IDA18 Mid-Term Review Crisis Response Window: Review of Implementation October 24, 2018 ACRONYMS AND ABBREVIATIONS Fiscal year (FY) = July 1 to June 30 Cat DDO Catastrophe Deferred GFRP Global Food Crisis Response Drawdown Option Program CDB Caribbean Development GNI Gross National Income Bank GRiF Global Risk Financing CERC Contingent Emergency Facility Response Component HFA Humanitarian Food CPF Country Partnership Assistance Framework IBRD International Bank for CPS Country Partnership Strategy Reconstruction and CPIA Country Policy and Development Institutional Assessment ICRC International Committee of CRED Center for Research on the the Red Cross Epidemiology of Disasters ICRR Implementation Completion CRW Crisis Response Window and Results Report Review DaLA Damage and Loss IDA International Development Assessment Association DPF Development Policy IEG Independent Evaluation Financing Group DPL Development Policy Loan IMF International Monetary Fund DPO Development Policy IPF Investment Project Financing Operation IRM Immediate Response DRM Disaster Risk Management Mechanism DSA Debt Sustainability Analysis ISR Implementation Status and EAP East Asia Pacific Results ECCB Eastern Caribbean Central LIC Low Income Country Bank MDTF Multi Donor Trust Fund EM-DAT Emergency Events Database MENA Middle East and North Africa ERL Emergency Recovery MTR Mid-Term Review Lending NGO Non-Governmental EU European Union Organization EVD Ebola Virus Disease PBA Performance-based FAM Famine Action Mechanism Allocations FAO Food and Agriculture PDNA Post Disaster Needs Organization of the United Assessment Nations PEF Pandemic Emergency FCS Fragile and Conflict-affected Financing Facility States PforR Program-for-Results FCV Fragility, Conflict & PLR Performance & Learning Violence Review GCRP Global Crisis Risk Platform PSNP Productive Safety Net GDP Gross Domestic Product Program GFDRR Global Facility for Disaster Reduction and Recovery REDISSE Regional Disease UN United Nations Surveillance Systems UNHCR The UN Refugee Agency Enhancement UNICEF United Nations Children’s RFM Risk Financing Mechanism Fund RRA Risk and Resilience UNISDR United Nations Office for Assessment Disaster Risk Reduction SCD Systematic Country WASH Water Supply, Sanitation and Diagnostic Hygiene SEDRP Somalia Emergency Drought WBG World Bank Group Response Project WFP World Food Programme SIL Specific Investment Lending WHO World Health Organization TABLE OF CONTENTS EXECUTIVE SUMMARY ........................................................................................................... i I. INTRODUCTION............................................................................................................. 1 II. BROAD TRENDS IN THE CRISIS LANDSCAPE ...................................................... 1 III. ROLE AND EVOLUTION OF THE CRW ................................................................... 4 A. ROLE OF CRW IN THE WBG CRISIS TOOLKIT...................................................................... 4 B. CRW EVOLUTION THROUGH THE YEARS ............................................................................. 7 IV. SUMMARY OF CRW UTILIZATION .......................................................................... 9 V. CRW PERFORMANCE AGAINST KEY OBJECTIVES ......................................... 11 VI. CRW CONTRIBUTIONS TO THE WBG CRISIS AGENDA .................................. 24 VII. POTENTIAL POLICY IMPLICATIONS AHEAD .................................................... 26 A. ADVANCING THE PREVENTION AND PREPAREDNESS AGENDA ........................................... 26 B. FACILITATING EARLIER RESPONSE TO SLOWER-ONSET CRISES ......................................... 27 C. REVISITING CRW ACCESS ................................................................................................. 28 D. OTHER ISSUES .................................................................................................................... 29 REFERENCES AND BIBLIOGRAPHY ................................................................................. 38 LIST OF ANNEXES Annex 1: List of CRW Allocations from IDA16 to Present ......................................................... 31 Annex 2: IDA’s Experience with Cat DDOs ................................................................................ 32 Annex 3: Pilot CRW in IDA15 ..................................................................................................... 34 Annex 4: Crisis Management Themes in the Bank’s Country Engagement Products ................. 36 LIST OF BOXES AND FIGURES Boxes Box 1. The Global Risk Financing Facility .................................................................................... 6 Box 2. Exceptional CRW Allocation for Somalia and Deepening Partnerships .......................... 18 Box 3. Strengthening Resilience to Extreme Weather Events in the Pacific Island States .......... 20 Box 4. Ethiopia’s Productive Safety Net Program ....................................................................... 21 Box 5. CRW Allocation to Combat Ebola in West Africa in 2014 .............................................. 22 Box 6. CRW Support for the 2010 Haiti Earthquake ................................................................... 23 Figures Figure 1. Countries Generally Face Higher Frequency of Shocks ................................................. 2 Figure 2. Poor Countries Suffer Disproportionate Impact Relative to Richer Ones ...................... 3 Figure 3. The WBG Has an Extensive Crisis Toolkit for IDA Countries ...................................... 5 Figure 4. Stylized Representation of How the CRW Fits into the WBG Toolkit for IDA Countries ......................................................................................................................................... 6 Figure 5. Demand for CRW Has Been Modest in IDA18 .............................................................. 8 Figure 6. By Disaster Type, the CRW Has Mainly Been Used for Natural Catastrophes............ 10 Figure 7. By Region, Africa is the Largest CRW Recipient in Absolute Terms .......................... 10 Figure 8. MNA is the Largest CRW Recipient by PBA Share ..................................................... 11 Figure 9. CRW Has Supported Interventions across Multiple Sectors, Especially in Social Protection ...................................................................................................................................... 11 Figure 10. IPFs Have the Largest Commitment Volumes for Operations that Receive CRW Funds ............................................................................................................................................. 11 Figure 11. Compared to Overall IDA, CRW-funded Operations Saw Greater Use of Additional and Supplemental Financing ......................................................................................................... 11 Figure 12. The CRW Has Responded to the Most Severe Crises ................................................. 12 Figure 13. The Size of CRW Allocations is Positively Correlated with Disaster Severity .......... 13 Figure 14. Processing Times ......................................................................................................... 14 Figure 15. CRW-funded Operations Have Disbursed Faster........................................................ 15 Figure 16. Sources of Post-disaster Financing in CRW Episodes ................................................ 16 Figure 17. CRW Allocations and PBA Resources........................................................................ 17 Figure 18. CRW-funded Operations are Generally Satisfactory or Moderately Satisfactory ...... 22 Figure 19. Efforts on Prevention and Preparedness are Cost-effective ........................................ 25 EXECUTIVE SUMMARY i. The Crisis Response Window (CRW) provides additional resources to respond to major natural disasters, public health emergencies and severe economic crises, so as to help IDA countries return to their long-term development paths. Piloted in IDA15 and officially established in IDA16, the CRW is the pièce de résistance in IDA’s crisis management toolkit. Crises pose an integral development challenge in IDA countries. The poor tends to suffer disproportionately, as they are likelier to settle in hazard-prone areas and lack the wherewithal to cope with disasters. Crises can also wreak irreversible effects on education and health that entrench the intergenerational transmission of poverty, such as by forcing households to cut healthcare expenditures or withdraw children from school. The CRW is a valuable source of funding to help countries respond to shocks. Overall, IDA is estimated to have provided some US$40 billion in total1 crisis-related financing since IDA16, of which around US$10 billion2 was disaster response financing. Within this, the CRW has allocated US$3.1 billion3 or around 30 percent of IDA’s disaster response since its formal establishment in IDA16. Annex 1 sets out the list of CRW allocations from IDA16 to the present. ii. The CRW has pioneered several “firsts” in IDA’s approach to crisis management. These achievements have been developed and honed through the years with guidance from IDA Deputies and Borrower Representatives (IDA Participants), as well as IDA Executive Directors. Prior to the CRW’s creation, crisis support from IDA was mainly ad hoc and relied on project restructuring or reallocations from country portfolios to provide emergency funding, which entailed trade-offs with core development activities. The lack of a dedicated mechanism to provide surge financing for severe crises was a key gap in IDA’s toolkit. The CRW’s creation therefore marked a paradigm shift in IDA’s crisis response from ad hoc interventions to a more systematic approach. Important strides were also made in implementation arrangements, as evidenced by a two-step governance process that entailed early guidance from the Board on the potential use of CRW resources. The CRW Eligibility Note was another innovation, arising from the Board’s request for a standardized and rigorous framework to assess applications for CRW funding. iii. After around a decade of operation, this paper seeks to review the CRW’s implementation against five key objectives. These are: (i) whether it has responded to the most severe crises; (ii) its timeliness; (iii) the extent of burden-sharing vis-à-vis other development partners; (iv) its implementation flexibility; and (v) how it has helped countries to “build back better” and augment resilience to future disasters. This paper is also anchored within the WBG’s efforts towards a more coherent and strategic approach to crisis management, as set out in the Global Crisis Risk Platform (GCRP) paper discussed by the Board on July 19, 2018.4 Notably, a central tenet of the GCRP is to facilitate a “pivot” to crisis prevention and preparedness. 1 This comprises financing for upstream resilience-building activities as well as post-disaster response. 2 The estimation methodology entailed gathering project metadata (e.g., project title, Project Development Objectives, abstracts, project components and results indicators), and using a taxonomy of crisis-related keywords to search the key project metadata so as to identify crisis-related projects. This methodology is also similar to the one used for the July 2018 Board paper on the Global Crisis Risk Platform (IDA/SecM2018-0126). 3 The Pilot CRW in IDA15 is presented separately in Annex 3. 4 World Bank, 2018. - ii - iv. The review finds that the CRW has generally performed admirably against its core objectives. Overall, the CRW has responded to the most severe crises for which countries have requested support. There is also evidence of reasonable burden-sharing vis-à-vis other development partners, which suggests that usage of CRW resources has been judicious. While operating within the implementation framework set out in successive IDA replenishment agreements, the CRW has exercised flexibility in highly severe cases where there was a clear imperative for forbearance in the application of its rules. A notable example was when it marshaled US$420 million to tackle the 2014 Ebola crisis, back when public health emergencies were not yet under its coverage. Finally, the CRW also made important contributions to help countries to “build back better” after a crisis. v. Consistent with the CRW’s adaptation to evolving crisis demands, there is scope for continued improvement ahead. Efforts are underway to further streamline the processing of CRW requests. The paper also poses some questions on potential future policy directions for IDA Participants’ guidance and feedback. Given the WBG-wide pivot towards prevention and preparedness, a key question is how and whether the CRW should be used more strategically to help galvanize this shift, while preserving its core function as a vehicle for crisis response. In particular, Management asks if IDA Participants would support developing a more full-fledged proposal for IDA19 on having the CRW respond at an earlier juncture to food insecurity episodes that are related to CRW-eligible events. Unlike sudden onset disasters such as earthquakes where the devastating aftermath is immediately clear, food insecurity tends to be a slower onset development and earlier CRW financing could prevent emergent stresses from escalating into more severe catastrophes. vi. Finally, demand for CRW resources has been modest thus far in IDA18, and the present pace of CRW usage suggests that US$1 billion to US$1.5 billion could be untapped. To manage funds for greatest impact, Management proposes to redeploy this potential surplus to other uses. This is also discussed in the IDA18 MTR Chapeau paper.5 5 See the IDA18 MTR paper on, “IDA18 Mid-Term Review Implementation and Results Progress Report. Towards 2030: Investing in Growth, Resilience and Opportunity. Delivering on innovation and transformation, and managing IDA resources for greatest impact.” I. INTRODUCTION 1. Piloted in IDA15 and officially established in IDA16, the CRW is the cornerstone of IDA’s crisis management toolkit. It provides additional resources for recovery and reconstruction in the aftermath of crises, to restore countries to their long-term development paths. CRW funds are additional to IDA country envelopes and help to protect core development spending that could otherwise be jeopardized amid more pressing crisis needs. The CRW’s creation marked a paradigm shift in IDA’s crisis response from ad hoc interventions to a more systematic approach, underpinned by a unique governance arrangement. After around a decade of operation, it is timely to reflect on the CRW’s experience and discern key lessons for the future. 2. This paper is structured as follows. Section II sets out key trends in the broader crisis landscape, followed by Section III which discusses the CRW’s role and its evolution through the years. Sections IV and V assess the utilization of the CRW and how it has fared against its main policy imperatives. Section VI discusses how the CRW has contributed to the WBG’s crisis agenda, and Section VII concludes with potential areas for continued improvement. II. BROAD TRENDS IN THE CRISIS LANDSCAPE 3. Crises pose an integral and urgent development challenge in IDA countries. Many IDA clients face an increasingly complex risk landscape, characterized by multiple and often recurrent hazards that threaten to undermine development gains and progress towards the Sustainable Development Goals by 2030. Between 1950 and 2011, most of the relatively faster growth of high-income countries resulted not from experiencing faster growth but rather from shrinking less, and less often, from crises or wars compared to lower-income countries.6 The Sendai Framework for Disaster Risk Reduction also states that the least developed countries face specific challenges and warrant particular attention given their higher vulnerability and risk levels. 4. IDA clients are facing greater frequency and complexity of shocks (Figure 1). The incidence of natural disasters peaked in the early 2000s but remains elevated. The rate of hydrological, climatological and meteorological disasters has nearly tripled7, and the frequencies of disease outbreaks and the number of refugees are likewise trending up. Global economic losses from disasters average up to US$300 billion a year.8 Economic and financial crises occur irregularly but inflict significant damage. In addition, research9 on tipping points and potential climate impacts on the earth system underscore the critical need to speed up action to address such threats. Food insecurity is another challenge, with around 124 million people across 51 countries facing Crisis levels of acute food insecurity or worse10, and requiring urgent assistance.11 Crises also increasingly entail “compound risks”—multidimensional hazards that span natural disasters, 6 World Bank, 2017a. 7 Thomas and López, 2015. 8 UNISDR, 2015. 9 Nolan et al., 2018 and Steffen et al., 2018. 10 This means Integrated Phase Classification (IPC) Phase 3 and above or equivalent. 11 World Food Programme, 2018 -2- disease outbreaks, economic shocks, food insecurity and conflict—sometimes with cross-border spillovers that propagate and amplify the impact of shocks.12 An example is the 2017 episode in the Horn of Africa and Yemen where drought, conflict and food insecurity interacted to pose a significant and—in some places—continuing threat. Figure 1. Countries Generally Face Higher Frequency of Shocks a. Number of Natural Disasters b. Number of Infectious Disease Outbreaks Source: EM-DAT database and staff calculations. Source: Smith et al., 2014. c. Number of Refugees (Millions) d. Number of Systemic Banking Crises Source: UNHCR. Source: Laeven and Valencia, 2012. Note: Figure 1.a covers six disaster types in the EM-DAT database: geophysical (e.g., earthquakes, volcanic activity); meteorological (e.g., storms); hydrological (e.g., floods); climatological (e.g., droughts); extraterrestrial (e.g., asteroids); and biological (e.g., epidemics). Technological disasters (e.g., industrial accidents) are excluded here. 5. Poor countries also suffer disproportionately compared to richer ones (Figure 2). Crises have an asymmetric impact, with deterioration in development indicators during bad times being much larger than improvement during good times.13 IDA clients face higher costs in percent of GDP from natural catastrophes compared to more developed countries. Lower-income countries 12 World Bank, 2018. 13 Arbache, Go and Korman, 2010. -3- also bear the brunt of mortality from natural hazards14 and pandemics.15 Small islands are especially vulnerable as many face significant exposure to natural hazards and high event frequency. Within countries, poor communities are especially susceptible as they are likelier to settle in hazard-prone areas such as hillsides and near riverbanks, and lack the wherewithal to cope with shocks. Natural disasters are estimated to drive 26 million people into poverty every year.16 Notably, disasters can wreak irreversible effects on education and health that entrench the intergenerational transmission of poverty, such as by forcing households to cut healthcare expenditures or withdraw children from school. Figure 2. Poor Countries Suffer Disproportionate Impact Relative to Richer Ones a. Impact of Natural Disasters in IDA and b. Number of Deaths from Natural Non-IDA Countries Hazards by Income Group (1996–2015) Source: EM-DAT database and staff calculations. Source: UNISDR et al. (2016). c. Expected Annual Losses from Pandemics (% of National GDP) Source: Smith et al. (2017). 14 UNISDR and CRED, 2016. 15 Madhav et al., 2017. 16 Hallegatte et al., 2017. -4- 6. There has been a call for a new modus operandi between humanitarian and development partners.17 IDA’s role is typically not one of first responder, but is rather a complement to national authorities, humanitarian agencies and others which are the primary providers of emergency relief. IDA has therefore focused on strengthening countries’ resilience before crises strike, and thereafter supporting recovery. This demarcation of roles is clear in most CRW cases. That said, there is an increasing confluence of humanitarian and development imperatives, particularly for protracted crises where the delineation between urgent humanitarian relief and longer-term development is less clear. Crises in conflict-affected settings, where the conditions for sustained recovery are weaker, are a case in point. Disease outbreaks are another example, where development and humanitarian actors intervene at a similar stage to contain their spread. At the 2016 World Humanitarian Summit, the UN and World Bank hence committed to a “new way of working” which aimed to bridge the humanitarian-development nexus, including aligning more effectively around collective outcomes. For instance, IDA’s 2017 support to Yemen as part of the regional famine response entailed seeking policy wavers that facilitated disbursement of CRW funds through UNICEF and WHO, as IDA had no on-the-ground presence in the country. III. ROLE AND EVOLUTION OF THE CRW A. ROLE OF CRW IN THE WBG CRISIS TOOLKIT 7. The WBG has developed an extensive crisis toolkit for IDA countries18 (Figure 3). These entail traditional instruments such as investment project and development policy operations, targeted crisis tools like the Contingent Emergency Response Component (CERC), IDA Catastrophe Deferred Drawdown Option (Cat DDO), and market-based mechanisms such as weather derivatives. These are in turn funded by various sources of financing. Core country programming comprises the bulk of crisis-related funding, supplemented by various windows, regimes and Bank-administered resources such as trust funds and regional risk pools. 8. WBG support also spans different phases of crisis management. Prevention and preparedness entail upstream interventions to bolster resilience. Prevention refers to elimination or mitigation of risks, such as by investing in vaccinations and resilient agriculture production systems. Preparedness refers to situations where risks cannot be significantly mitigated, and the focus is on limiting the ramifications, e.g., early warning systems, targeted application of safety net instruments, codifying disaster response protocol and adopting layered risk financing strategies. In this phase, contingent tools such as CERCs, Cat DDOs and the Pandemic Emergency Financing Facility (PEF) constitute pre-arranged financing that can be accessed rapidly should a disaster strike. Next, the response phase involves addressing needs in the immediate aftermath of a crisis, such as restoring basic services and livelihoods. Finally, the recovery phase entails longer- term reconstruction and “building back better” to strengthen resilience to future shocks. After the 2014 Ebola crisis for instance, the IDA Regional Program provided funds to the Regional Disease Surveillance Systems Enhancement (REDISSE) project to, among other things, augment capacity 17 Please see the IDA18 MTR paper on “IDA18 FCV Special Theme Progress for Mid -term Review”. 18 IFC and MIGA also developed initiatives, e.g., in the 2008/09 global crisis, IFC pioneered the Global Trade Liquidity Program and Global Trade Finance Program, while MIGA developed the Financial Sector Initiative. -5- for disease surveillance and epidemic preparedness. These phases are not always distinct, especially in slower-onset events such as drought and disease outbreaks. Figure 3. The WBG Has an Extensive Crisis Toolkit for IDA Countries Source: Adapted from World Bank, 2017b. 9. The CRW is an integral part of this architecture. A few points bear emphasis. First, the CRW is a source of funds to support response and recovery, as indicated by phases B and C in Figure 4. It does not finance upstream resilience-building before a disaster occurs (phase A), as these are regular development activities to be funded by other means such as performance-based allocations (PBA). Humanitarian responses are likewise not under the CRW’s ambit. Second, the CRW is instrument-neutral and its monies can be channeled through Investment Project Financing (IPF), Development Policy Financing (DPF) and Program-for-Results (PforR) financing. Third, it caters to discrete episodes involving natural disasters, public health emergencies and economic crises. It is not intended to finance reconstruction needs arising from war and conflict; for that IDA generally relies on regular or exceptional IDA country allocations as well as trust fund resources.19 Fourth, the CRW complements other elements such as Cat DDOs (Annex 2) and trust funds, including the new Global Risk Financing Facility (Box 1). As a source of surge financing, it also complements reallocations of project funds through CERCs and the Immediate Response Mechanism (IRM). For severe disasters in particular, the different tools often provide additive financing to facilitate a WBG response at scale. CERCs and IRM can also be used to address smaller disasters for which the CRW is not mobilized. 19 The CRW has responded to CRW-eligible disasters (e.g., drought and health-related outbreaks) in conflict-affected environments, but it would not be deployed based on conflict alone. -6- Figure 4. Stylized Representation of How the CRW Fits into the WBG Toolkit for IDA Countries Box 1. The Global Risk Financing Facility (GRiF) The GRiF is a new financing mechanism of US$145 million administered through a Multi Donor Trust Fund (MDTF) that aims to strengthen financial resilience of vulnerable countries. It seeks to complement the existing climate finance and aid architecture, and over time complement financing for other types of crises as well. The GRiF will align with the Vision and Principles of the InsuResilience Global Partnership for Disaster and Climate Risk Finance and Insurance. The GRiF focuses on scaling up and strengthening existing risk financing initiatives, and catalyzing pilot investment approaches not yet explored by other TA-focused programs. Its scope covers natural disasters, extreme weather events and other climate risks, but could over time extend to other shocks such as famine. The GRiF will provide financing to lower barriers for implementation and create the enabling environment for effective risk financing by co-financing WBG investments to: • establish risk financing mechanisms, e.g., investment in start-up or operating costs and up-front capital contributions for risk pools or disaster funds; • reduce the cost of risk financing mechanisms, e.g., co-payment of insurance premiums, buying down interest rates of financial instruments, providing concessional grants to enhance contingent financing arrangements by governments to complement risk transfer solutions; and • link the funding to mechanisms for delivery of effective crisis response and recovery, e.g., adapting safety nets with scalability components, integrating disasters in public financial management systems, or improving claims handling systems for public insurance schemes. There are several potential complementarities between the CRW and GRiF. For instance, the GRiF could provide funds to pay for premiums of catastrophe insurance that complement the CRW as an additional source of funds in the event of disasters. As insurance payouts are based on parametric triggers for tail risks, the CRW could be mobilized for disasters that do not breach these triggers. The GRiF could also eventually fund technical assistance and capacity-building in crisis risk management, areas that clients may not be able to finance from core IDA envelopes given competing and often more immediately visible priorities. Such upstream interventions would in turn strengthen the clients’ ability to use CRW funds more effectively. Finally, the GRiF could pilot innovative approaches which, where proven effective, could be mainstrea med into the CRW and IDA’s conventional toolkit over time. -7- B. CRW EVOLUTION THROUGH THE YEARS 10. The CRW began as a pilot in IDA15 to respond to the 2008/09 global economic and financial crisis. Back then, IDA countries faced the formidable “triple challenge” of mitigating the fallout from the Global Financial Crisis, returning to their pre-crisis development trajectories, and regaining lost ground. It was recognized that there was a gap in the international aid architecture to systematically help low-income countries to cope with large shocks. This sentiment was echoed in the 2009 G20 and Development Committee20 communiques, with the latter calling for a “new crisis response facility in IDA to protect LICs from future crises”. The proposal was discussed at the IDA15 Mid-Term Review (MTR) and approved by IDA Executive Directors in December 2009. US$1.4 billion of Pilot CRW funds were committed through 75 operations to 46 countries. 11. IDA16 saw the permanent establishment of the CRW, which marked a paradigm shift in IDA’s crisis response from ad hoc interventions to a more systematic approach. IDA’s crisis support had hitherto entailed restructurings and reallocations that diverted resources from long-term development, exceptional allocations that reduced funding for other IDA countries, or trust funds that took time to set up. The CRW was thus envisaged to strengthen the predictability and timeliness of IDA’s crisis response, and enhance its capacity to participate in global disaster recovery efforts.21 A transparent governance process was designed, which entails seeking the Board’s early guidance22 on potential use of CRW resources before projects are prepared and submitted for approval. The CRW Eligibility Note was another innovation, arising from the Board’s request for a more systematic way to assess applications for CRW funds. Several Participants at the second IDA16 replenishment meeting also recommended that the CRW be accompanied by efforts on trust fund consolidation23, envisaging that donors could make additional contributions to the CRW that are earmarked for specific crises, instead of going through trust funds. Finally, to address the concern that CRW resources could sit idle if unused, it was agreed that Management would propose reallocations of unused funds at the time of IDA MTRs. 12. In IDA17, strong demand led to a mid-replenishment increase in CRW resources and expansion of its scope to cover public health emergencies. The initial CRW allocation of US$900 million was almost fully committed in the first year of IDA17 due to exceptionally high demand. At the IDA17 MTR, another US$900 million was therefore distributed to the CRW, which was likewise fully allocated by end-IDA17. Overall, IDA17 saw multiple severe crises that received significant CRW funding—notably, US$420 million for the Ebola epidemic in West Africa, US$360 million for famine response in the Horn of Africa and Yemen, and US$300 million for the Nepal earthquake. Another key development was the expansion of CRW coverage to include public health emergencies, a change prompted by the 2014 Ebola epidemic. IDA Participants noted that the CRW was critical to the WBG’s response and lauded it as “a regional and global public good with positive externalities in containing and mitigating a fast-moving epidemic”.24 20 World Bank, 2009a. 21 World Bank, 2010. 22 The practice of consulting the Board early was later adapted to other IDA windows like the Refugee Sub-window. 23 World Bank, 2010. 24 World Bank, 2015. -8- 13. A further modification was introduced in IDA18 such that IDA financing terms would more quickly reflect the impact of a disaster on a country’s risk of debt distress. For countries where damages and losses from a natural disaster exceed one third of GDP, the IDA financing terms in that fiscal year would be adjusted to either 50 or 100 percent grants if warranted, based on an updated Debt Sustainability Analysis (DSA) following the disaster. In addition, to simplify and expedite the work of country units with client countries that are experiencing crises, a one- page guide was developed, including templates for how to apply for CRW funding. 14. Demand for CRW resources has been modest in IDA18 compared to earlier replenishments (Figure 5). To date, US$350 million or 12 percent of CRW resources have been allocated in IDA18, with the largest being US$200 million for a cholera outbreak in Yemen. The subdued demand is due to the fact that there have been very few severe crises that meet the CRW eligibility criteria so far in IDA18. Moreover, in some cases countries have been able to access other sources of funding. For example, following the Ebola outbreak in the Democratic Republic of Congo in May 2018, the PEF deployed a US$12 million grant from its cash window and another US$80 million of credits from CERC activation was also available. Section VII discusses the redistribution of unused CRW resources. Figure 5. Demand for CRW Has Been Modest in IDA18 Source: CRW Eligibility Notes and staff calculations. Notes: 1/ The shaded parts of the columns show CRW allocations as at end-September of the second FY in each IDA replenishment. 2/ The total height of the IDA17 column—shaded and unshaded—reflects the initial CRW size of US900 million and the subsequent IDA17 MTR top-up of another US$900 million. -9- IV. SUMMARY OF CRW UTILIZATION 15. This section presents key analytics on CRW utilization from IDA16 to the present. Since its official establishment in IDA16, the CRW has allocated US$3.1 billion across 24 crisis episodes, contributing some 30 percent of IDA’s overall disaster response financing during this period. The IDA15 Pilot CRW is presented separately in Annex 3.25 16. The CRW has mainly been used for natural catastrophes. Natural disasters comprised 73 percent of utilization across 20 cases (Figure 6), with the largest ever CRW allocation being US$508 million for the Haiti earthquake. The CRW has been deployed twice for public health emergencies, and twice for economic shocks. Annex 1 sets out the list of CRW allocations since its formal establishment in IDA16. 17. Africa is the largest CRW recipient in absolute dollar terms, while the Middle East and North Africa (MENA) ranks first by share of PBA (Figures 7 and 8). That Africa is the biggest beneficiary is no surprise given its high concentration of IDA countries and several large disasters with multi-country spillovers—notably the 2014 Ebola epidemic, the 2016 El Niño drought, and the drought and famine crisis in 2017. MENA is the largest recipient by percent of PBA (Figure 7), mainly due to Yemen.26 All in all, around 80 percent of CRW beneficiaries were IDA-only clients and almost half were fragile and conflict-affected states (FCS). The largest beneficiary per capita is Dominica, which received US$680 of CRW funds per capita or a total of US$50 million after Hurricane Maria. Dominica suffered extensive devastation after the Category 5 hurricane wreaked damages and losses of 226 percent of GDP. 18. CRW funds have financed programs mainly in the social protection, health, water, waste & sanitation, as well as transportation sectors (Figure 9). Social protection programs in particular were a common vehicle for channeling CRW funds as they could be rapidly scaled up during crises and often had pre-identified beneficiaries that targeted vulnerable groups. Overall, the sectoral composition accords well with the early objective in IDA15 to use the CRW to protect core spending primarily in health, education, safety nets, infrastructure, and agriculture.27 19. Similar to overall IDA financing, the bulk of CRW funds has been channeled through IPFs (Figure 10). Different types of instruments serve different purposes, with varying parameters and requirements. As the CRW has been mainly used for natural disasters, IPFs are an appropriate choice for activities such as infrastructure reconstruction. Notably, the share of commitments channeled through additional and supplemental financing operations in the CRW is higher than that under the overall IDA portfolio (Figure 11). Such routes allowed faster and more effective delivery of the crisis response, as teams could tap into existing implementation channels. 25 The Pilot CRW is presented separately in Annex 3 due to several idiosyncrasies of that phase. In IDA15, the Pilot CRW was used to mitigate the impact of economic shocks following the 2008/09 global crisis, while usage thereafter has mainly been for natural disasters. The CRW allocation methodology in IDA15 was also different. Finally, the CRW Eligibility Notes— which provide standardized information on assessments of CRW requests, much of which is used in this paper—was instituted subsequently. 26 Yemen accessed the CRW twice in short succession in 2017. See Annex 1. 27 World Bank, 2009b. - 10 - Figure 6. By Disaster Type, the CRW Has Mainly Been Used for Natural Catastrophes CRW Allocations by Disaster Type Outer ring shows top five largest CRW allocations Source: CRW Eligibility Notes and staff calculations. Figure 7. By Region, Africa is the Largest CRW Recipient in Absolute Terms CRW Allocations by Recipient Outer ring shows top five CRW country recipients Source: CRW Eligibility Notes and staff calculations. - 11 - Figure 8. MNA is the Largest CRW Figure 9. CRW Has Supported Recipient by PBA Share Interventions across Multiple Sectors, Especially in Social Protection Figure 10. IPFs Have the Largest Figure 11. Compared to Overall IDA, Commitment Volumes for Operations that CRW-funded Operations Saw Greater Use Receive CRW Funds of Additional and Supplemental Financing Source: CRW Eligibility Notes and staff calculations. Note: In Figures 9 to 11, the columns marked “CRW” show commitments of IDA operations that received CRW funds. Where these were Additional or Supplemental Financing operations with distinct p-numbers, the data only shows such financing and not that of the parent operations. V. CRW PERFORMANCE AGAINST KEY OBJECTIVES 20. This section assesses how the CRW has fared over the years against its key objectives. It covers five areas, namely: (i) whether the CRW has responded to the most severe crises; (ii) its timeliness; (iii) the extent of burden-sharing vis-à-vis other development partners; (iv) its implementation flexibility; and (v) how it has helped countries to “build back better”. Several points bear emphasis here. First, the assessment takes into account the CRW’s original intent as “a systematic mechanism to provide timely, additional and predictable financing to IDA countries - 12 - impacted by severe exogenous shocks”.28 Second, besides evaluating whether the CRW has met its ambitions, it is also important to bear in mind what it is not intended to do. Third, as the CRW’s role is in response and recovery to restore countries to their long-term development paths, it is not always germane to compare its experiences with that of other actors such as humanitarian agencies. i. Has CRW support been for the most severe disasters? 21. The CRW has generally responded to the most severe crises that countries requested help for (Figure 12). The size of CRW allocations also shows a positive correlation with the extent of disaster severity. This relationship is evident across different metrics, both as a percent of GDP and in per capita terms (Figure 13). Moreover, the CRW responded to three out of the five crises with the highest fatalities, namely, the earthquakes in Haiti and Nepal, and the 2014 Ebola epidemic.29 Figure 12. The CRW Has Responded to the Most Severe Crises Top Severe Crises in IDA countries, 2010–17 (Damages as % of GDP) Shaded Columns = Received CRW Source: CRW Eligibility Notes, EM-DAT database and staff calculations. Notes: 1/ Years in parenthesis reflect the period of the disaster, which may not coincide with the year of the CRW allocation. 2/ Dominica did not request CRW support for Tropical Storm Erika, while Pakistan used its own resources and funding from the Asian Development Bank to meet its recovery and reconstruction needs after the 2010 flood. Zimbabwe could not receive regular IDA funding due to its arrears status. 3/ Data is from CRW Eligibility Notes where available, otherwise supplemented by data from EM-DAT and the Regions. Health disasters are excluded as EM-DAT does not have damage and loss data for this category. 28 World Bank, 2012. 29 The other two episodes were the Somalia drought in 2010–11 (Somalia could not receive regular IDA funding due to its arrears status), and the India flood of 2013. - 13 - Figure 13. The Size of CRW Allocations is Positively Correlated with Disaster Severity a. CRW Allocations (% of GDP) b. CRW Allocations Per Capita Source: CRW Eligibility Notes, EM-DAT and World Economic Outlook database. Note: Excludes outliers (e.g., Dominica due to very high Damages and Losses) and cases with incomplete data (e.g., Somalia). ii. Has the CRW responded in a timely manner? 22. Management has responded quickly to crises, deploying teams to the field to help coordinate post-crisis efforts. In the case of severe natural disasters for instance, IDA mobilizes quickly to set in train a sequence of events. As IDA is not a first responder, its initial point-of- entry is typically to send a team, within a week of receiving a request from the government, to support a country-led rapid assessment and/or Post Disaster Needs Assessment (PDNA) in coordination with the EU, UN and other development partners. The PDNA is a thorough analysis of the physical damages, economic losses and the costs of meeting recovery needs, and typically takes up to six weeks to complete. It is an important input which forms the basis for assessing CRW eligibility and the size of such support. Around 90 percent of CRW responses to natural disasters and public health emergencies have been informed by some type of needs assessment.30 The PDNA also informs discussions at donor pledging conferences, where the outcome is another useful input for assessing CRW eligibility31. After sufficient information has been collected to determine CRW eligibility, an in-principle Management decision on the use of CRW resources is made within days, and Management then signals this intention to the Board. The median time elapsed between the disaster date and the Board Technical Briefing is 86 days (Figure 14.a). In 30 In some cases, a formal PDNA or DaLA (Damage and Loss Assessment) was not undertaken, for instance due to the country being in active conflict. In Yemen for instance, the Bank instead undertook a Dynamic Damage and Needs Assessment (DNA) that focused exclusively on an estimate of physical recovery and reconstruction costs in a subset of sectors and cities. 31 The outcome from the pledging conference helps to illuminate the extent of burden-sharing between IDA and other development partners. It also provides a sense of the size of the financing gap which the CRW could help to fill as a last resort to complement other efforts. The concepts of burden-sharing and last resort are discussed in paragraphs 25 to 27. - 14 - cases of extreme severity, this has been as short as around two weeks, with the acknowledgement that information not available at that juncture would be subsequently provided to the Board. After obtaining the Board’s informal agreement at the Technical Briefing to deploy the CRW, teams would prepare the relevant operations under accelerated emergency procedures and submit these for Board approval on an absence-of-objection basis. It also bears noting that IDA would mobilize a package of measures which could include reprogramming of core IDA funds through fast- disbursing CERCs and project restructuring, with the CRW as a tool of last resort. Figure 14. Processing Times a. Median Time Elapsed for Processing Cycle of CRW-financed Operations Source: CRW Eligibility Notes and staff calculations. *Note: Stage 1 of Figure 14 excludes slower-onset disasters such as droughts and health-related outbreaks, as the starting dates of such events are difficult to determine. b. Median Time from Approval to First c. Median Time from Approval to First Disbursement Disbursement Source: World Bank, Business Intelligence. Source: World Bank, Business Intelligence. 23. CRW-funded operations have also disbursed faster than standard IDA operations. The median times from approval to disbursement is shorter for CRW-funded operations, in line with its crisis response mandate (Figures 14.b and 14.c). In addition, the percentage of disbursed funds from CRW operations was also higher than for whole-of-IDA operations approved in the - 15 - same period. This observation holds true when differentiated by instrument and across replenishment cycles (Figure 15). Figure 15. CRW-funded Operations Have Disbursed Faster Percentage of commitments (by IDA cycle) that have disbursed Column = CRW-funded; Diamond = Whole-of IDA Source: Business Intelligence; Analysis for Office. Note: PforRs are not depicted as there is only one CRW-funded PforR operation. 24. The CRW’s role is to support recovery and reconstruction, rather than to act as first responder. It thus complements the work of national authorities, humanitarian agencies and other partners which are the primary providers of emergency relief. As a corollary, it can be challenging to find suitable benchmarks for determining if the CRW processing times are considered timely in the context of its role. iii. Has there been appropriate burden-sharing and was CRW a last resort? 25. The concepts of fair burden-share and last resort are important considerations that go into CRW allocation decisions. The size of CRW allocations is based on multiple factors such as disaster severity, country capacity, the WBG’s burden-share vis-a-vis funds from other partners, and the scope for reallocations or deploying unutilized core IDA resources. Fair burden-share and last resort do not, however, mean that the CRW seeks to minimize or apply a pre-determined burden-share, or wait till it is the last entity to provide funds. Rather, given its goal to restore countries to their long-term development paths, considerations of fair burden-share and last resort are balanced against the need to reduce disruptions to clients’ core development portfolios. In practice, Bank teams are required to assess if there are other readily available resources for crisis response before tapping the CRW. This entails demonstrating that externally, “CRW resources are… part of a concerted international effort32”; and internally, that there is limited or no space to use core IDA resources. Such assessments form part of the CRW Eligibility Notes submitted to the Board and have been consistently applied. The analysis below uses a similar approach: first by assessing the funding shares vis-à-vis external sources; and second by exploring if, within the WBG, the CRW has stepped in when there is limited scope to use a country’s core IDA funds. 26. CRW vis-à-vis External Resources: For disasters which mobilized the CRW, funds from the CRW have comprised 14 percent of overall crisis response (Figure 16). The figure rises to 35 32 World Bank, 2012. - 16 - percent if total WBG response—including from sources such as PBA, CERC activations and WBG-administered trust funds—is counted. The bulk of the “Other Donors’ category comprised of bilateral and multilateral donors, while insurance mechanisms contributed relatively modest payouts mainly in the Pacific and Caribbean regions. Overall, it is difficult to establish what percentage from the CRW would constitute “fair” burden-sharing, especially given the varied factors that influence the size of CRW allocations. Data uncertainties, such as to what extent the pledges under the “Other Donors” category have actually materialized, further complicate the picture. To illustrate, if “Other Donor” pledges were excluded and one were to only count commitments in that category, the shares of the CRW and WBG contributions rise to 22 percent and 56 percent respectively. On balance, the evidence points to reasonable burden-sharing, with CRW support complementing significant financing from other WBG sources and external donors. Figure 16. Sources of Post-disaster Financing in CRW Episodes Source: CRW Eligibility Notes. Note: This presents a point-in-time snapshot at the time of the CRW Technical Briefings to the Board, as there is no known data source that systemically tracks the amounts and sources of financing across different crises, including for non-WBG funds. The amount of donor pledges shown can vary from the commitments that eventually materialize. 27. CRW vis-à-vis PBA: Within the WBG, countries with greater space in their PBA portfolios to respond to crises have received smaller CRW allocations per capita, and vice versa (Figure 17). The x-axes in Figure 17 measure uncommitted PBA funds33, adjusted for crisis severity using damage and loss data. Therefore, the x-axes indicate the adequacy of available PBA resources for crisis response. A negative correlation is observed across both non-small and small states. This suggests that CRW usage has been judicious—where there is larger scope to respond using PBA resources, the CRW has provided relatively modest support.34 33 If a crisis strikes near the end of an IDA cycle for instance, there may not be much PBA resources left for response. 34 It should be noted that the CRW also seeks to protect core development spending, so the objective is not to mobilize CRW funds only when PBA resources have been depleted. - 17 - Figure 17. CRW Allocations and PBA Resources a. Non-small States b. Small States Note: Uncommitted PBA resources were as at the date of the CRW Technical Briefings to the Board. For the 2010 Haiti earthquake, the uncommitted PBA as at the date of the first CRW-funded operation was used instead. The CRW allocation to Haiti was agreed by IDA Participants as part of the IDA16 replenishment discussions. iv. Has the CRW been flexible? 28. For very severe crises, the CRW has been used flexibly to facilitate a response at scale. While the CRW has generally operated within its implementation framework, such flexibility is deemed necessary in cases of exceptionally high disaster impact. 29. On financing terms, the CRW exercised forbearance under extreme circumstances. For instance, the CRW funding terms for Dominica were softened from credit to grant terms given the significant damages and losses from Hurricane Maria that totaled 226 percent of GDP. IDA18 also now permits a mid-FY adjustment of IDA financing terms based on an updated DSA, in the event of severe natural disasters with damages and losses of over a third of GDP. Tonga was the first to benefit from this change.35 Another example was the CRW allocation to Somalia as part of the 2017 regional famine response package. Significant innovation and flexibility was exercised in that case, with the CRW resources channeled through the International Committee of the Red Cross (ICRC) and the Food and Agriculture Organization of the UN (FAO), as Somalia could not directly receive IDA resources because of its arrears situation (Box 2). 35 After Tonga was struck by Tropical Cyclone Gita that wreaked damages and losses of 38 percent of its GDP, Management approved a shift of Tonga’s IDA financing terms from a 50 percent grant, 50 percent credit mix to 100 percent grant financing. - 18 - Box 2. Exceptional CRW Allocation for Somalia and Deepening Partnerships Somalia was on the brink of famine in 2017 primarily due to severe drought. As of April 2017, an estimated 6.2 million—half its population—were acutely food insecure and in urgent need of humanitarian assistance, with over a million children projected to be acutely malnourished. The Federal Government of Somalia however has been in non-accrual status with the Bank since 1991, and IDA had very limited avenues to engage in the country given its policy of not financing into arrears. Recognizing the extreme circumstances, the Board approved a US$50 million CRW allocation to support famine response in Somalia. To avoid setting a broader precedent given Somalia’s arrears status, the CRW allocation embedded a set of carefully ring-fenced criteria to emphasize the exceptionality of this case. In particular, it was positioned as part of a regional famine response where excluding Somalia could create displacement spillovers to neighboring countries, thereby risking the success of IDA’s intervention for the whole region. CRW funds were channeled through ICRC and FAO, and demonstrated IDA’s comparative advantage in fostering partnerships. The Somalia Emergency Drought Response Project (SEDRP) represents the Bank’s first collaboration with ICRC through IDA funding, given Somalia’s non -accrual status, thereby allowing ICRC to secure and utilize critically-needed additional resources at the height of the crisis. The experience showed that both institutions could work together in a crisis, thus bridging the perceived humanitarian-recovery-development divide in a principled and pragmatic fashion. ICRC’s response supported both r esilience-building and early recovery. The partnership with FAO was also critical in extending the impact of CRW funds, and helped forge a closer working relationship between FAO and the government. Both ICRC and FAO have demonstrated flexibility and adaptability that allowed the project to respond to on-the-ground conflict dynamics for maximum impact. The SEDRP outperformed its original objectives. Across both ICRC and FAO components, results achieved include: improved food security and Water Supply, Sanitation and Hygiene (WASH) access for over 700,000 people (over half of which are female), distribution of cash grants to over 55,000 households, vaccination of 8.5 million livestock, access to drought-related health services for over 665,000 people, and rehabilitation of 100 vital water infrastructures completed or underway. The project exceeded original beneficiary targets for multiple sub- components due to efficiencies and cost savings. 30. The CRW also demonstrated operational flexibility. For instance, in the case of the 2014 Ebola crisis, Management moved quickly to consult with shareholders in order to deploy US$420 million of CRW resources to Guinea, Liberia and Sierra Leone. The extreme and rapidly evolving situation warranted an exceptional CRW response, including on grounds of providing an urgently- needed public good to help prevent the epidemic from spreading beyond the three affected countries. In IDA17, public health emergencies were formally included as an eligible category in the CRW criteria. The Somalia experience in Box 2 was also testament to IDA’s core strength in developing partnerships to channel support through third parties where IDA has no on-the-ground presence. Board waivers allowed FAO and ICRC to receive CRW funding to benefit Somalis threatened by famine, with these organizations helping to implement the IDA projects. 31. There is nonetheless room for improvement. A Value Stream Mapping exercise was conducted with frontline departments to assess how CRW processing could be made more agile. CRW guidelines have since been simplified, with a one-page guide and templates made accessible on the Bank’s intranet and systematically shared with country teams to expedite the process. Teams also felt that preparations for the CRW Technical Briefings could divert attention from on-the- ground crisis management. This Technical Briefing was originally conceived to apprise the Board of the crisis situation and is accompanied by a CRW Eligibility Note from Management to convey its intention to use CRW resources. This gives staff an informal green-light from the Board to - 19 - proceed with project preparation, with the confidence that CRW funds would be forthcoming. Over time, the considerations underpinning CRW eligibility have in practice been streamlined and encapsulated in the CRW Eligibility Note that is submitted to the Board ahead of the Technical Briefing. To further simplify the process, the Technical Briefing could no longer be a default practice but rather be held upon the Board’s request. CRW Eligibility Notes would continue to be submitted to the Board, both as a standardized way to assess CRW requests and to convey Management’s intention to deploy CRW resources. The Board would approve CRW-funded operations on an absence-of-objection basis, as is current practice. The foregoing strikes an appropriate balance between enhancing agility while preserving the CRW’s governance arrangements. It also reflects that the CRW process has matured and its rigor can be maintained using a lighter-touch approach. v. How has the CRW helped countries to “build back better” and increase resilience to crises? 32. Building back better is a key tenet of the CRW, as crises present important opportunities as well as political and social impetus to strengthen resilience to future shocks. Besides helping countries tide over catastrophes, the CRW also supports building back better to prevent recurring devastation that undermines development gains. The design of crisis responses should therefore lay the foundations towards this longer-term objective. As a vehicle for crisis response, CRW funding also needs to be supplemented by other resources such as domestic finances, PBA and development partner contributions to underpin the path towards recovery. 33. A recent paper on “Building Back Better” characterizes such efforts as supporting a stronger, faster and more inclusive recovery.36 A stronger recovery helps countries withstand future shocks by better managing risks, such as through locating new buildings outside of flood zones and building structures to resist high winds. A faster recovery enables afflicted populations to restore their incomes and assets more quickly. Finally, a more inclusive recovery ensures that the poorest and most vulnerable can access the support they need to rebuild their lives and avoid the long-term consequences of crises, such as those caused by health issues and disability, loss of education or the inability to save or borrow to replace lost assets. This is key to prevent poor people from falling into poverty traps that can magnify and entrench the impact of disasters. 34. Overall, the CRW has provided valuable support to help countries to build back better. CRW-funded operations have generally integrated elements of resilience-building and several showcased innovative technologies. CRW-funded operations were also an important platform for IDA to demonstrate its value-added in crisis management. For instance, the CRW provided US$200 million37 to the Earthquake Housing Reconstruction Project in Nepal. As most houses destroyed by the earthquake were traditional brick or stone structures built with mud mortar, the project focused on restoring houses with multi-hazard resistant features. IDA resources for this project were complemented by a MDTF whose partners include the US, UK, Switzerland and Canada, while Japan and India provided parallel financing. The CRW-funded operation alone 36 Hallegatte et al., 2018. 37 The total CRW allocation to Nepal was US$300 million, of which US$200 million went to the Earthquake Housing Reconstruction Project, US$50 million to the Second Financial Sector Stability Credit, and the remaining funds for other critical recovery and reconstruction needs. - 20 - aimed to rebuild 154,000 homes with multi-hazard resilient features, of which 40 percent has been completed. A key feature of this project was that housing grants are disbursed directly into the beneficiaries’ bank accounts. The project also assisted the authorities to develop criteria for extending special support to the poorest who were disproportionately affected. 35. IDA has developed strong partnerships with actors both international and local. For instance, it forged valuable collaborations in Yemen with UNICEF, WHO and local service delivery institutions for the implementation of the CRW-funded response. As IDA has no on-the- ground presence in Yemen, these partnerships enabled IDA to provide 640,665 people with the largest treatment for cholera out of a million suspected cases. CRW support also allowed IDA to provide 1.5 million households, or over eight million individuals, with cash transfers that were critical to combat food insecurity in Yemen. Another example was the response to Hurricane Maria in Dominica, where IDA coordinated the PDNA in conjunction with the UN, the Eastern Caribbean Central Bank (ECCB), the Caribbean Development Bank (CDB) and the EU. The PDNA results showed significant devastation to Dominica and helped mobilize support from other partners. IDA’s close consultation with the government also led to quick buy-in and consensus on the priority areas for support. 36. Boxes 3 to 6 present more examples on how the CRW has enhanced resilience- building, the results achieved and lessons learnt. It should be noted that attribution of results specifically to CRW financing is difficult. This is because CRW-funded operations are often combined with other interventions and/or reforms to address a crisis, and could also have mixed funding sources that include PBA and trust funds, for instance. Box 3. Strengthening Resilience to Extreme Weather Events in the Pacific Island States The Pacific island states face many geographical and structural challenges such as remoteness, geographical dispersion, small size, limited natural resources and thin capacity. They are also significantly exposed to natural disasters such as cyclones, earthquakes, tsunamis and floods. Recovery and reconstruction needs often far exceed the financial capacity of Pacific islands, resulting in sizable financing gaps. They also have limited or no access to international capital markets. Since IDA16, the Pacific islands have accessed the CRW six times for a total of US$115 million to respond to cyclones and floods. All CRW-funded projects have a strong emphasis on building back better. Notably, following Tropical Cyclone Gita in February 2018, the CRW provided a US$10 million grant to the Tonga: Pacific Resilience Project under the Pacific Resilience Program which supports building back stronger, including through school recovery and reconstruction, retrofitting existing school buildings —which often serve as shelters during disasters—to withstand greater windspeeds, and resilience-building for critical infrastructure. Another example is the Samoa Agriculture and Fisheries Cyclone Response Project which supported farmers and fishers through vouchers and grants after Tropical Cyclone Evan in December 2012, to restore productive capacity and help build back faster. This project was also able to—for the first time in the Pacific —introduce innovative technology to provide e-vouchers that enabled the project to reach over 7,300 subsistence farmers or fishers, with 95 percent of respondents indicating that they had increased production due to assistance from the project. The experiences of the Pacific islands have yielded important lessons. Resources for project preparation and implementation support need to be set realistically. The rollout of the e-voucher system in Samoa for instance did take longer than anticipated in light of limited capacity in the implementation agency. The lessons also underscore the value of IDA providing extensive hands-on support throughout project implementation. Importantly, after the exigencies of a crisis have subsided, there are opportunities to ensure that stronger Disaster Risk Management (DRM) be mainstreamed into national frameworks. In Samoa, Disaster Risk Management Strategies for agriculture - 21 - and fisheries were prepared through support from the IDA project and have been incorporated into the Samoan National Disaster Risk Management Plan. Box 4. Ethiopia’s Productive Safety Net Program (PSNP) Ethiopia has faced recurrent episodes of drought and food insecurity which threaten to undermine its development gains. The 2015/2016 El Niño drought was a case in point. Up to 22 percent of households surveyed reported distress asset sales to meet basic food needs, and livestock holdings fell by nearly 50 percent between 2014 and 2016 in hard-hit areas. Rural households suffer protracted repercussions, taking an average of four years to recover. The government is undertaking significant efforts on prevention and preparedness, such as through the Resilient Landscapes and Livelihoods Project and the Livelihood Resilience and Commercialization Project in Pastoral and Agro-Pastoral Areas. The Productive Safety Net Program (PSNP) is Ethiopia’s first line of defense to safeguard the livelihoods of the poorest people in drought-afflicted areas. It provides regular cash or food transfers to people in areas most affected by drought. It is designed to scale up in response to drought, by extending the duration of support to existing clients or broadening its reach to more beneficiaries. While safety nets do not directly reduce the risk of crises occurring, they are an important part of resilience-building as they help people to absorb, cope with and recover from disaster impacts. The CRW has been a vital source of support, providing a total of US$270 million on three occasions for drought (Annex 1). In 2012, US$70 million of CRW funds supplemented the PSNP Drought Risk Financing Mechanism (RFM). The intervention through the PSNP RFM occurred within two months, compared to an average response time through the humanitarian system of eight months. It supported an additional 3.1 million people outside of the regular PSNP caseload, and extended the duration of transfers for 6.5 million existing beneficiaries. This response was widely credited with alleviating the worst impacts of the drought. The next CRW allocation of US$100 million in 2016 provided additional assistance to 7.2 million PSNP beneficiaries, which helped them survive the El Niño induced drought. The lessons from 2016 helped improved subsequent response. It addressed inefficiencies in how the PSNP and Humanitarian Food Assistance (HFA) tackled drought, improving the ‘continuum of response’. This included strengthening coordination between development and humanitarian partners. Procedures for beneficiary selection and grievance redress were consolidated, and emergency response implementation is now based on a single, integrated, PSNP-HFA plan. Information systems were also expanded to cover non-PSNP districts, to improve transparency when delivering benefits in a crisis. In 2017, the CRW provided assistance to about 2 million drought- affected people outside of regular PSNP areas. The CRW hence had an important demonstration effect on the transparency, accountability and efficiency of government systems. This in turn encouraged several development partners to channel about US$80 million for emergency response through the government’s contingency budgets and systems. 37. Most CRW-funded operations are rated as Satisfactory and Moderately Satisfactory (Figure 18). Given that the CRW was only formally established in IDA16, it is not surprising that only 12 operations have closed and been rated by the Independent Evaluation Group (IEG) thus far—seven in IDA16 and five in IDA17.38 The Implementation Completion and Results Report Review (ICRR) ratings ascribed by the IEG generally fall into the Satisfactory and Moderately Satisfactory range. 38 See Annex 3 for information on the performance of CRW-funded operations in the IDA15 Pilot CRW phase. - 22 - Figure 18. CRW-funded Operations are Generally Satisfactory or Moderately Satisfactory a. IEG ICRR Ratings b. ISR Development Objective Ratings* * Only includes projects for which no IEG evaluation was performed. ** Does not sum to 100% as one IDA17 project has no ISR. Note: The classification by IDA cycle reflects ratings of CRW-funded operations that were approved in that IDA cycle. The IEG ascribes a single ICRR rating that covers the parent operation and any Additional/Supplemental financing. The ICRR rating is recorded under the parent p-number, and the parent operation may not fall within the IDA cycle as shown. Box 5. CRW Allocation to Combat Ebola in West Africa in 2014 The CRW provided a US$420 million allocation to Liberia, Sierra Leone and Guinea, the three countries that were hit hardest by the Ebola Virus Disease (EVD) epidemic. Notably, US$117 million was disbursed in just nine days after approval to the countries and implementing agencies. CRW funds helped to contain the spread of infections, and supported the governments in strengthening the resilience of their public health systems. At the height of the crisis, the CRW supported Ebola education campaigns; expedited lab testing, surveillance, and safe burials; secured additional health workers via work arrangements and provision of training and financial incentives such as hazard pay; provided medical supplies and treatments; and rendered assistance to communities such as through cash-for-work projects. After the successful control of the outbreak, the use of CRW funds shifted focus towards building resilient public health systems and strengthening preparedness. Activities covered included health workforce planning and training; essential infrastructure including isolation centers, triages and housing units; as well as epidemic preparedness and response. The EVD response was planned and implemented with the UN, WFP, WHO agencies, the African Union, private hospitals and communities. Working with these partners, the WBG supported the recruitment, training and deployment of more than 1,300 qualified foreign health workers to bolster the response in Guinea, Liberia and Sierra Leone. It also supported mobile money systems which allowed domestic workers to stay in their jobs and get paid in a timely and accountable manner. A key lesson was to keep an eye on medium-term goals even in the face of a severe crisis. IDA’s EVD program showed that it may be possible to “kill two birds with one stone” in some instances by ensuring compatibility of emergency response with ongoing development efforts. Examples include the boost to longer-term rural transport connectivity via the emergency effort to improve access to EVD facilities, and the bolstering of health standards and personnel training in the EVD emergency program that have contributed to longer-term strengthening in health care delivery. - 23 - Box 6. CRW Support for the 2010 Haiti Earthquake After the 2010 Haiti earthquake, the CRW provided US$508 million to support reconstruction and longer- term restoration of capacity in the country. The size and breadth of the CRW allocation was atypical, given the crisis magnitude and the depth of Haiti’s preexisting development challenges. The Bank’s support complemented efforts from partners such as the InterAmerican Development Bank, the UN, the US, EU and Canada. All CRW-funded interventions integrated resilience-building in some form. The CRW enabled building back stronger with the development of a new building code with earthquake- and hurricane-resistant standards, retrofitting of critical transport infrastructure and better land-use planning. Innovative tools—such as high- resolution digital elevation models, flood models, the first Atlas of multi-hazard maps and data to track population movements—were developed. IDA also helped build back more inclusively, including through support to raise farmers’ productivity and rebuild assets. To prevent human capital erosion, signif icant resources were spent on maintaining and then expanding access to primary education, and on health services. Haiti faced extremely challenging implementation conditions in the aftermath. Its economy had collapsed and the public sector faced severe capacity constraints, including widespread fatalities. 19 ministries were rendered non-operational. In addition, a succession of political crises further delayed recovery efforts. While waiting for CRW resources to become available mid-2011, early response was supported with IDA15 funds approved within 10 weeks of the disaster. Despite difficult conditions, the enormous needs and a strong drive for response and recovery allowed the Bank—working with the government and civil society partners—to frontload the CRW allocation, with 60 percent of CRW funding committed by the midpoint of the IDA16 cycle, and 80 percent committed by the end of the second year. The CRW was pivotal to Haiti’s recovery and reconstruction, and contributed to these results: • Infrastructure reconstruction: provision of rental subsidies for 50,000 people to return from camps to safer housing; structural inspection of 450,000 buildings; neighborhood upgrading for over 200,000 people; treatment of over 1,000,000m3 of debris; and rehabilitation of over 90km of roads. • Education and health: provision of over 440,000 tuition waivers and daily meals to 380,000 children (50 percent of whom were girls); provision of maternal and child health services; quality improvements that contributed to increased percentages of prenatal care (25% to 31%) and fully vaccinated children under five (41% to 50%) between 2012 and 2016. • Disaster preparedness: coverage of nearly 80 percent of the population with emergency radio systems and certified level 2 or higher local civil protection committees; support for the new legal framework under the DRM National System; and training of civil protection units in all communes. The lessons were manifold. First, coupled with analytical work, CRW funds have supported evidence-based policymaking that demonstrate to governments the benefits of disaster risk preparedness and management. This was exemplified by Haiti’s orientation of its IDA17 and IDA18 programming to focus strongly on such efforts. Second, investments in public goods—such as infrastructure, health, education and financial inclusion—were crucial for restoring livelihoods. Some forms of public investment, such as roads and neighborhood upgrading, complement direct support to the private sector and served as effective triggers for restarting private investment. Third, sector-level coordination and project-level inclusion mechanisms were critical to stakeholder coordination in the absence of strong government leadership capacity after disasters, as local inputs are needed to contextualize and provide legitimacy for external interventions. Finally, where mechanisms for immediate response do not exist in affected sectors, a single flexible IPF designed for multi-sector rapid response is often an effective tool. - 24 - VI. CRW CONTRIBUTIONS TO THE WBG CRISIS AGENDA 38. The CRW has made valuable contributions to the WBG’s crisis agenda. It pioneered several “firsts”, with its inception marking a paradigm shift in IDA’s crisis response from ad hoc interventions to a more systematic approach, thereby closing a gap in IDA’s financing toolkit. A unique two-step governance arrangement was developed, underpinned by early engagement with the Board. The CRW is also an integral source of financing that allowed IDA to mount crisis responses at scale. From IDA16 onwards, the CRW has allocated US$3.1 billion or close to 30 percent of IDA’s disaster response funding. It is a mainstay of the WBG crisis architecture, with complementarities vis-a-vis other tools such as Cat DDOs, PEF and the GRiF. In addition, CRW- funded activities are an important showcase of the WBG’s value-add in crisis management. 39. IDA Participants and IDA Executive Directors have led efforts in steering the establishment and evolution of the CRW. Participants at the IDA15 Mid-term Review and IDA16 replenishment meetings actively shaped the CRW’s early design . The CRW has also adapted admirably over the years to evolving crisis demands. Notably, following the experience of the 2014 Ebola epidemic, IDA participants agreed at the IDA17 Mid-term Review to expand the CRW’s scope to cover public health emergencies. 40. More broadly, the WBG has undertaken a diverse array of initiatives in crisis management.39 To strengthen risk monitoring and analyses, a rich set of diagnostic tools has been developed and such analyses should inform strategic programming, with the objective of mainstreaming crisis management into the Bank’s country engagement products over time (Annex 4). Similarly, much progress has been made to enhance operational effectiveness in crisis contexts.40 Innovative technology—such as remote sensing and geo-enabled Monitoring and Evaluation—has been harnessed to overcome operational challenges, especially in areas that Bank staff cannot access. These various initiatives are brought under the auspices of the GCRP, which was created to engender a more coherent and strategic approach to crisis management in the WBG. 41. Notwithstanding the WBG’s longstanding contributions to crisis management, a priority ahead is to galvanize greater prevention and preparedness. Such efforts are cost- effective across different types of disasters (Figure 19). Besides actions that directly reduce disaster risk, policies such as social protection, financial inclusion and contingent finance are critical for building resilience and reducing losses to well-being.41 Social safety nets, for instance, help poor people to absorb, cope with and recover from disaster impacts. Stronger, faster and more inclusive reconstruction could generate major benefits totaling US$173 billion annually, or 31 percent of current well-being losses due to natural disasters.42 Within the WBG, a core GCRP objective is to facilitate a “pivot” to prevention and preparedness. The WBG is well-positioned to do this, building on decades of experience in resilience-building in client countries. Multiple initiatives are underway, such as the development of a mechanism to identify and model 39 World Bank, 2018. 40 Examples include the new Procurement Framework and agile processes to endorse or approve policy waivers. 41 Hallegatte et al., 2017. 42 Hallegatte et al., 2018. - 25 - multidimensional risks, and efforts to mainstream crisis management in country engagement products.43 Figure 19. Efforts on Prevention and Preparedness are Cost-effective Median Estimated Benefit-Cost Ratios Source: Wethli, 2014. 42. This “pivot” to prevention and preparedness would take time and the challenges are especially pronounced in IDA countries. Many IDA clients lack the capacity to undertake such endeavors, and the conditions for resilience-building may not be present in conflict-affected areas. Attention to crisis management also fluctuates over time, and the motivation to build back better could wane once the worst effects of a disaster subside. Political opportunity costs may also come into play, as the benefits of resilience-building could materialize years after a government’s tenure has ended. Further, countries face trade-offs in using concessional IDA resources for contingent tools such as Cat DDOs, where funds are not disbursed if there is no crisis. The incentive dilemma is magnified for slower-onset events such as droughts, food insecurity and disease outbreaks. Policymakers may adopt a wait-and-see approach in the earlier stages, given uncertainty as to how such risks would unfold. The impetus to act pre-emptively is further diluted as funds from the international response system are likely to come onstream if the crisis escalates. 43. One policy question for the future is whether and how the CRW could contribute to the shift towards prevention and preparedness. It bears emphasizing that this “pivot” is already being galvanized by an extensive set of initiatives in the WBG.44 Any proposals for the CRW to spur this shift should also be mindful to preserve its core character as a crisis response mechanism, as the CRW is not intended to finance upstream interventions before a crisis occurs. Where the CRW could potentially contribute to resilience-building is twofold. First, the CRW could enhance building back better after a crisis, through strengthening the links to a country’s core IDA programming. Second, it could bolster financial resilience—potentially through pre-arranged financing—to support earlier responses to slower-onset disasters. The next section sets out some preliminary considerations for the discussion and guidance of IDA Participants. 43 World Bank, 2018. 44 World Bank, 2018. The paper sets out the full suite of initiatives in the WBG on crisis risk management. - 26 - VII. POTENTIAL POLICY IMPLICATIONS AHEAD 44. Management’s consultations with internal stakeholders in the course of preparing this review pointed to a number of potential changes to the design and operation of the CRW. This section lays out some of these ideas, which could be examined more closely in the course of the IDA19 replenishment discussions. In particular, Management asks if IDA Participants would support developing a more full-fledged proposal for IDA19 on having the CRW respond at an earlier juncture to food insecurity episodes that are related to CRW-eligible events (paragraph 48). Management also welcomes IDA Participants’ views on the other forward-looking ideas set out below. A. ADVANCING THE PREVENTION AND PREPAREDNESS AGENDA Enhancing Linkages with Core IDA Programming 45. CRW usage could be an opportunity to solidify crisis management as a thematic focus in the core programming of recipient countries. As crisis responses are reactive by nature, core programming remains the central IDA platform for mainstreaming effective crisis management. Going forward, countries that access the CRW could be required to demonstrate a stronger and more systematic focus on prevention and preparedness in subsequent IDA programming, so as to engender longer-term solutions to address their vulnerabilities. The objective is to demonstrate how prevention and preparedness has percolated beyond CRW-funded operations to broader core programming agendas, and report on progress. This is particularly relevant for clients that are structurally crisis-prone, such as those facing recurrent extreme weather events. 46. Where relevant, teams should be encouraged to embed contingent components in project design that could be scaled up quickly to respond to shocks. A well-designed CERC for instance would facilitate contingent components that are ringfenced and time-bound to ensure differentiation from non-crisis activities. This additionality of outcomes is key, as project design for crisis-prone countries should already take into account the clients’ structural vulnerabilities. Outcome additionality could therefore be by way of contingent components for Water Supply, Sanitation and Hygiene (WASH) projects in drought-susceptible areas that allow expanded water services if the drought turns out worse than normal. Social protection programs could also be scaled up during crises to temporarily cover affected households that are otherwise not beneficiaries of the regular programs. The objective is to embed contingency planning that enables a scaled-up response more quickly, including through requesting CRW support if needed. Where applicable, CERCs could be used for such purpose, especially if the country cannot qualify for Cat DDOs or for events where Cat DDO disbursement triggers are not activated. Encouraging Financial Preparedness Through CERCs45 47. The CRW could top up the funds that have been drawn down from existing projects due to CERC activation for crisis response. CERCs are contingent tools that help countries to be better prepared financially for disasters, by reallocating funds from existing projects to disburse 45 The Cat DDOs is another contingent tool. However, no further changes are envisaged to IDA Cat DDOs as they are only recently introduced in IDA18 and already enjoy incentives for usage. See Annex 2. - 27 - quickly for crisis response. Both regular CERCs and IRM-CERCs have been used. As of September 2018, there are 119 active projects with project-specific CERCs and 28 active projects with IRM-CERCs in IDA. Countries and task teams could be reluctant to trigger CERCs as new monies may not be forthcoming to make the original projects whole again. In particular, buy-in from client countries and a strategic approach to identify suitable projects are critical ingredients for CERCs to work well. The proposed CRW top-ups could hence give greater confidence to countries and Bank teams that funding would not be jeopardized, thereby encouraging CERC take- up and usage. Further thought needs to go into potential skewed incentives where countries may use CERCs to withdraw an excessive amount of funds, and measures to avoid such outcomes where necessary. In the event of severe catastrophes, CERC activation could also be supplemented by additional crisis responses that are funded by the CRW in the conventional manner. B. FACILITATING EARLIER RESPONSE TO SLOWER-ONSET CRISES 48. The CRW could allow pre-arranged financing that disburses at an earlier juncture to tackle slow-onset crises, such as food insecurity related to CRW-eligible events. Natural disasters like droughts, or economic shocks such as the 2007–08 food price crisis, could lead to food insecurity. Food insecurity could in turn cause health repercussions like weakened immune systems, and the negative impacts of malnutrition and stunting could afflict children through to adulthood. Unlike sudden-onset crises like earthquakes where the devastation is immediately clear, the impetus to respond to slower-onset events may only come much later. Early response could thus prevent emergent stresses from escalating into full-blown crises. This coheres well with the Famine Action Mechanism (FAM) which tackles food insecurity across prevention, preparedness and response. Work is underway to identify opportunities to build resilience, including possibly through regional approaches. On response, the FAM would link early warnings to early financing and action. The Agriculture Global Practice’s Ag Observatory could potentially provide warnings of food insecurity and price shocks months in advance. The CRW could also contribute by providing funds for early response, which is likely to be cost-effective for IDA compared to dealing with a larger and more expensive crisis down the road. This entails thinking through aspects such as CRW governance, the types of instruments best suited for such interventions, and whether the last resort principle should still apply in the context of slower-onset crises. Preliminary considerations are as follows. • Scope: The pre-arranged CRW support could cover food insecurity related to CRW- eligible events, namely, natural disasters, economic shocks and public health emergencies. As food insecurity itself is not covered by the CRW, the link to a CRW-eligible disaster is necessary to qualify, similar to the rationale in the 2017 regional famine response. It is not proposed to broaden CRW coverage to food insecurity that originates only due to political- or conflict-driven causes, i.e., in the absence of a CRW-eligible event.46 This is because there has been a history of famine being used as an instrument of war.47 If such manipulation leads to a predictable source of CRW funds, it could magnify moral hazard and the risk of CRW funds being misappropriated, or inadvertently cause harm to people. It also raises financial sustainability concerns, as there could be recurrent and substantial 46 To be clear, this does not preclude food insecurity related to CRW-eligible events in fragile and conflict-affected environments, as was already the case during the 2017 regional famine response. 47 De Waal, 1993. - 28 - calls on the CRW, potentially over prolonged periods if the underlying conflict is not resolved. • Stage of Intervention: The CRW could provide early financing for predictions of food insecurity based on modelling of animal and plant health, food prices, weather and water patterns and critical crop seasons, for instance. This could leverage on the real-time forecasting tools being developed by the FAM modelling team and the Agriculture Global Practice’s Ag Observatory, or use project-level triggers, or a mix of these. The principle here is to use objective and verifiable threshold triggers to avoid moral hazard and manipulation. • Activities: Countries could face chronic or seasonal food insecurity especially in arid regions, and activities to tackle such structural challenges should be funded by PBA resources. To ensure outcome additionality and avoid substituting PBA, the CRW should provide funds for worse-than-expected food insecurity episodes. Operations could embed CRW-funded contingent components that are activated to enable a scaled-up response in the event of heightened food insecurity, with the regular non-crisis components being financed by other sources. Box 4 presents an example of a potential modus operandi. Going forward, a few pilot projects could be identified to develop and test such solutions. • Size of CRW Allocations: More thought also needs to go into appropriately sizing these interventions. An “early action” tranche could be created within the existing CRW envelope, but managed flexibly such that it could also be used for non-food security related crises if needed, e.g., if CRW funds outside of the tranche have been exhausted. C. REVISITING CRW ACCESS 49. To incentivize countries to focus on crisis prevention and preparedness, one option is to make CRW access conditional on prior resilience-related programming. This would however impose conditionalities that would exclude the most fragile IDA clients as they likely lack the capacity to undertake such endeavors. While less stringent criteria could apply to countries with weaker capacity, designing appropriate conditionalities remains difficult. Moreover, there are fundamental policy issues with using CRW access as a “stick” to motivate change. Withholding CRW support may inadvertently cause countries to regress or delay their recovery, and IDA could face reputational risk if a severely-affected client is denied CRW financing. Such proposals are also unlikely to be effective without concomitant reforms to the broader international response system. 50. Other ideas—such as introducing parametric triggers48 and restricting repeated access—were also contemplated, but neither seem consistent with the mandate and operational effectiveness of the CRW. While parametric triggers increase the transparency of when CRW funds would be deployed, there are several drawbacks. Given the wide range of crises covered by the CRW, having multiple triggers could introduce considerable operational complexity. Situations involving “compound risks” further complicate how the triggers could 48 Parametric triggers are used in catastrophe insurance, where a party purchases a pre-defined amount of coverage which will pay out a claim based on pre-determined terms, e.g., when certain levels of earthquake intensity, hurricane wind speed or rainfall accumulation are breached. - 29 - work in such instances. Moreover, parametric triggers may conflict with the CRW’s public policy character. The current CRW framework49 uses multifaceted eligibility criteria such as the impact of the disaster especially on poor and vulnerable communities, domestic response capacity and cost of recovery. This approach is more germane given IDA’s poverty reduction mandate. Notably, the considerations are different in paragraph 48, where parametric triggers only apply to a small subset of disasters and the intent is to mobilize funds earlier to prevent deterioration. In that context, parametric triggers are deemed appropriate as thresholds to activate mitigating measures since a full assessment of the crisis impact—as is typically conducted for CRW requests—would not be possible at that stage. As for repeated access to the CRW, 37 percent of CRW recipients are recurrent beneficiaries, of which all are IDA-only clients. This reflects the fact that many IDA countries face recurring exposure to crises, and some risks such as seasonal climatic events and certain health hazards50 are recurrent. Rather than tighten CRW access, a more appropriate approach is to use CRW responses as opportunities to solidify resilience-building in IDA programming, as discussed in paragraph 45. 51. Management proposes to explore providing access to the CRW in IDA19 for recent IDA graduates and small states under certain limited conditions. Please see the IDA18 MTR paper on “Transitioning out of IDA Financing: A Review of Graduation Policy and Transition Process” for details. D. OTHER ISSUES 52. Twelve percent or US$350 million of the CRW’s US$3 billion envelope in IDA18 have been allocated so far, and the present pace of usage suggests that US$1 billion to US$1.5 billion could be left untapped. To maximize the use of resources in a given IDA cycle, this potential surplus could be redeployed to other uses. This is discussed in the IDA18 MTR Chapeau paper.51 Such an approach still leaves the CRW with firepower of US$1.15–1.65 billion to respond to crises over the remainder of IDA18. As a rough guide, the top five largest CRW allocations amount to US$368 million per crisis on average; in other words, the remaining CRW size of US$1.15–1.65 billion would be sufficient to fund 3.1 to 4.5 times such responses. Should the CRW turn out to be insufficient over IDA18, for instance in the event of an unexpected and very large crisis52, IDA can use the flexibility of its new hybrid leverage model to increase IDA18 commitments beyond the agreed US$75 billion, provided that adjustment could be made in future replenishments while maintaining IDA’s overall financial sustainability and adequate capital. 53. The Cat DDO is a recent addition to IDA’s crisis management toolkit and no changes to IDA’s Cat DDO framework are currently proposed. During the IDA18 replenishment discussions, staff had committed to review: (i) if IDA would require additional resources in the event that Cat DDO drawdown rates turn out higher-than expected; and (ii) whether the IDA Regional Program could be utilized for multi-country Cat DDOs. Annex 2 reviews IDA’s 49 World Bank, 2017c. 50 For instance, the Democratic Republic of Congo faced its 10th Ebola outbreak in August 2018. 51 See the IDA18 MTR chapeau paper on, “IDA18 Mid-Term Review Implementation and Results Progress Report. Towards 2030: Investing in Growth, Resilience and Opportunity. Delivering on innovation and transformation, and managing IDA resources for greatest impact”. 52 Such events cannot be ruled out. See paragraphs 4 and 5. - 30 - experience with Cat DDOs thus far and proposes that no further changes be made to the framework at this juncture. As the Cat DDO is a feature pertaining to Development Policy Operations (DPOs), the topic of multi-country Cat DDOs would also require first exploring whether to open up the IDA Regional Program to multi-country DPOs. This is discussed in the IDA18 MTR paper on “Review of the IDA Regional Program”. 54. The following issues are for IDA Participants’ discussion and guidance: • Do IDA Participants agree with the assessment of the CRW’s performance against its key policy objectives as set out in Section V? • Do IDA Participants see scope for the CRW to be used more strategically for galvanizing prevention and preparedness, while preserving its core function as a vehicle for crisis response? What are IDA Participants’ views on the preliminary ideas in sections VII.A and VII.B? Should these be further explored as part of the IDA19 replenishment discussions? • In particular, do IDA Participants support developing a more full-fledged proposal for IDA19 on having the CRW respond at an earlier juncture to food insecurity episodes that are related to CRW-eligible events? (paragraph 48) • Do IDA Participants agree to explore providing access to the CRW in IDA19 for recent IDA graduates and small states under certain limited conditions? (paragraph 51) Please see the IDA18 MTR paper on “Transitioning out of IDA Financing: A Review of Graduation Policy and Transition Process” for details. • Do IDA Participants agree that US$1 billion to US$1.5 billion of CRW resources could be redeployed for other purposes? (paragraph 52) Please also see the IDA18 MTR Chapeau paper for details. • Do IDA Participants agree to maintain the current IDA Cat DDO framework, and that there is no need at present to develop a framework to use the IDA Regional Program resources for multi-country Cat DDOs? (paragraph 53) Please also see Annex 2 of this paper and the IDA18 MTR paper on “Review of the IDA Regional Program” for details. - 31 - Annex 1: List of CRW Allocations from IDA16 to End-September 2018 IDA Calendar Fiscal CRW Allocation Type of Disaster Region Country Project Name(s)** Cycle Year* Year (US$ Million) Relaunching Agriculture: Strengthening Agriculture Public Services II Project Disaster Risk Management and Reconstruction Project Haiti Education for All Project - Phase II AF for Infrastructure and Institutions Emergency Recovery Project Rebuilding Energy Infrastructure and Access Project IDA16 2011 2012 Earthquake LCR Haiti 508 Business Development and Investment Project Improving Maternal & Child Health through Integrated Social Services Project Economic Reconstruction and Growth Development Policy Financing Cultural Heritage and T ourism Sector HT Center & Artibonite Regional Development Urban CDD Project AF Horn of Africa 250 Health Sector Support Project - Additional Financing AFR Kenya 87 Additional Financing for Water and Sanitation Services Improvement Project AFR Ethiopia 70 Additional Financing Productive Safety Net APL III Project IDA16 2011 2012 Drought AFR Regional 80 Horn of Africa Emergency Health and Nutrition Project AF for Rural Community Development and Water Mobilization Project MNA Djibouti 13 Social Safety Net Project Second AF for Power Access and Diversification Project National Water Resources Development Flood Response Project IDA16 2013 2013 Flood AFR Mozambique 71 Roads and Bridges Management Maintenance Project Phase II Development Policy Financing IDA16 2013 2013 Cyclone EAP Samoa 20 Enhanced Road Access Project Agriculture and Fisheries Cyclone Response Project IDA16 2014 2014 Cyclone EAP T onga 12 T onga Cyclone Reconstruction IDA16 2014 2014 Flood LCR St. Lucia 17 Disaster Vulnerability Reduction Project IDA16 2014 2014 Flood LCR St. Vincent 19 Regional Disaster Vulnerability Reduction Project - Additional Financing IDA16 2014 2014 Flood ECA Bosnia 100 BiH Floods Emergency Recovery Project Solomon Islands Rural Development Program II IDA17 2014 2015 Flood EAP Solomon Islands 10 Economic Reform and Recovery DPO AF(CRW) for SB-Rapid Employment Project West Africa 420 AFR Guinea 107 Emergency Macroeconomic and Fiscal Support Grant AFR Liberia 177 Poverty Reduction Support Credit II IDA17 2014 2015 Health (Ebola) Emergency Economic and Fiscal Support Operation (EEFSO) AFR Sierra Leone 136 Ebola Emergency Response Project Ebola Emergency Response Project - Additional Financing IDA17 2015 2015 Flood AFR Malawi 80 Malawi Floods Emergency Recovery IDA17 2015 2015 Cyclone EAP Vanuatu 50 Infrastructure Reconstruction and Improvement Project IDA17 2015 2015 Cyclone EAP T uvalu 3 Supplemental Financing for Second DPO Financial sector stability DPC2 IDA17 2015 2015 Earthquake SAR Nepal 300 Earthquake Housing Reconstruction Project NP: Road Sector Development Project IDA17 2016 2016 Flood EAP Myanmar 100 Flood and Landslide Emergency Recovery Project IDA17 2016 2016 Drought AFR Ethiopia 100 Additional Financing PSNP IV Drought 90 AFR Lesotho 20 Crisis response AF to SA project AFR Madagascar 20 Social Safety Net Drought Response IDA17 2016 2017 Drought Second Additional Financing for MASAF IV AFR Malawi 20 Malawi Drought Recovery and Resilience Emergency Resilient Recovery Project AF AFR Mozambique 30 MZ-Social Safety Net project Africa 150 AFR Chad 80 Development Policy Operations (DPO) West Africa Agric Prod Progrm (WAAPP-1C) AFR Guinea 30 IDA17 2016 2017 Economic Shock GN: Economic Governance T echnical Assist LR-PRSC III (FY16) AFR Liberia 20 Second AF LR Road Asset Management AFR Sierra Leone 20 Productivity and T ransparency Support Cr HT Disaster Risk Mngmt & Reconstruction Agriculture and Rural Development IDA17 2017 2017 Hurricane LCR Haiti 100 HT Sustainable Rural WSS Project HT Improving Maternal and Child Health Africa & Yemen 360 AFR Ethiopia 100 ET Productive Safety Nets Project 4 Drought / Famine AFR Kenya 35 AF for the NSNP IDA17 2017 2017 Response AFR Somalia 50 Emergency Drought Response and Recovery Project AFR South Sudan 50 Emergency Food and Nutrition MNA Yemen 125 Yemen Emergency Crisis Response Project Emergency Agricultural Livelihoods and Climate Resilience Project IDA18 2017 2018 Hurricane LCR Dominica 50 Housing Recovery Project IDA18 2017 2018 Economic Shock EAP Mongolia 80 Economic Management Support Operation First DPF IDA18 2017 2018 Health (Cholera) MNA Yemen 200 Yemen Emergency Health and Nutrition Project Second Additional Financing Supplemental Financing to the Second Inclusive Growth DPO IDA18 2018 2018 Cyclone EAP T onga 20 AF to T onga: Pacific Resilience Project under the Pacific Resilience Program * T he calendar and fiscal years here are based on the dates of the respective CRW allocations. ** T hese reflect commitments to date which can differ from the allocation envelope. - 32 - Annex 2: IDA’s Experience with Cat DDOs 1. The Catastrophe Deferred Drawdown Option (Cat DDO) enables countries to secure pre-arranged financing that can be disbursed quickly to address natural catastrophes, including health-related events. Funds are channeled as budget support after the drawdown trigger—typically the member country’s declaration of a state of emergency—is met. The Cat DDO has been available to IBRD countries and was extended to IDA clients in IDA18. To date, there have been 16 Cat DDOs approved totaling US$3.2 billion: 15 in IBRD countries and the first to an IDA client (Kenya). The IDA pipeline for FY19 currently includes six Cat DDO operations. 2. The IDA18 policy package includes innovations to Cat DDOs. As countries may be reluctant to use concessional IDA funds for contingent purposes, several enhancements were made. Notably, if a country funds its Cat DDO using concessional core IDA allocations, only 50 percent of the Cat DDO amount would come from its country envelope, with the balance matched by IDA overall resources. Upon cancellation or expiry of the Cat DDO, clients could also recommit the concessional core portion of undisbursed balances for other purposes, so that they do not lose these resources if no eligible crises transpire. 3. The Cat DDO and CRW contribute to IDA’s crisis management toolkit in complementary ways. First, the Cat DDO provides fast-disbursing funds in the immediate aftermath of a crisis, while other support such as CRW resources are being mobilized. Second, the Cat DDO strengthens the linkages between the ex-ante and ex-post phases of crisis management. To qualify, countries need to have: (i) an adequate macroeconomic policy framework; and (ii) preparation for, or existence, of a satisfactory DRM program. The latter condition means that the Cat DDO is a particularly valuable platform for the Bank to galvanize policy dialogue with countries to fortify preparedness and prevention. This would both over time reduce the impact of disasters and lessen the need for large amounts of CRW funding after a catastrophe, and help ensure that CRW monies—if mobilized—would be used more effectively. Third, as countries’ needs could be amplified in the event of significant crises, the Cat DDO and CRW could serve as additive sources of financing. 4. The first IDA Cat DDO was approved for Kenya in June 2018. This was a US$200 million operation with a focus on addressing climate-related vulnerabilities. 84 percent of Kenya’s land is arid or semi-arid, and the livelihoods of farmers and pastoralists are also susceptible to increased variability of rainfall patterns. Kenya’s Cat DDO therefore embedded Prior Actions such as approval of an integrated National DRM Policy; operationalizing the Water Act which supports water resource management and climate change adaptation; and approval of the National Disaster Risk Financing Strategy that defines priorities to enable effective and timely responses to crises. 5. The build-up of IDA’s Cat DDO portfolio would take time, given that this is the first year of implementation. The IBRD experience has shown that the Cat DDO provides Bank teams with a valuable platform to promote dialogue with governments on disaster risk management. This effort takes time. IDA clients are at varying stages of readiness to take up Cat DDOs, including amid challenging macroeconomic conditions which have delayed some Cat DDO operations. 6. No changes to the IDA Cat DDO are currently proposed. During the IDA18 replenishment discussions, staff committed to review: (i) if IDA would require additional resources - 33 - in the event that Cat DDO drawdown rates turn out higher-than expected, as Cat DDOs are not fully funded from IDA country envelopes; and (ii) whether the IDA Regional Program could be utilized for multi-country Cat DDOs. On (i), there is no Cat DDO drawdown to date. On (ii), there has not been pent-up demand for IDA Cat DDOs and it would be premature to design a new framework for Cat DDOs to tap the Regional Program financing. Moreover, as the Cat DDO is a feature pertaining to DPOs, this would require first exploring whether to open up the IDA Regional Program to multi-country DPOs—which is discussed in the IDA18 MTR paper on “Review of the IDA Regional Program”. The topic of regional Cat DDOs could be revisited in future if needed. - 34 - Annex 3: Pilot CRW in IDA15 1. The establishment of the Pilot CRW followed calls from the Development Committee as well as from G20 leaders for IDA to support low income countries affected by the global financial crisis.53 The proposal was discussed at the IDA15 MTR in November 2009 and approved by IDA’s Executive Directors in December 2009, and its implementation began in mid-January 2010.54 The Pilot CRW was set up with an allocation of US$1.55 billion (SDR1.05 billion), of which 77 percent was redeployed from internal IDA resources set aside for arrears clearance and re-engagement, with the remainder from IDA investment returns and an additional contribution from the UK. Allocation Methodology 2. Pilot CRW resources were allocated through a two-stage process, with 85 percent envisaged to be distributed in stage 1 and the rest in stage 2. Stage 1 allocations were based on broad macroeconomic indicators, namely, GDP impact and GNI per capita. Stage 2 allocations were made to countries with: (a) highest crisis impact based on more detailed fiscal, social, poverty and employment impact assessments; (b) largest pre-existing needs taking into account additional poverty- and fragility-related factors; (c) resource needs and availability; and (d) capacity to effectively use Pilot CRW resources to protect core spending. The eventual stage 1 and stage 2 ratio shifted from 85-15 to 80-20 due to adjustments in the East Asia and Pacific (EAP) region55. Fifty-six countries were eligible for Pilot CRW resources 56, of which 46 utilized the funds. The allocation methodology was broadly effective in rationing scarce Pilot CRW resources, with 84 percent of the resources going to countries with high crisis impact and/or high needs (Figure A3.1). Commitments 3. The Pilot CRW met expectations in terms of providing rapid and additional support to IDA countries. US$1.4 billion of Pilot CRW resources were committed through 75 operations to 46 countries that were fully or partially funded by the Pilot CRW. Of the 75 operations, 48 were new operations with the remaining 27 provided as additional financing. The majority of commitments were in Sub- Saharan Africa (Figure A3.2). A range of instruments was used. About half (35) of the projects were Specific Investment Lending (SIL) operations amounting to SDR480 million. There were 21 DPOs totaling SDR187 million, and 14 Emergency Recovery Lending (ERL) operations amounting to SDR188million. DPOs were typically used to protect core spending in countries facing falling fiscal revenues and widening deficits, while ERLs and additional financing of existing projects facilitated faster response. 53 World Bank, 2009a. 54 World Bank, 2009b and Implementation Arrangements for the IDA Crisis Response Window, internal note to staff, January 2010. 55 Given the small size of individual Pilot CRW allocations and the need for economies of scale, EAP distributed the combined stages 1 and 2 allocations to the three worst-hit countries, namely, Lao PDR, Mongolia and Samoa. 56 Twenty-three countries were excluded, which comprised: (i) 15 blend countries; (ii) four IDA-only countries (Angola, Chad, Republic of Congo and Nigeria) that were significant net oil exporters which, while having been impacted by the crisis, had also accumulated significant windfall gains due to the high oil prices preceding the crisis; and (iii) four countries with arrears to IDA and/or IBRD (Myanmar, Somalia, Sudan and Zimbabwe). - 35 - Figure A3.1. Resource Distribution Based Figure A3.2. Regional Distribution on Crisis Impact and Country Needs (FY10 and FY11) Source: World Bank, 2012. Source: World Bank, 2012. 4. In line with the aim of protecting core spending, some three quarters of Pilot CRW resources supported investments in agriculture and rural development, health, education, social protection and infrastructure. While Pilot CRW resources did not have sectoral or thematic earmarking, country teams were encouraged to focus on interventions that: (a) deliver rapid impact in protecting core spending; (b) address the poverty, social and economic repercussions of the crisis; and (c) build resilience to future shocks. The diversity of instruments available and the absence of thematic earmarking were important factors that enabled country teams to tailor their interventions to each country’s crisis-specific needs and circumstances. Timeliness and Disbursement 5. Pilot CRW operations were on average prepared faster than regular IDA operations. The average times from project concept to approval, and from approval to effectiveness, were 6.8 months and four months respectively. Flexibility in the choice of instruments led to frequent use of additional financing, DPLs and ERLs which contributed to faster processing times. Responses were also quicker where interventions could rely on existing systems for implementation. Almost all operations under the Pilot CRW have fully disbursed. Performance 6. Along with alleviating the crisis impact, a large share of Pilot CRW projects also helped to build resilience to future shocks and protect investments envisaged in country strategies that could be jeopardized due to the crisis. One lesson was that the large swathe of eligible countries resulted in a sizable number of smaller interventions, which added to implementation and supervision challenges. Of the 64 operations that have closed, 53 were rated by the IEG as being in the Satisfactory range (i.e., Highly Satisfactory, Satisfactory or Moderately Satisfactory). These operations account for 86 percent of Pilot CRW commitments. - 36 - Annex 4: Crisis Management Themes in the Bank’s Country Engagement Products 1. This Annex sets out preliminary findings on the extent to which the Bank’s key country engagement products have focused on crisis management in client countries. A sample of IDA countries susceptible to natural disasters, public health emergencies and economic crises were covered.57 All 13 Systemic Country Diagnostics (SCDs) reviewed had discussed the importance of crisis management in some fashion, either by identifying resilience-building as a priority or identifying the crisis risks as key binding constraints. Of the 10 country programming documents reviewed, all but two had a focus on strengthening resilience to disaster risk in various forms. 2. Examples include the mainstreaming of disaster risk management into policy-making in the Pacific Island States. Resilience-related programming also featured prominently in cases such as enhancing the management of climate risks in Mozambique, increasing smallholder farmers’ resilience to weather shocks in Madagascar, and improving safety nets in Ethiopia. In Myanmar, the WBG’s strategy is described as incorporating “emergency response contingencies” in IDA operations to enable rapid reprogramming of funds for crisis response. The IEG also commended the systemic integration of Risk and Resilience Assessments (RRAs) in the SCDs and CPFs of FCV countries.58 3. The following lessons were drawn from IEG evaluations59 and four PLRs: • Strategic programming. Countries underscored the importance of flexibility in adapting to changing country circumstances. The Nepal PLR60 noted that this flexibility was built into its Country Partnership Strategy (CPS) by focusing on the long-term foundations of inclusive and sustainable growth, which enabled the Bank to respond quickly to changes in country context due to exogenous shocks. This flexibility was further embedded in the Nepal CPF for FY19–23. Clients also emphasized the importance of tailoring the WBG interventions to country capacity. • Instruments. DPOS were an important part of the WBG’s crisis responses. These however require follow-up for effective long-term resilience-building.61 DDOs seem the most effective in strengthening resilience when drawdown is genuinely deferred. In addition, some countries valued the variety of instruments available to address disasters.62 • Partnerships. The IEG and Bank PLRs63 noted that a holistic approach to strengthening resilience requires strong partnership between the WBG, its development partners, regional bodies and client countries. The IEG posited that coordination is critical during crises, as the aftermath of disasters is often a chaotic scene where the delineation of responsibilities is not always clear. There is scope for international efforts in determining 57 The documents reviewed spanned 13 SCDs; 10 CPFs, CPSs and CENs; and four PLRs. 58 Please see the IDA18 MTR paper Fragility, Conflict and Violence (FCV). 59 Independent Evaluation Group, 2017. 60 Performance and Learning Review of the Country Partnership Strategy for Nepal for FY14 –FY18. 61 Independent Evaluation Group, 2017. 62 Performance and Learning review of the Country Partnership Strategy for Kenya for the Period FY14-FY18. 63 Performance and Learning Review of the Regional Partnership Strategy for OECS for FY15 –FY19. - 37 - how to better handle and address challenges following a crisis, particularly for natural disasters. • Global programs. Relatedly, the IEG noted that global programs could have an important role in enhancing the effectiveness of crisis response and strengthening resilience to future shocks. For instance, the Global Food Crisis Response Program (GFRP) and the Global Program for Avian Influenza made valuable contributions, as did the Global Facility for Disaster Reduction and Recovery (GFDRR) in disaster reduction and climate change adaptation. - 38 - REFERENCES AND BIBLIOGRAPHY Arbache, J., D. Go, & V. Korman. 2010. Does Growth Volatility Matter for Development Outcomes? An Empirical Question Using Global Data, and World Bank staff calculations. De Waal, War and Famine in Africa. 1993. Article first published in 1993, and republished in the Institute of Development Studies Bulletin, Archive Collection, Volume 49 (April 2018). Hallegatte, S., Rentschler, J., & Walsh, B. 2018. Building Back Better: Achieving Resilience through Stronger, Faster, and More Inclusive Post-Disaster Reconstruction. World Bank, Washington, DC. © World Bank. Hallegatte, S., Vogt-Schilb, A., Bangalore, M. & Rozenberg, J. 2017. Unbreakable: Building the Resilience of the Poor in the Face of Natural Disasters. Climate Change and Development. Washington, DC: World Bank. © World Bank. Independent Evaluation Group. 2017. Crisis Response and Resilience to Systemic Shocks: Lessons from IEG Evaluations. Washington, DC: World Bank. Laeven, L. & Valencia, F. 2012. Systemic Banking Crises Database: An Update. IMF Working Paper No. 12/163. Madhav N., Oppenheim B., Gallivan M., et al. 2017. Pandemics: Risk, Impacts, and Mitigation. In: Jamison DT, Nugent R, Gelband H, et al, eds. Disease Control Priorities, 3rd Edition, Volume 9. Washington, DC: World Bank. Nolan, C. et al. 2018. Past and Future Global Transformation of Terrestrial Ecosystems Under Climate Change. Science 31 August 2018, Vol. 361, Issue 6405, pp. 920-923. Smith, K. F., Goldberg, M., Rosenthal, S., Carlson, L., Chen, J., Chen, C., & Ramachandran, S. 2014. Global rise in human infectious disease outbreaks. Journal of the Royal Society Interface, 11(101), 20140950. Steffen, W. et al. 2018. Trajectories of the Earth System in the Anthropocene. Proceedings of the National Academy of Sciences. Thomas, V. & López, R. 2015. Global Increase in Climate-Related Disasters (November 2015). Asian Development Bank Economics Working Paper Series No. 466. United Nations Office for Disaster Risk Reduction. 2015. Global Assessment Report on Disaster Risk Reduction 2015. New York: UNISDR. United Nations Office for Disaster Risk Reduction and Center for Research on the Epidemiology of Disasters. 2016. Poverty and Death: Disaster Mortality 1996–2015. Wethli, K. 2014. Benefit-cost analysis for Risk Management: Summary of Selected Examples. Background paper for the World Development Report 2014. - 39 - World Bank. 2009a. World Bank-IMF Annual Meetings 2009 Development Committee Communiqué. October 5, 2009. __________. 2009b. Proposal for a Pilot IDA Crisis Response Window. November 23, 2009. __________. 2010. Technical Note on the Establishment of a Crisis Response Window in IDA16. __________. 2012. IDA16 Mid-Term Review: Enhancing IDA’s Capacity to Respond to Crisis. __________. 2015. IDA17 Mid-Term Review: Update on IDA’s Crisis Response Window. __________. 2017a. World Development Report 2017: Governance and the Law. Washington, DC: World Bank Group. __________. 2017b. World Bank Group Global Crisis Response Platform. __________. 2017c. Report from the Executive Directors of the International Development Association to the Board of Governors: Additions to IDA Resources – Eighteenth Replenishment (English). Washington, D.C.: World Bank Group. __________. 2018. Global Crisis Risk Management Platform. World Food Programme. 2018. Global Report on Food Crises 2018.