reporting, audit, and accounting improved significantly. Also, the share of local governments without adverse or Asset O&M was not monitored during management N/A disclaimed audits was more than 98 percent throughout the program. implementation. / O&M • Human resource management also showed improvement, with the program making rapid progress in hiring core staff and achieving a relevant DLI for the first three years of implementation. Because of a state-level policy freezing the recruitment of new staff, however, the hiring of local staff stalled, and the DLI had to be adjusted, resulting in staff and capacity shortages among the local governments. Performance-Based Fiscal Transfers for Urban Local Governments 85 Part II Results and Lessons from Two Decades of World Bank Financing Access to infrastructure and service delivery also significantly improved The system was designed to pool available funds and then disburse over the program period. The program provided more than 5,000 km of them to different sets of local governments—that is, the funds from the roads and 250 km of drainage, and the implementation ratio of planned World Bank (IBRD) were disbursed to Phase II local governments, while infrastructure projects increased from 67 to 85 percent. The lower rate SFC and CFC funds supported the remaining Phase I governments. of execution during the first years of implementation was the result Performance assessment system. An APA system developed under of delays in and limited oversight of procurement processes.34 The the program was effectively institutionalized in the SFC grant system. majority (88 percent) of the UPG financing was spent on infrastructure The APA processes were contracted out to an independent verification and services and benefited all local governments in the state. agent (IVA) to ensure their credibility and to address local governments’ 3. Operational design key challenges, as identified from preparation diagnostics and Phase Grant design. The program was designed to provide small grants I experiences. All the SFC grants then began replicating the newly amounting to about US$1 per capita (US$20,000 per local government, created APA system from the UPG and gradually adopted overall on average), partly because of the drastic geographical expansion to performance-based grant systems. This, in turn, created a (healthy) cover 3,000 local governments. Despite their small per capita size, the competitive environment among the local governments and nurtured UPGs were expected to be sufficient to create meaningful impacts. This the development of several key systems at the local level. The APAs were was because a statewide performance-based grants system, which was found to be cost effective, with the total costs of this comprehensive, a combination of an SFC grant and the World Bank UPGs, constituted effective assessment exercise accounting for only 0.54 percent of the a substantial share (about 19–23 percent) of the overall development annual UPG amount. grants to the local governments at the design stage. A substantial On the other hand, the adjustment of the indicators during the increase in the Central Financial Commission (CFC) grant, however course of the program without sufficient communication with local (a centrally disbursed nonperformance grant), during implementation governments made it difficult for them to comprehend and respond reduced performance-based grants as a share of total development to the updated APA system. The delays in executing the APAs, which grants allocated to local governments to only 9 percent in 2019/20. This were, as mentioned, mainly due to contract holdups and the COVID-19 contributed, in turn, to a reduction of the UPGs’ impacts in improving pandemic, also negatively affected the important role of the assessment the governments’ performances and shifted the focus of the grants to results as financial incentives for local governments. Another challenge smaller-scale infrastructure investments, at about US$5,700 per project. related to the APAs was that the local governments’ efforts were mostly Despite concerns raised by the smallness of the grants, covering all limited to complying with the MCs, with less emphasis on achieving the the local governments in the state was deemed vital to achieving more qualitative PM targets. This was deemed to be in part because of the program objective of developing an integrated and harmonized low awareness of UPG systems and the small size of the grants. performance-based grant system that would not exclude any of them. Quality assurance. The program employed the “two-layer” quality assurance review system to ensure the robustness of the APA results. A third-party verification team hired by the government to review 34 During the first three years of Phase II implementation, the World Bank’s procurement system was applied, as the practice of the Phase I, which was an IPF, was carried over. Given that Phase a sample of 100 local governments was complemented by a quality II was a PforR, however, which required using the existing system in the country, the result was confusion and blind spots in state scrutiny as well as in Bank oversight, leading to a major assurance team hired by the World Bank, which reviewed a sample of 10 challenge to the program. This issue was assessed during the midterm review and was shifted Gram Panchayats. Most recently, the government-side verification team to the state systems for the remainder of the program implementation. Performance-Based Fiscal Transfers for Urban Local Governments 86 Part II Results and Lessons from Two Decades of World Bank Financing and the Bank team reviewed the same sample of local governments to Sustainability. The review found that UPGs in India were highly likely ensure strong checks and balances. This multilayered quality assurance to be able to sustain their impacts on institutional performances and system, together with the reconciliation and approval of results by system development beyond the program period, for three reasons: a steering committee, provided incentive to the participating local • The UPG system had been fully embedded in the statewide IGFT governments to strive for stable performance enhancement and build system, and it also had impacts on the grant system at the central overall trust in the system among stakeholders. level. Capacity building. Based on capacity gaps identified in Phase I, the • The core systems developed under the program in the areas of PFM, program came up with a unique design for capacity-building activities reporting, grievance handling, and accountability were likely to be that enabled it to remain flexible and ensured the local governments continued beyond the program phase. would have multiple avenues to express their capacity needs on an • The state government provided a high level of co-funding, annual and day-to-day basis. A key feature was the mentor deployment contributing 65 percent of the program costs, as well as pooling of program, which assigned over 400 specialized mentors to local funds from the World Bank and the SFC grants; this indicated strong governments to provide day-to-day handholding support for issues government ownership. related to the IT system, planning and budgeting, financial management In addition, an APA was in the process of being designed using data systems, service delivery, and procurement activities, among others. available in a transparent, objective, and timely manner from the now The mentor program was one of the most successful aspects of Phase I established and operational e-governance system. The objective was to and was retained and refined under Phase II. In addition, conventional enable the government to continue with the APA using its own systems annual training programs were provided, including training needs after the program closed. assessments involving participation of the local governments and the Environmental and social management. Overall, ISGPP improved its district offices. Annually, 15,000–25,000 officials were trained. E&S management by incorporating E&S issues into APAs and providing In view of the benefit derived by the Phase I local governments from targeted capacity-building support in these areas. The program mentoring support, Phase II was designed to provide more intensive facilitated the universal application of the Earth System Modeling mentoring support to the Phase II governments. The ratios of mentoring Framework across the state and implemented an online updating teams to government staff were 1/40 for Phase I and 1/20 for Phase system of land records containing information on all the private lands II. It was observed, however, that the performance of Phase I local utilized to implement infrastructure works under the program. Local governments showed a decline compared to Phase II local governments. capacity and HR gaps continued to persist, however. A new grievance This result implied the importance of continued and balanced capacity- redressal management system developed under the program, although building support for sustaining local performance enhancement and representing the first important steps in bringing about such an integrating technical assistance and capacity-building components into apparatus that was fully operational, was not fully utilized for registering financial incentives for local governments. To resolve the issue, the ratio complaints because of lack of awareness of the system and the existence and intensity of mentoring support for the Phase I local governments of parallel state platforms (in particular, the chief minister’s grievance was adjusted to target all local governments equally and to pay more redressal management system). attention to the weaker local governments in both groups. Performance-Based Fiscal Transfers for Urban Local Governments 87 Part II Results and Lessons from Two Decades of World Bank Financing 9. Kenya: Kenya Devolution Support Table 9.1. Factsheet for the Case UPG Programs in Kenya Program Program title Kenya Devolution Support Program-for-Results (KDSP) 1. Background Financing In alignment with such national decentralization initiatives as the PforR instrument National Capacity Building Framework, the Kenya Devolution Support Program-for-Results (KDSP) was introduced as the first PforR UPG Financing IDA: US$200 million amount operation in the country to strengthen the capacity of national and (Financier) Kenya: US$68.9 million county institutions and improve delivery of devolved services at the local level. At the time of this review, KDSP was augmenting the Duration (Board approval March 2016 to September 2021 government’s equitable share grant, which distributed a minimum of to closing) 15 percent of national revenues (US$2.5 billion) to local governments35 Geographical through a progressive formula to benefit historically marginalized local All 47 local governments (counties) coverage governments and enhance the capacity of all local governments in the country. Capacity-building activities at the center of the program Project To strengthen capacity of core national and county institutions to development design were to address capacity weaknesses observed throughout the improve delivery of devolved services at the county level. objective devolution process. To sustain and scale up the impacts of KDSP over the years, a subsequent program was under preparation per request from the central, county, and local governments. The design and implementation experiences under KDSP had also informed other ongoing UPG programs, such as the Kenya Urban Support Program and a program focused on climate resilience, called Financing Locally Led Climate Action. Table 9.1 shows key facts of the UPG program reviewed in this study. 35 In the case of Kenya, “local governments” refers to county governments. Performance-Based Fiscal Transfers for Urban Local Governments 88 Part II Results and Lessons from Two Decades of World Bank Financing 2. Key performances Table 9.2. Key Performances of the Case UPG Programs in Kenya At the time of the review, the UPGs under KDSP (called capacity and performance grants) had, overall, successfully movitated the participating local governments and the national government to achieve Details (changes over the program Key performance areas Improvement period) the targeted areas. Two funding windows comprised level 1 grants for capacity-building support and level 2 grants for local development General MC (see the grant design section, below, for more details). All participating performance compliance Yes Increased from 28% to 81%. rate (%) local governments (counties) complied with the MCs for level 1 grants, and the rate of compliance with MCs for level 2 grants increased by Average APA 53 percentage points over the program period. The average PM score performance Yes Increased from 42 to 71. score increased by 69 percent for all the local governments during the same period, and the scores of 83 percent of the governments consistently Key urban OSR collected increased by 66% in variables OSR Yes improved on a year by year basis. KSDP made good progress overall nominal figures. across the following key result areas (Table 9.2): Asset APA score on O&M budget increased management Yes from 0.1 points to 2.2 points out of a • OSR increased by 66 percent in nominal figures. / O&M maximum of 4 points. • Asset management showed relatively slow progress. On the sufficiency Average APA score on PFM increased of O&M budgets to ensure the sustainability of infrastructure (that is, from 43% to 71%; quality of financial on having at least 5 percent of capital budgets allocated for O&M), statements improved by 18% and PFM Yes the local government scores improved, on average, from 0.1 points financial reporting by 13%; and the to 2.2 points out of a maximum of 4 points. Forty percent of the counties’ performance on the internal audit showed an upward trend of 26%. governments, however, still scored zero under this indicator. • PFM improved in various aspects. According to the APAs, scores on Average APA score on accountability monthly financial reporting improved by 13 percent, and the counties’ increased by 64%; score on communication framework and performance on the internal audit trended upward by 26 percent. Accountability Yes engagement improved by 30%; and • Accountability improved, with scores on communication framework score on public participation (e.g., and engagement rising by 30 percent and on public participation budget forums held, publication of bills) increased by 19%. (budget forums held and publication of bills) by 19 percent. • HR management scores increased on average by 16.8 percent over Human Average APA score on HR the program period. Achievements recorded included improvements resource Yes management increased from 35 to 69 management out of a maximum of 100 points. in review of organizational structures and staffing patterns and in planning for staff recruitment and development, and strengthened systems for recruitment, promotions, and appointment of staff. Performance-Based Fiscal Transfers for Urban Local Governments 89 Part II Results and Lessons from Two Decades of World Bank Financing The implementation of infrastructure and service delivery improved in size from US$400,000 to $600,000 for very small, low-performing significantly, as well. Although execution rates took a couple of years governments to US$5 million for well-performing ones, based on the to be on target, they continued to improve, and almost all the planned results both of an equitable share formula37 and the PMs. Given that infrastructure projects in the local development plans were being the best-performing local governments could earn reward grants of implemented, according to the latest APA results. At the program up to 300 percent of the lowest performers’ development grants (or closing, 171 subprojects had been implemented in the areas of health fourfold), the overall grant size was appropriate. The participating local (comprising 44 percent of the total subprojects), water (19 percent), governments were able to invest mostly in medium-scale infrastructure trade, including rural markets (9 percent), agriculture (6 percent), projects (US$200,000 on average) through the level 2 grants. education (4 percent), and rural roads (6 percent). During the COVID-19 Performance assessment system. In accordance with the grant design pandemic, KDSP investments in health facilities provided a huge boost discussed above, the APA consisted of three sets of measurement for the counties, with the refurbishment and equipping of highly indicators: basic MCs for level 1 grants on the preparation and execution specialized hospitals benefiting over 1.2 million people. of capacity-building activities; extended MCs for level 2 grants on the 3. Operational design status of core systems in place (for example, PFM, E&S management, Grant design. KDSP grants were generally smaller than other UPGs. and complaints handling); and PMs focused on qualitative aspects of While the grant size across local governments varied exponentially the core systems, including OSR, asset management, audits, planning, with performance, a rough estimation put them at approximately HR management, and accountability. The APAs were formulated to US$1.80 per capita on average—far smaller than the grants under other address the identified weaknesses and bottlenecks impeding effective UPG programs.36 The UPG share of the total grants allocated to local service delivery by local governments, as well as insufficient systems governments was also small, at only 5 percent. and procedures at the national level in terms of supporting these governments. The APA system performed well in operationalizing the For the limited grants to provide adequate incentives to local governments, grant design as envisaged at the design stage, despite recurrent delays, several mitigation measures were adopted. First, two different sets as observed in other cases. of grants were introduced: level 1 grants allocated approximately US$300,000 per local government for compliance with basic MCs, and Quality assurance. The results of the APAs were verified both by the level 2 grants allocated US$1.5 million per local government annually national government and the World Bank. The reviews by the Bank for infrastructure projects for compliance with extended MCs. The team, however, were mainly carried out through desk reviews without rationale for applying the two funding windows with different sets hiring a specialized external consultant. Although this is considered less of MCs was to provide strong incentives to well-performing local rigorous than a systematic approach entailing field visits by specialized governments through level 2 grants while ensuring all the governments consultants to randomly selected sample local governments, the Bank would have easy access through level 1 grants to a fixed share of the task team reported no critical discrepancies in the submitted results. money to serve as startup funding for addressing core performance In fact, however, there were some disagreements early on among gaps. In addition, the level 2 grants were uniquely designed to provide stakeholders with regard to the measurements and results of certain exponential incentives to well-performing local governments, ranging safeguards performance measures, which could have been avoided with Based on an average investment grant of about $1.8 million per county, with the counties 36 37 The equitable share formula was based on population, poverty, equal shares, land area, and a having about 1 million people on average. fiscal component. Performance-Based Fiscal Transfers for Urban Local Governments 90 Part II Results and Lessons from Two Decades of World Bank Financing a predefined systematic review process. During the midterm review, a Environmental and social management. General improvements were focus on E&S safeguards and a grievance redressal management system observed in local E&S management throughout the program period. was strengthened as part of the APA. The incorporation of E&S performance measures in APA systems led to an increase of 15 percent in the number of local governments preparing Capacity building. The KDPS level 1 grant was introduced as a new, annual environmental and social audits or reports. In addition, 90 systematic approach to capacity-building support. The grant combined percent of the local governments had E&S management plans, according national-level support (such as guideline development and peer-to-peer to the latest results. A grievance redressal management system was also learning) with demand-driven support informed by APA results and put into place in all 47 local governments. Among the challenges noted comprehensive annual capacity-building plans. It was also well-aligned during implementation was the recruitment of qualified safeguards with other initiatives under Kenya’s devolution capacity-building officers and the raising of awareness of the issues at the local level. systems, especially under the National Capacity Building Framework, and was complemented by analytical and technical assistance funded by the Kenya Accountable Devolution Program, a multidonor trust fund. The packaged approach performed well over the program period, with the quality of capacity-building plans significantly improved and the overall coordination of the capacity-building support provided to local governments strengthened. Sustainability. The overall impact of the program was found highly likely to continue beyond the project period. The program established a new framework for performance-based grants and tested the use of treasury functions for the routing of funding from development partners. In addition, it paved the way to include devolved UPG systems in the government’s Division of Revenue Act and the County Allocation and Revenue Act by directly influencing national budgeting processes and the equitable allocation formula for horizontal resource sharing among local governments. It also laid the foundation for a coordinated framework for interministerial program implementation, in close coordination with the Council of Governors, the National Treasury, the Ministry of Devolution and Arid and Semi-Arid Lands, and other line ministries. To continue the momentum of its successful implementation, a second phase of KDSP was being planned. Performance-Based Fiscal Transfers for Urban Local Governments 91 Part II Results and Lessons from Two Decades of World Bank Financing 10. Tanzania: Urban Local Table 10.1. Factsheet for the Case UPG Programs in Tanzania Government Strengthening Programme de Développement Urbain et de Gouvernance Program Program title Locale, PDUGL 1. Background Financing PforR instrument In Tanzania, a decentralization reform initiative known as Decentralization by Devolution has been a key pillar of the country’s Financing amount US$255 million national growth and poverty reduction strategies, together with fiscal (Financier) transfer programs, including the Local Government Development Grant. The World Bank has supported the country’s grant through the Duration (Board approval October 2012 to December 2020 Local Government Support Project (US$98 million), which became to closing) effective in 2004, and, since 2012, the Urban Local Government Strengthening Program (ULGSP). As one of the piorneering UPG Geographical 18 urban local governments (70% of all urban local governments in coverage Tanzania) programs, ULGSP introduced the first multisectoral performance- based grant using the new lending instrument, Program-for-Results Project To improve institutional performance for urban service delivery in (PforR), at the Bank, as well as the first grant introducing a urban window development program urban local government authorities. objective in the country. The new instrument was also innovative in terms of directing the DLIs toward addressing core central and local government capacity gaps. The successful ULGSP model has been replicated and customized in numerous other countries. At the time of this case study, a follow-on project, the Tanzania Cities Transforming Infrastructure and Competitiveness Project (TACTIC), was under preparation to complement and continue to advance the impact of ULGSP. Table 10.1 shows key facts of the UPG program reviewed in this study. Performance-Based Fiscal Transfers for Urban Local Governments 92 Part II Results and Lessons from Two Decades of World Bank Financing 2. Key performances Table 10.2. Key Performances of the Case UPG Programs in Tanzania Overall, the ULGSP showed good achievements throughout the program period. The compliance rate with the MCs reached 100 percent Details (changes over the program of the participating local governments, starting from 83 percent, and the Key performance areas Improvement period) average PM scores increased by 21 percent. Institutional performances in the following key result areas improved, despite some structural General MC Yes Increased from 83% to 100%. performance compliance challenges (Table 10.2): rate (%) • OSR improvement was a key achievement of ULGSP. Increased Average APA Yes Increased from 60 to 80 points out of revenue from property tax was measured by DLIs as one of the performance 100. five key performance areas. The absolute amount of OSR collected score increased by 163 percent during the program period. Since property Key urban OSR collected increased by 163% in tax valuation and collection had been centralized since 2016, variables local currency for the first five years however, the indicators corresponding to property tax under ULGSP OSR Partially yes of implementation, but the indicator were dropped at that time. dropped beginning in 2017 due to the government’s centralization initiatives. • Asset management was enhanced. In 2017, 94 percent of local governments became able to spend more than 70 percent of their Accountability score of local Accountability Yes annual O&M budgets, although the funding allocated for O&M was governments increased by 86%. still less than adequate, especially for roads. Share of local governments that spent Asset • Financial management in local governments steadily improved more than 70% of their annual O&M management Yes during implementation. The local governments’ average scores on budgets on O&M increased from 0% / O&M to 94%. fiduciary system efficiency increased by 51 percent. The procurement compliance rate reached 87.2 percent in fiscal year 2019, surpassing Local governments’ average scores the target level of 80 percent. on efficient fiduriary systems have PFM Yes increased from 12.8 to 19.3 out of a max • Staffing, especially at the regional level, was insufficient relative score of 25. to what was envisaged at the design stage, because of delays in processing recruitments. Human Insufficient staffing at the regional resource Yes level. • Accountability and oversight improved, as well. The APA score on management accountability increased by 86 percent over the program period. Performance-Based Fiscal Transfers for Urban Local Governments 93 Part II Results and Lessons from Two Decades of World Bank Financing The program contributed significantly to the provision of urban the protracted procurement of a new IVA firm delayed the fourth infrastructure and service delivery, as well. The completed infrastructure APA and affected the assessment and disbursement cycle for the under ULGSP included 179.3 km of roads with solar streetlights and following year. The key performance areas of the APA comprised roadside drains; 6.1 km of standalone stormwater drains; 10 bus terminals the following five: constructed; 2 lorry parking facilities constructed; 4 modern markets • Improved urban planning systems constructed; 3 dumpsites rehabilitated; and 290 waste collection points • Increased revenues from property tax installed. Successful implementation of infrastructure projects larger • More efficient fiduciary systems than before also indicated improved institutional performance at the local level.. Moreover, the construction of urban roads was associated • Improved infrastructure, implementation, and Operation & with an increase in land values in areas surrounding ULGSP-funded Maintenance (O&M) project sites by around 100–300 percent across local governments. At • Strengthened accountability and oversight systems least 52,402 jobs were created during construction and through small These indicators were also monitored through a DLI. businesses along upgraded roads, markets, and bus terminals financed Quality assurance. In addition to the government’s review of APA under the program. results, the World Bank carried out an independent review in a timely 3. Operational design manner to enhance the effectiveness of the assessment process and the robustness of its results. The Bank review team prepared a report Grant design. The grant as designed was adequate in size, at around based on a draft APA submitted by an IVA firm, on the basis of which US$15–20 per capita, and during implementation it was actually the assessments by the Bank and the government were reconciled. US$10.31 on average. The UPG-funded infrastructure projects across Once agreement was reached on the results, a final synthesis APA all 18 participating local governments were medium size on average, report was published. at around US$650,000 per project, falling in the range of US$0.5 million–$1 million as designed. The UPGs also accounted for a sizeable Capacity building. At the design stage, the program envisaged three share of total grants to local governments, comprising 24.7 percent forms of capacity-building support: of total funding streams and 76.8 percent of development grants. The • Activities implemented by the local governments, using up to 5 allocation formula under the program was conventional, in that the size percent of the UPG funds of the grant varied linearly with the annual PM score achieved by each • Support from the central government to the local governments on local government. preparation of required annual capacity-building plans Performance assessment system. Building upon UPG experiences in • Ongoing capacity-building support through mentoring and issue- the country and the Africa region, the program’s APA system was well specific zonal response teams designed. No substantial changes were made to indicators throughout During the first years of the implementation, intensive capacity- implementation, which made the system predictable and stable. For building support for the local governments was rendered both by the the first three years, the central government procured an independent central and regional governments. Later, however, these efforts did verification agent (IVA) in a timely manner through a single contract, not provide the anticipated level of support to the local governments, which reduced transaction costs and made the procurement and nor did they adequately target the skills and systems they needed mobilization of APAs more efficient. From the fourth year, however, to deliver urban infrastructure services. As a result, the fifth Performance-Based Fiscal Transfers for Urban Local Governments 94 Part II Results and Lessons from Two Decades of World Bank Financing disbursement under a DLI on completion of annual capacity-building activities by the central government was dropped and reallocated. Sustainability. Overall, the program successfully contributed to improving institutional performance and system development that could be sustained beyond the program period. It was well aligned with the existing government grant system in the country. Strong government ownership and high visibility of the UPG system led to the preparation of a follow-on project. The continuation of the UPG system was put at risk, however, when the central government centralized property tax collection and gradually reduced fiscal and operational support to local governments. Although the responsibility for property tax collection was returned to the local governments in 2021, such temporary policy changes revealed the vulnerability of UPG systems to external factors. Environmental and social management. The program effectively addressed E&S issues by incorporating E&S indicators in both APAs and DLIs, as well as by providing targeted capacity-building support in these areas. Local governments were asked, for example, to report on E&S issues every year during annual workshops, leading to consistent compliance with MCs on these issues and the inclusion of E&S management plans in construction contracts. The relevant E&S scores in APAs improved steadily, despite some notable lapses in health and safety standards in site-specific E&S management plans. Performance-Based Fiscal Transfers for Urban Local Governments 95 Part II Results and Lessons from Two Decades of World Bank Financing 11. Tunisia: Urban Development and Table 11.1. Factsheet for the Case UPG Programs in Tunisia Local Governance Program Programme de Développement Urbain et de Gouvernance Program title 1. Background Locale, PDUGL Urban Development and Local Governance (Programme de Financing PforR Développement Urbain et de Gouvernance Locale, or PDUGL) is instrument a Program-for-Results (PforR) operation with a strategic focus on IBRD: US$300 million supporting the country’s ongoing reform of capital grants (the Municipal Financing amount Tunisia: US$63 million Development Fund [Caisse des Prêts et de Soutien des Collectivités (Financier) Locales, or CPSCL])38 and associated capacity-building assistance. The IBRD additional financing: US$130 million program began in 2014, covering all 264 local governments and their Duration July 2014 to December 31, 2019 7 million urban residents (two-thirds of Tunisia’s total population). In (Board approval Extended to June 30, 2023 2018, additional financing was approved to accommodate an increased to closing) population, which had risen to 9.8 million as a result of communal Geographical All 264 local governments in the country border extensions, and to strengthen administrative transparency, coverage accountability, and HR capacity by further aligning the intergovernmental Project (a) Strengthen local governments’ performance to deliver fiscal transfer (IGFT) system with the government’s decentralization development municipal infrastructure; and (b) improve access to services in framework. In July 2020, in its response to the COVID-19 pandemic, the objective targeted disadvantaged neighborhoods. program was restructured to adapt activities and targets to the sanitary context and support municipalities’ economic recovery in the medium term. Table 11.1 shows key facts of the UPG program reviewed in this study. CPSCL is primarily accountable to the Ministry of Interior and the Ministry of Finance, 38 as a separate entity providing local governments with access to credit as well as to central government funding to finance municipal investments. Performance-Based Fiscal Transfers for Urban Local Governments 96 Part II Results and Lessons from Two Decades of World Bank Financing 2. Key performances Table 11.2. Key Performances of the Case UPG Programs in Tunisia The review found PDUGL generally on track in terms of improved performance in the targeted areas. The MC compliance rate of the local governments in the first year of implementation was 89 percent, which Details (changes over the program Key performance areas Improvement period) was a drastic increase from 40 percent at the time of the pilot carried out before the program launch. The score consistently exceeded planned General MC Yes Increased from 89% to 97%. targets over the program period. The average PM score of all the local performance compliance rate (%) governments also significantly increased, by 41 percent, between the first and fourth APAs, and 76 percent of the local governments met the Average APA Yes Increased from 58 to 82. PM minimum score in 2017, exceeding the program’s end target of 70 performance score percent. Key urban Share of local governments preparing Although quantitative data were relatively limited in some areas because variables OSR collection plans increased from 87% of challenges to data collection and to the M&E system, the program to 97%; 84% of the local governments OSR Partially yes reportedly showed achievements in the following performance areas were able to collect 70% of envisaged OSR. Nominal figure of OSR collection is (Table 11.2): unknown. • Asset management improved over time. The percentage of local PM scores of local governments on governments that allocated more than 12 percent of their O&M Human preparation and implementation of human budgets increased from around 65 percent in the first year to 90 resource Partially yes resource management plans increased percent in the fourth APA. In addition, almost all (98 percent) management from 62 to 98 out of 100 in 2018, but staffing level was limited. executed over 10 percent of their O&M budgets in the fourth year. An asset accounting model was also developed and at the time of the Share of local governments allocating review was being used by six pilot local governments, with results to more than 12% of their O&M budgets increased from around 65% in the first be fed into the assets and inventory management procedures manual O&M Yes year to 90% in the fourth APA; almost all to be published at the end of the program. the local governments (98%) executed • Accountability was enhanced. The share of local governments over 10% of O&M budgets in the fourth year. reaching the maximum score on access to information and complaints management exceeded 90 percent according to the latest A DLI on financing management and APA result—an increase of 52 percentage points from the first year. reporting was achieved. While financial PFM (timely reporting was almost nonexistent at the • OSR management improved, but the absolute amount of the increase reporting and Yes local level prior to the program, it was was unknown. According to the APAs, the share of local governments auditing) integrated into local governments’ modus that prepared OSR collection plans increased to 97 percent, and operandi during the program period. 84 percent of the governments were able to collect 70 percent of Share of local governments that reached envisaged OSR, according to the fourth APA. It was noted, however, the maximum score on access to that collecting data on increased OSR amounts in nominal figures Accountability Yes information and complaints management across the local governments was challenging, mainly because of the increased from 40% to 93%. Performance-Based Fiscal Transfers for Urban Local Governments 97 Part II Results and Lessons from Two Decades of World Bank Financing lack of a centralized monitoring system in the country and the limited was lowered in 2019.39 A multiyear approach to investment planning authority and capacity of local governments to collect OSR. The instead of annual planning and allocations was adopted to improve central government’s new decree on decentralization was expected efficiency and scale up impacts. A VfM audit report in 2020, however, to contribute, however, to improving the local governments’ capacity suggested a number of areas for improvement to ensure the quality and and authority to collect and monitor local taxes. efficiency of the provided infrastructure and services; these included • Timeliness of reporting and the quality of auditing improved. A budgeting, contracting, timeliness, and participatory processes. DLI on financing management and reporting was met, and, while financial reporting was almost nonexistent at the local level prior 3. Operational design to the program, it was integrated into the local governments’ modus Grant design. The grants were roughly US$5 per capita, which was operandi during implementation. To assure the quality of reporting, smaller than those from other UPG programs. While initially designed in 2019 the central government (through CPSCL) asked the program to be larger, at around US$25 per capita, during implementation the to provide local governments with grants in two installments per size was reduced, especially after the additional financing extended year, with the second subject to the quality of the local governments’ the duration of the program and increased the number of targeted reporting. beneficiaries to 9.8 million. It is the task team’s view that, while increasing • Human resource management overall showed mixed results under the per capita grant size might have been beneficial by strengthening the program. HR management was monitored as part of the APA the incentives and impacts, keeping it small was a strategic decision to thematic areas focusing on the preparation and implementation of general HR plans, which showed a significant improvement over the ensure all the local governments and population would have equitable program period. Recognizing that HR staffing levels and competencies access to the grants in a countrywide system and to introduce and had been relatively poor, however, the additional financing further operationalize overall nationwide capital grant reforms. The limited per strengthened the focus on HR management. It included a new DLI capita grant size was complemented by the substantial share (40 percent) on providing incentive for the improved staffing of key functions and represented by UPGs of the total grants flowing to local governments on structural reform toward an ICT/data-driven HR approach, with, average. The share was even higher for the smaller local governments, for example, development of an online job market and management providing them with stronger incentives to improve their performances. system. The UPGs focused on medium-size infrastructure projects, amounting The program significantly improved access to infrastructure and service to roughly US$200,000 per project, although this varied greatly across delivery in all the local governments. As of April 2021, 89 percent of the local governments. planned investment projects under the program had been completed, including 515 km of roads and streets rehabilitated or constructed, along with 5,400 streetlights installed. The delivery of the planned investments was also effective. Approximately 95 percent of the UPGs were directly spent on infrastructure investments. The execution level improved from 36 to 64 percent during the program period, although the levels were hard to compare over time, as the performance bar of the related DLI 39 Initially, a DLI was designed as the percentage of local governments that had executed at least 70 percent of the expenditures involved in their annual investment plans. Because of their recurrent difficulties in reaching these targets, however, it was agreed to lower it to at least 30 percent of their annual plans. Performance-Based Fiscal Transfers for Urban Local Governments 98 Part II Results and Lessons from Two Decades of World Bank Financing In addition to the conventional UPG that provided unconditional capital Quality assurance. The APA results were verified by an external IVA, grants to all local governments, the PDUGL program had another who was hired initially by the World Bank but, after the additional funding window: conditional grants for disadvantaged neighborhoods financing, by the government. The Bank team continued to assess the in selected local governments. This was a vehicle to support the quality of the APAs through implementation support missions. The government’s policy priorities with regard to upgrading service levels in Bank team also supported a reconciliation of inconsistencies between historically disadvantaged neighborhoods. Through intensive regional the IVA and the Controller General of Public Services, with no critical consultations, 229 neighborhoods across 144 local governments were fiduciary issues selected, and they prepared potential investment menus to be funded Capacity building. The provision of capacity-building support was by the grants. The application of the APA system to this conditional based on an annual local capacity-building plan prepared by each local grant was limited to MCs on a one-time basis, when the construction of government to cater to the varying demand from all. Support was also the approved investment plans was projected to commence. provided to the central institutions and systems with significant roles Performance assessment system. The program’s APA system was in the effective operation of the local governments. The capacity- introduced progressively, with a manual developed in 2015 and a full- building activities were funded from the central government budget scale dry run successfully implemented in 2016 before the APA became line, amounting to a total of US$10 million, without a specific capacity- fully operational in 2017. The APA focused on three thematic areas: building grant allocated. While less attention was paid to capacity governance (participatory planning and budgeting, transparency, and building at the beginning of the program due to a lack of interest on the procurement, among others); sustainability (asset inventory, financial part of the client in adding this component, the final program design recovery plans, OSR collection plans, and so on); and management. The incorporated the execution of such activities in both APAs and DLIs to APAs were carried out by a government agency external to the program provide strong incentives for this support. The result was the improved (the Controller General of Public Services), following a common practice capacity of the municipal staff to implement capacity-building plans in the Middle East and North Africa region. Another government and provide technical assistance and trainings; in 2019, 84 percent of agency (CPSCL) was responsible for the assessment of MCs, however, the local governments received at least 70 percent of the training actions which resulted in increased transaction costs and confusion, as well requested in their annual capacity-building plans. as delays in the release of the grants.40 It was the task team’s view that Sustainability. The program was introduced as the first PforR the MC and PM processes needed to be combined and conducted by in the country to address challenges in sustaining the impacts of the same agency, the Controller General of Public Services, to improve previous infrastructure-focused projects financed by the Bank and the efficiency of the APA process. During the midterm review in 2017, to provide better support for the country’s policy priorities, such as APA indicators were switched from being process based to more results decentralization and the empowerment of local governments and based, with a new focus on HR included. The new set of indicators was governance. As envisaged, the program had impacts on the development supposed to kick in from 2020, but this had not taken place at the time of core urban systems, such as a capacity-building support system, an of the review due to COVID-19. asset management system, and a performance-based grant system, all of which became the de facto modus operandi of local governments. The central government’s performance was also satisfactory in leading 40 The two agencies also assessed PMs and MCs at different times That is, PMs were assessed in September of the calendar year and MCs between January and February of the next calendar the implementation of this new type of instrument as part of operating year. Performance-Based Fiscal Transfers for Urban Local Governments 99 Part II Results and Lessons from Two Decades of World Bank Financing its own IGFT system and in looking into additional ways to increase financing for local governments, indicating a high level of ownership. Environmental and social management. The program effectively improved E&S management in APAs; the share of local governments achieving a maximum score in this area increased from 65.7 to 93.4 percent over the program period. The program also gave the central government incentive to provide capacity-building support to local governments in this area and provided targeted trainings to local governments on E&S issues, influencing the country’s E&S-related legal framework and guidelines. Performance-Based Fiscal Transfers for Urban Local Governments 100 Part II Results and Lessons from Two Decades of World Bank Financing 12. Uganda: Uganda Support Table 12.1. Factsheet for the Case UPG Programs in Uganda to Municipal Infrastructure Urban Support to Municipal USMID Additional Financing Development Program Program title Infrastructure Development Program (USMID AF) Program (USMID) 1. Background Financing PforR PforR Uganda was one of the first countries to apply performance-based grants instrument in their IGFT systems. In 1997, the country piloted such grants in five IDA: US$150 million IDA: US$310 million districts with support from the United Nations Capital Development Financing Fund (UNCDF). They were subsequently scaled up with the World Uganda: US$10 million Uganda: US$40 million amount Bank’s first UPG program, the Local Government Development Program (Financier) Host community and refugee sub-window: US$50 million I & II (2000–2007), followed by the Local Government Management and Service Delivery Program (2007–13). Based on the lessons learned from Duration March 2013 to May 2018 May 2018 to December 2023 the three programs, the Urban Support to Municipal Infrastructure (Board approval to closing) Development Program (USMID) was launched in 2013 with the new lending instrument, Program-for-Results (PforR), for urban local 14 urban local governments 22 urban local governments and Geographical governments. Having been successful in improving institutional 8 selected (refugee-hosting) coverage districts performances and urban service delivery in large urban local governments, USMID was scaled up in 2018 with additional financing. To enhance institutional Project performance of program local Given the substantial changes made to overall UPG design during the development governments to improve urban additional financing, this study assessed the results of USMID and the objective service delivery. additional financing program (USMID AF) separately. Table 12.1 shows key facts of both reviews. Performance-Based Fiscal Transfers for Urban Local Governments 101 Part II Results and Lessons from Two Decades of World Bank Financing 2. Key performances Table 12.2. Key Performances of the Case UPG Programs in Uganda While USMID showed significant improvements over the five years of its implementation, USMID AF had mixed results in some performance areas. Under USMID, all of the participating local governments Key performance areas Improvement Details (changes over the program period) complied with MCs consistently, and the average PM scores improved by 52 percent over the five years, surpassing the end targets as defined in General MC Yes for both Consistently 100% under USMID; the initial design. From 2018 under USMID AF, the MC compliance rate performance compliance increased from 78% to 100% under AF. rate (%) continued to show significant progress with all the local governments, but their overall performance did not improve. This is mainly because Average APA Yes for USMID Increased from 61 to 92 under USMID; the geographical coverage was gradually expanded from 14 to 18 local performance only decreased from 71 to 69 under AF, score mainly due to the increased number of governments in the first year and then to 22 local governments, and urban local governments. many of those newly added were smaller, less capacitated, and/or in Key urban Yes for USMID Amount of OSR collection increased by newly constituted urban areas. The geographical expansion was based OSR variables only 83% (USMID); not applicable under AF. on technical criteria as well as political economy and regional balance, which made performance improvements more challenging. Moreover, Human Measured as an MC, which was resource Yes for both consistently met under USMID and AF. the performance bars of the APA tool were raised significantly, with management more demanding indicators and a focus on thematic areas that were harder to improve in the short term. Share of local governments executing 75% of their O&M budgets increased O&M Yes for both Key performance results included the following (Table 12.2): from 21% to 100% under USMID and from 33% to 100% under AF. • OSR collection increased by 82.5 percent in nominal figures under USMID, and the collection rate of the budgeted OSR amount APA scores on the PFM thematic area Yes Yes for PFM increased by 67% under USMID and by exceeded 80 percent in 13 out of 14 local governments in 2018. From both 25% under AF. 2018, however, changes to legislation related to OSR collection and valuation restrained the policy options of local governments to APA scores on accountability and Transparency Yes for USMID transparency improved from 8.9 to 11.2 collect urban taxes. and only points out of maximum of 13 points Accountability • Asset management improved continously under USMID and USMID (USMID). AF. The share of local governments complying with 75 percent execution of their O&M budgets increased by 79 percentage points, from 21 to 100 percent under USMID and from 33 to 100 percent under USMID AF. • Financial management scores on APAs increased by 67 percent and 25 percent under USMID and USMID AF, respectively. • Human resource management was measured as an MC, which was consistently met under USMID and USMID AF. Performance-Based Fiscal Transfers for Urban Local Governments 102 Part II Results and Lessons from Two Decades of World Bank Financing • Transparency and accountability improved, as evidenced by Performance assessment system. The program’s APA tool was firmly a 26 percent increase in APA scores in this area over the program anchored in lessons learned from previous World Bank operations, as period. The program also enhanced access to information through well as performance weaknesses identified from field testing with a development of an online information-sharing system. large number of local governments before the program launch. APAs USMID enhanced both the quantity and quality of infrastructure were conducted by an independent verification agent (IVA) and further and services across local governments. The completed infrastructure strengthened to focus more on key urban management challenges under included 78 km of roads, 46 km of bicycle lanes, 111 km of paved pedestrian the additional financing. A set of DLIs was effective in supporting local walkways, 2,738 streetlights, and 56,400 square meters of green areas, governments in institutional improvements and in their gaining access to benefiting over 1.3 million people. According to VfM audit reports, the the grants. As often observed in other UPG programs, however, repeated USMID-funded infrastructure projects recorded higher VfM scores and delays in the APAs hampered allocation planning and predictability, as were more cost efficient than other non-USMID investments in similar well as synchronization with the country’s budget cycle. To address this sectors (Government of Uganda 2019). The program also enabled issue, the program sought to improve the effectiveness and timeliness integrated approaches to the design and implementation of larger of the APA processes by integrating them with a new online reporting infrastructure projects across local governments through enhanced system, such as the Online Performance Assessment and Monitoring procurement systems and close cooperation among the governments. System (OPAMS).41 Associated challenges were some delays in the execution of large Quality assurance. A specialized consultant recruited by the Bank projects and limited capacities of contractors, but these were effectively reviewed the results of the APAs to ensure quality, objectivity, and fairness addressed by close procurement and contract implementation support, to all local governments. The government worked closely with the Bank technical assistance, and improved guidelines, among others. team and carried out dry runs prior to formal APA assessments. The 3. Operational design results were further scrutinized and approved by a program technical Grant design. The program’s per capita grant size was one of the highest committee comprising different ministries, departments, and agencies. in the world among multisectoral performance-based grants. It was The APAs, together with quality assurance, provided a credible review increased systematically from US$20 to US$45 per capita under USMID process throughout both USMID and USMID AF. AF, which was significantly higher than the US$2 per capita provided Capacity building. The program provided targeted capacity-building under the previous programs (LGDP I & II) and the US$1–2 per capita for support to less capacitated local governments, focusing on such aspects other performance-grant schemes in the country. The grants comprised of overall institutional strengthening as the enhancement of working a major share of the total grants and development grants provided environments, career development, and local planning capacities. The to local governments (39 percent and 92 percent, respectively). This capacity-building support contributed to a higher retention rate and significant grant size allowed local governments to invest in integrated positively affected the nationwide capacity-building support system. infrastructure projects larger than US$1 million, often with multiyear Capacity-building activities at the local level on HR development and implementation plans. Also important to note is that the UPGs under operating system enhancement were supported through a specific the program were the only multisectoral development grants available to local governments, whereas other grants comprised a type of 41 This OPAMS framework was supported by another WB operation supporting the countrywide earmarked sector funding. application of performance grants, including multisectoral and sector grants under the Uganda Intergovernmental Fiscal Transfer Reform Program, with support from the government of the United Kingdom and implementation support from the Overseas Development Institute. Performance-Based Fiscal Transfers for Urban Local Governments 103 Part II Results and Lessons from Two Decades of World Bank Financing capacity-building grant, amounting to approximately 6 percent of total UPGs, with fewer access conditions than development grants. The local governments could receive these grants if they had capacity-building plans in place and had executed such activities utilizing previous capacity-building grants within agreed-on parameters. Sustainability. In addition to the physical infrastructure provided, the program supported the development of core systems and procedures that were likely to be long lasting: • It developed systems of VfM audits that were applied for other infrastructure investments. • Its APA system positively affected the country’s performance assessment reform in 2018 and was replicated in a nationwide performance assessment manual with strong political support. • It improved procurement procedures for larger-scale infrastructure projects (for example, the clustering of cWcts across local governments). • It promoted interministerial collaboration. USMID also influenced overall the urban roads design manual to include such issues as incorporating nonmotorized transportation and associated infrastructure. The program developed the Integrated Revenue Administration System, an automated system for tax assessment and collection, which was adopted by other development partners and was being rolled out to other local governments even outside the program. Some constraining factors were present, too, however; these included changes to the legal framework for OSR, HR capacity gaps, and changes in the administrative boundaries and status of local governments (including increasing fragmentation of the overall structure). Environmental and social management. E&S management was one of the most substantially improved performance areas under the program, with PM scores on E&S thematic areas increasing from 8.6 to 14.3 points out of a maximum of 15. The program also provided targeted capacity- building support in E&S areas, both at the central and local levels, and incorporated the program’s E&S system into the country’s procurement framework and guidelines.