MAY 2016 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 1 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT A Summary of Findings from the Disaster Risk Finance Impact Analytics Project 2 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT ACKNOWLEDGMENTS This note was authored by Daniel Clarke of the Disaster Risk Financing and Insurance Program at The World Bank Group and Ruth Hill of the Poverty and Equity Global Practice at The World Bank Group, with the support of Johanna Avato, Darcy Gallucio, and Susann Tischendorf (consultants, The World Bank Group) and with contributions from Mehari Hiluf Abay, Bianca Adam, Christopher Adam, Jesse Anttila-Hughes, Guush Berhane, David Bevan, Laura Boudreau, Sarah Coll-Black, Samantha Cook, Naomi Cooney, Alain de Janvry, Alejandro del Valle, Mareile Drechsler, Anna Edwards, Benjamin D. Fox, Stéphane Hallegatte, Roman Hohl, Munenobu Ikegami, Nathaniel Jensen, Andrew Jinks, Roberto Ley-Borrás, Olivier Mahul, Shadreck Mapfumo, Karlijn Morsink, Andrew Mude, Catherine Porter, Richard Poulter, Elizabeth Ramirez Ritchie, Elisabeth Sadoulet, Mohan Sharma, Wolter Soer, Tse-Ling Teh, Emily White, and Liam Wren-Lewis. The note is an output of the Disaster Risk Finance Impact Analytics Project, and it greatly benefited from the technical expertise of the Project’s steering committee, comprising Charlotte Benson (Asian Development Bank), Richard Choularton (World Food Programme), Stefan Dercon (Centre for the Study of African Economies, University of Oxford, and U.K. Department for International Development), Francis Ghesquière (Global Facility for Disaster Reduction and Recovery), Robert Muir-Wood (Risk Management Solutions), and Reto Schnarwiler (Swiss Re). We are grateful for the inputs and reviews from all those who acted as peer reviewers for the underlying research that this note attempts to distill: Diego Arias Caraballo, David Bevan, Richard Choularton, Sarah Coll-Black, Samantha Cook, Tito Cordella, Julie Dana, Anna Edwards, Ghada Elabed, Alan Fuchs Tarlovsky, Xavier Gine, Alejandro Guerson, Stéphane Hallegatte, Niels Holm- Nielsen, Oscar Anil Ishizawa Escudero, Barry Maher, Olivier Mahul, Juan Jose Miranda Montero, Robert Muir-Wood, Richard Poulter, Ian Rogers, Elisabeth Sadoulet, Wendell Samuel, Emmanuel Skoufias, Charles Stutley, Panos Varangis, and Liam Wren-Lewis. We also appreciate the overall guidance of Alfonso Garcia Mora and Samuel Maimbo of the Finance and Markets Global Practice and Francis Ghesquière of the Disaster Risk Management Practice Group, with the support of Olivier Mahul of the Disaster Risk Financing and Insurance Program and Emmanuel Skoufias of the Poverty and Equity Global Practice. Finally, we thank the team at Studiografik for the design and layout of the note and Sabra Ledent for proofreading and copyediting services. The Disaster Risk Financing and Insurance Program—a joint initiative of The World Bank Group’s Finance and Markets Global Practice and the Global Facility for Disaster Reduction and Recovery—and The World Bank Group’s Poverty and Equity Global Practice are grateful for the financial support received from the Global Facility and the U.K. Department for International Development’s Humanitarian Innovation and Evidence Programme. DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 3 TABLE OF CONTENTS Acknowledgments 2 Overview 5 1 Increasing Commitment Through Disaster Risk Finance 9 Dull Disasters? How Planning Ahead Will Make a Difference 10 Solving Commitment Problems in Disaster Risk Finance 12 Disasters and Discipline: The Political Economy of Natural Disasters and of Sovereign Disaster Risk Finance and Insurance in Mexico 13 2 The Benefits of Timely Reconstruction 15 Indirect Cost of Natural Disasters and an Economic Definition of Macroeconomic Resilience 16 Insuring Growth: The Impact of Disaster Funds on Economic Reconstruction in Mexico 17 3 The Benefits of Timely Support to Livelihoods 19 The Application of a Probabilistic Catastrophe Risk Modelling Framework to Poverty Outcomes 20 Effects of Timing of Public Work Payments on Welfare: The Case of Ethiopia’s Productive Safety Net Programme 21 Early Warning, Early Action: The Use of Predictive Tools in Drought Response Through Ethiopia’s Productive Safety Net Programme 22 How to Measure Whether Index Insurance Provides Reliable Protection 23 Integrating Social Protection Strategies for Improved Impact: A Comparative Evaluation of Cash Transfers and Index Insurance in Kenya 24 Weather Index Insurance and Shock Coping: Evidence from Mexico’s CADENA Program 25 4 Saving Money Through Disaster Risk Finance 27 Evaluating Sovereign Disaster Risk Finance Strategies: A Framework 28 Evaluating Sovereign Disaster Risk Finance Strategies: Guidance and Case Studies 29 A Methodology to Assess Indicative Costs of Risk Financing Strategies for Scaling Up Ethiopia’s Productive Safety Net Programme 30 Financing the Reconstruction of Public Capital after a Natural Disaster 31 Using Probabilistic Models to Appraise and Decide on Sovereign Disaster Risk Financing and Insurance 32 References 33 4 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 5 OVERVIEW Extreme natural events such as droughts, floods, earthquakes, tropical cyclones, and pandemics can threaten lives, livelihoods, and even entire economies. Disaster risk finance aims to increase the resilience of vulnerable countries to the financial impact of disasters as part of a comprehensive approach to disaster risk management. By increasing resilience, disaster risk finance offers the promise of protecting and promoting development. But does it actually work in practice? Critics of disaster risk finance often argue that investing to avoid or reduce risk is more cost-effective than investing in post-disaster expenditures. They also argue that insurance and other risk transfer instruments can be opaque and expensive, providing poor value to governments. Generating the evidence to better-guide investments in sovereign disaster risk finance programs, to maximize their expected humanitarian and development impacts, and to ensure that public investments deliver value for money requires robust methodologies—ones that rigorously monitor and evaluate existing schemes and new products. Since 2013 The World Bank Group has partnered with the Global Facility for Disaster Reduction and Recovery and the U.K. Department for International Development to address some of these gaps in evidence and methodologies. The Disaster Risk Finance Impact Analytics Project has made significant contributions to the understanding of how to monitor and evaluate existing or potential investments in disaster risk finance from a development perspective, and to the evidence base for where such investments have development impact. This note summarizes the findings of this project, presenting the key messages of a book, a technical report, and 14 research papers, categorized into four themes. Increasing Commitment Through Disaster Risk Finance. Independent central banks can, it is widely understood, resolve a commitment problem in macroeconomic policy—if governments allow themselves full discretion to set interest rates, short-term political incentives will tend to lead to interest rates that are too high, at a huge cost to the economy. By establishing an independent institution with the power to set 6 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT interest rates, but subject to pre-agreed rules flood—and supporting the individuals who or principles, governments around the world have lost their livelihoods as a result of a resolve this commitment problem. Financing disaster—for example, financial assistance to post-disaster needs faces a similar commitment households that lost their harvest as a result problem—if governments, firms, people, and of drought. Two elements of response, speed development partners allow themselves full and reliability, can bring significant benefits to discretion to decide who will pay for what relief individuals and economies. and reconstruction after a potential disaster, short-term political incentives will tend to Timely Reconstruction. Part 2 explores the lead to slow, fragmented, unreliable response, economic gains from speed and reliability as well as to underinvestment in adaption and through two papers. Hallegatte’s elegant rule risk reduction. This happens again and again of thumb estimates the total economic cost in rich and poor countries alike, and leads to of a disaster, beyond the direct loss of assets. unnecessarily high human and economic costs De Janvry, del Valle, and Sadoulet report on an from natural hazards. impact evaluation of Mexico’s fund for natural disasters, FONDEN: the faster reconstruction Part 1 presents insurance and insurance- of infrastructure assets made possible by like institutions—both public and private— FONDEN’s disaster risk finance strategy as potential solutions to this commitment contributes, on average, to an increase in post- problem. Clarke and Dercon present the disaster local economic activity of 2–4 percent. overarching argument in their book Dull Disasters; Clarke and Wren-Lewis provide Timely Support to Livelihoods. Part 3 investigates a more theoretical economic analysis of the gains to household welfare from speed the commitment problem and the range and reliability and partitions this large body of potential insurance-like solutions; and of research into empirical evidence and Boudreau presents evidence from Mexico methodologies. Porter and White show that a suggesting that disaster risk finance programs rural safety net program in Ethiopia lessens a can indeed work as a commitment device for drought’s impact by 25 percent. The speed with governments. which this safety net makes transfers, Berhane, Abey, and Hoddinott demonstrate, affects the But building and using public or private benefits realized by individuals. For example, institutions that provide post-disaster when households receive financial support financing according to pre-agreed rules has a by the beginning of the lean season, they are cost, and are the benefits really worth it? better able to increase consumption, to reduce The end use of the financing is crucial to answer malnutrition, and to keep children in school. this question. After a disaster, governments In a very different setting, de Janvry, Ramirez and partners may respond in a myriad of Ritchie, and Sadoulet find that when drought ways but most funding channels to two strikes, insurance payouts from the disaster broad categories of response: reconstructing risk finance program CADENA increase buildings and other damaged or destroyed Mexican farmers’ income by 38 percent and physical capital—for example, a bridge lost their consumption by 27 percent. in an earthquake or a road washed away in a DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 7 But a financial instruments’ quality also of risk finance strategies and to help decision determines the degree to which livelihoods makers choose the least-cost approach. and welfare improve. Taking agricultural index insurance as an example, Morsink, Clarke, But can choosing the right combination of and Mapfumo provide quantitative measures financial instruments markedly impact the to answer two key questions: How well does cost of a more timely response? the insurance insure what it set out to insure? And does the insurance really help reduce the In a worked example of scalable social income risk that poor households face? Jensen, protection in Ethiopia, the authors find that Ikegami, and Mude examine two instruments an alternative disaster risk finance strategy well-suited to protect the livelihoods of could reduce the average cost of financing pastoralists in northern Kenya—index scalability expenditures by 25 percent. insurance and scalable social protection—and Applying a macroeconomic model to Jamaica, discover both impact welfare positively. Bevan and Adam show that reallocating budget expenditure on operations and maintenance Saving Money Through Disaster Risk Finance. to finance more timely reconstruction was Part 4 considers the cost of providing timely three times more expensive than insurance, financing for ex ante response and lays out which was in turn slightly more expensive how well-structured risk finance strategies than raising taxes. Finally, Ley-Borrás and can reduce the economic costs of managing Fox explain catastrophe risk models and fiscal volatility. Three papers—Clarke, Mahul, apply them to designing, implementing, and Poulter, and Teh; Clarke, Cooney, Edwards, monitoring disaster risk finance strategies and Jinks; and Clarke, Coll-Black, Cooney, that ensure disaster risk analytics are based on and Edwards—develop and then apply a sound physical science. methodology to quantify the costs of different combinations of budgetary and financial To sum up, this body of research presents a instruments that can be used to finance a compelling case for disaster risk finance as a disaster response. The approach results in a tool for development. But the details matter. simple formula to capture the opportunity cost 8 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 9 INCREASING COMMITMENT THROUGH DISASTER RISK FINANCE With Disaster Risk Finance plans are dependable 10 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT Dull Disasters? How Planning Ahead Will Make a Difference Daniel Clarke, The World Bank Group Stefan Dercon, Centre for the Study of African Economies, University of Oxford, and U.K. Department for International Development Natural disasters remain all too common, and A Plan—But Not Just Any Plan. Good planning is the aftermath of such disasters is full of high- based on an iterative dialogue among scientists, stakes political leadership and debate, media bureaucrats, implementers, and financiers about attention, public appeals, and well-intentioned what or who is to be protected, how it or they actions. Yet well-intentioned responses by are to be protected, and what the cost will be. governments and the international community Bad planning happens when at least one of these often fall short of their aims. In this book, Clarke parties is missing from the dialogue. Planning and Dercon (2016) argue that the fundamental is a political choice; it is not just a technical problem is the funding model, whereby after a exercise. Political statements by governments or disaster, farmers and homeowners, subnational development partners about how much money governments, and national governments are would be made available or how many people required to plead for help to benefactors, would be mobilized in the event of a disaster are such as subnational governments, national not conducive to good planning. Useful political governments, and the international community, statements focus on target outcomes and leave all of whom retain discretion over how to the details on the “how” to be worked out by the allocate their budgets until after a disaster implementing agencies and financiers. strikes. This ad hoc post-disaster funding model does not work well. It is too slow, leads to a Benefactors who want to maximize the fragmented and underfunded response, and development impact of their support should encourages underinvestment in risk reduction think through different natural disaster and preparedness, thereby increasing the scenarios, assess what support they would economic and human costs of catastrophes. provide in each scenario, and own up to this contingent liability when in discussions with The Solution? The solution is for governments other partners. A benefactor with either no and their partners to adopt pre-agreed, pre- contingency plan or its own stand-alone financed, rules-based preparedness plans that contingency plan will fall short in its efforts can be implemented after a disaster strikes to help people. Benefactors can channel their without the need for further political decisions. financial support into precise sets of plans in Specifically, the responses to disasters can be which it is clear who exactly is being protected, more business-like and more effective (indeed, how, and who is paying. duller) if three things are in place beforehand: Behavioral biases against good planning are 1. A sound, coordinated plan for post-disaster strongest for the kinds of disasters that did not action agreed in advance occur in the recent past—that is, for nearly all 2. A fast, evidence-based decision-making future disasters. To combat these biases, there process is a particular need to invest in science-based 3. Financing on standby to ensure that the plan risk information and clear communication of can be implemented. this information to ensure that everyone knows what contingencies they need protection for. DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 11 Sound Decision Making—But Based on Good enough to withstand the whirlwind of highly Rules, Good Data. By ensuring that as little as charged post-disaster politics (see table). possible must be decided by stakeholders when a disaster strikes, rules can promote decisive, When designing and implementing disaster timely action. The data driving these decisions risk finance strategies, details matter. Financial need to be resistant to manipulation and experts add value. It is important to pay for strike the right balance among cost, speed, and financial advice and build in-house expertise. accuracy. The triggers in the financial strategy should Any data that could trigger action will depend match the triggers in the plan. Traditional on investments before a disaster in design of reinsurance can be particularly useful for the data collection system, including an audit locking in plans for reconstruction, and indexed function, and in the human and technological reinsurance can play the same role to finance capacity to collect data in a timely manner. indexed early actions. Three types of data are particularly useful for triggering post-disaster action: ground data on Partially subsidized financial instruments can be the damage to or losses of people and buildings, used to encourage others to contribute toward area average index data on damage and losses, the cost of well-defined plans. and parametric indices. Leaders should focus on providing protection, No rule is perfect, and so there should be not relief, and using financial incentives to some discretionary backup system to deal with encourage others to own up to and finance their situations in which the rules fail. share up-front. Ad hoc, post-disaster support is still needed, but it should act as a backup Standby Financing—But Based on Smart Choices when plans fail. It should not be the first line of Instruments and Triggers. Financial and of defense for droughts, floods, earthquakes, budgetary instruments are the glue that hold tropical cyclones, or pandemics. credible plans together and make them strong FINANCIAL AND BUDGETARY INSTRUMENTS Goal Ex ante instrument Ex post instrument (arranged before a disaster) (arranged after a disaster) Risk retention Contingency fund or budget Budget reallocation (changing how or when one allocation pays) Tax increase Line of contingent credit Post-disaster credit Risk transfer Traditional insurance or reinsurance Discretionary post-disaster relief (removing risk from the Indexed insurance, reinsurance, or balance sheet) derivatives Capital market instruments 12 6 Conclusions DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT Using the model built in this paper, we have been able to analyse a number of commitment problems arising in disaster risk finance. One important result to note from this analysis is that the effects of these various commitment problems frequently differ. For instance, whilst the Samaritan’s dilemma causes recipients to protect themselves too little, fear of aid misallocation may result in them protecting themselves in inefficient ways. As a result, we cannot recommend a single solution as being appropriate for all situations. Instead, it is Disaster Solving Commitment Problems inimportant Risk to diagnose which Finance commitment problems are most severe and then choose the system of disaster risk finance that is most adapted. Daniel Clarke, The World Bank Group Table 1 provides a first attempt to try to categorize how policies involving private risk Liam Wren-Lewis, Paris School of Economics finance may interact with various commitment problems. Pluses represent cases where a policy should mitigate a commitment problem, minuses cases where the policy may worsen such a commitment problem, and zeros mean it is not clear there would be any impact. Clarke and Wren-Lewis (2016) examine the DRF INSTRUMENTS AND COMMITMENT PROBLEMS ways in which risk transfer instruments— Table 1: Private finance solutions for commitment problems insurance, reinsurance, derivatives, and capital Recipient Common Benefactor Disaster market instruments—can act as commitment Commitment problem insurance insurance payout index subsidies triggers devices, helping governments and development partners to commit ahead of the disaster to Samaritan’s dilemma + + 0 + restrict their post-disaster discretion for the Aid misallocation + 0 +/- +/- good of the country. Delayed disbursements + + 0 0 Note 1 +/- policy mitigates/worsens the commitment problem, 0 effect unclear. They identify three distinct problems that The Samaritan’s dilemma can be mitigated by insurance for recipients or by insurance for benefactors. Recipient insurance decreases the need for discretionary aid, whilst benefactors can arise from an inability of benefactors to can simultaneously invest in disaster insurance, to provide finance when it is needed, and commit: reduce functionality the their access to more and credibility discretionary of funds the and lines of credit, to increase contingency respective relief the marginal cost institutions of post-disaster finance. not feasible. isDisaster indices may also help, though only if they are sufficiently accurate that benefactors are then not sufficiently motivated to learn the true 1. Disaster relief may be prone to a moral hazard Instead, investing in a system of risk transfer to needs of recipients. problem and the classical “Samaritan’s third parties A simple waycould be to reduce a misallocation aid more effective solution is through subsidizing recipient insurance, assum- dilemma” in particular. Those at risk and hasthe ing that become ofreliable part is private sector enough disaster countries’ risk Combining payout triggers not to misallocate. deliberately underprotect themselves finance strategies. 21 because they know governments or donors will come to their rescue. Clarke and Wren-Lewis consider four 2. Benefactors do not undertake the steps needed properties of schemes to transfer risk to third to avoid the misallocation of disaster relief. parties, each having different implications for Many who should receive relief do not, and the commitment problems (see table). sometimes funds are diverted to those who suffered no losses at all. Before a potential • Recipient insurance subsidies. Benefactors disaster, benefactors would like to reduce purchase or mandate the purchase of insurance misallocation, but if they cannot commit to for the poor and vulnerable. Or benefactors doing this, recipients will self-insure. This subsidize a fixed or proportional part of the serves to diminish the incentive to pay to premium payment. reduce misallocation. • Benefactor (re)insurance. Benefactors, prior to disaster, coordinate insurance coverage 3. Finally, disaster relief frequently arrives too with donors and purchase insurance. late. Besides practical reasons for relief not • Common payout triggers. Benefactors ensure arriving in a timely fashion, benefactors uniform relief triggers for public monies may wait to see what others give before and private insurance. giving. This strategy may be motivated by • Disaster indices. Benefactors gather and the wish to gain clarity on burden sharing publish statistics on disaster-loss proxies among donors before making payouts. (such as satellite data on wind speed, rainfall) and construct disaster-loss indices Especially in countries with poor governance, to trigger payout. solving commitment problems by improving DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 13 Disasters and Discipline: The Political Economy of Natural Disasters and of Sovereign Disaster Risk Finance and Insurance in Mexico Laura Boudreau, The World Bank Group Natural disasters are often highly political Governors and the federal government respond events that are closely followed by the to the incentives set by voters—requests and public—especially the government’s response. approvals for funds from FONDEN increase Evidence from around the world suggests that in presidential election years. Moreover, post- voters’ responses to these events may give disaster relief increases in the run-up to these governments incentives to prepare for and elections. These results do not imply that the respond to natural disasters in ways that are FONDEN system is ineffective at disciplining suboptimal, or costly, for society. politicians during election cycles—the A 2011 policy behavior in the absence of FONDEN is change triggered a rise in the In light of this dynamic, one possible benefit unknown. However, FONDEN may be helping likelihood of of disaster risk finance instruments may be to to address a commitment problem resulting municipal sectors discipline governments and other benefactors from elections that would otherwise be more insuring by 15 – 25 to abide by rules and commitments determined severe. percent. before disasters occur. Specifically, disaster risk finance instruments may help governments In fact, the evidence is consistent with credibly commit to cover certain risks and help FONDEN helping to discipline benefactors. to hold governments accountable to voters. In the early years of FONDEN, almost all The role of disaster risk finance instruments applications for funds were granted by the as commitment and accountability devices, federal government. Over time, the ratio however, is largely unexplored. of approvals-to-applications has fallen significantly, particularly for events when Boudreau (2016) provides preliminary parametric thresholds are used to determine empirical evidence that risk financing municipalities’ eligibility. Furthermore, instruments can serve as a commitment device FONDEN’s efforts to increase the insurance based on the experience of Mexico’s Fund for coverage of assets appear to be successful. National Disasters, FONDEN. The author’s In 2011 FONDEN implemented rule changes findings support the claim that Mexico’s that promote the take-up of insurance and disaster risk finance program has disciplined increase the overall coverage of assets. Since politicians in light of the incentives provided that time, the proportion of loss events with by voters. insurance relative to those without insurance has increased markedly. At the municipality Mexican voters punish politicians for the level, events after the 2011 policy change are occurrence of natural disasters in the run-up about 15–25 percent more likely to be covered to elections but reward them for the allocation by insurance. of post-disaster relief. Voters also respond to delays in post-disaster reconstruction by punishing the incumbent political party in the upcoming election. 14 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 15 THE BENEFITS OF TIMELY RECONSTRUCTION Without Disaster Risk Finance fast reconstruction does not happen 16 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT Indirect Cost of Natural Disasters and an Economic Definition of Macroeconomic Resilience Stéphane Hallegatte, The World Bank Group Being able to quantify the total loss to an in economic output beyond the pre-disaster economy from a disaster is a key component level and can be seen as positive even though when trying to determine the importance a classical stimulus policy could have the of disaster risk finance. Essentially, two same effect without the negative welfare and categories of losses must be considered. human impacts of the disaster. Direct losses are the assets lost because of the disaster. Indirect losses, also called output In this model, one dollar of direct loss in losses, are the reduced production and income productive capital translates into a decrease stemming from the asset losses, including in instantaneous (annualized) output that is all adverse long-term consequences for equal to the average productivity of capital. economic growth resulting from the disaster. This decrease in output is about three times The latter can be extremely challenging to the interest rate and may be increased by a quantify accurately. factor that represents ripple effects and the duration of reconstruction. Hallegatte (2015) develops a theoretical model to motivate a simple, intuitive rule of Along these lines, the rule of thumb includes thumb for measuring output losses. The rule the interest rate, the decreasing return in takes into account constraints that render it the production function (also equal to the impossible to immediately reallocate assets to share of profits in national income), and the their most efficient uses, as well as ripple and instantaneous and dynamic resilience—that is, stimulus effects. the ability to limit damage and reconstruction. The latter comprises the reconstruction Ripple effects appear in infrastructure and duration (longer reconstruction increases utility services—for example, a house itself welfare losses) and a ripple effect factor that may not be damaged, but the owner still has to increases or decreases immediate losses. This relocate because there is no running water. Or factor is negative if enough idle resources are the damaged part of a road prevents the rest of available to cope; positive if cross-sector and the road from being used. More generally, the supply-chain issues impair the production of stock of capital consists of complementary nonaffected capital. assets; the destruction of one part may reduce the productivity of other parts. A disaster that causes capital losses equal to US$500 million, for example, in a country The stimulus effect refers to the ability of with a 10 percent interest rate and with a an economy to react to the new production reconstruction period likely to span three constraints (such as road closure) and the years would lead to total (asset plus output) increase in demand for reconstruction losses of US$725 million (145 percent of direct through input substitution, production capital losses), with a discounted value of rescheduling, or mobilization of existing idle US$650 million (130 percent of direct losses). resources. This effect can lead to an increase DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 17 Insuring Growth: The Impact of Disaster Funds on Economic Reconstruction in Mexico Alain de Janvry, University of California–Berkeley Alejandro del Valle, Georgia State University Elisabeth Sadoulet, University of California–Berkeley Given the cost of financing post-disaster The researchers use a regression discontinuity reconstruction, it is important to understand design to identify this effect. They turn to whether such reconstruction can have a high-resolution satellite measures of night considerable impact on local economic lights as a proxy for the differential economic activity. This research presents unique performance created by the provision of rapid evidence on the impact of disaster risk finance reconstruction funds through FONDEN. And FONDEN boosts on economic activity in a long-established they use quarterly employment data as a high- local economic program for the reconstruction of public frequency measure of local employment after activity between assets. a disaster. 2 and 4 percent in the year following Federal and state governments in Mexico Overall, they estimate that FONDEN boosts an event. spend almost US$1.5 billion annually on the local economic activity between 2 and 4 reconstruction of public assets and low- percent in the year following an event. income housing after natural disasters, and this amount can be much larger in bad years. Is this effect worth the resources absorbed In 2010 alone, the reconstruction after major by FONDEN? The researchers find that the floods totaled over US$5 billion. In response benefit to the economy in the year after the to the recurrent need for post-disaster budget disaster is substantially greater than the total reallocations to finance reconstruction, the government expenditure, with a benefit–cost Government of Mexico established the Fund ratio of between 1.52 and 2.89. Although this for Natural Disasters (FONDEN) in 1996. Its range of ratios is quite broad, it does suggest original mandate was to provide adequate that FONDEN is likely to provide benefits in financial resources for federal and state excess of its cost. reconstruction efforts without compromising committed government spending. The scale of gains to local economic activity brought about by the availability and rapid De Janvry, del Valle, and Sadoulet (2016) disbursement of disaster funds in the Mexican focus on the peril that represents 68 percent program could encourage policy makers in of all events that led to financial support other countries to consider using disaster risk from FONDEN—heavy rainfall. They analyze finance schemes such as FONDEN to enhance a 10-year period (2004–2013) in which road their own response capabilities. reconstruction dominated expenditures and estimate FONDEN’s impact on economic activity at different points in the post- disaster period. The expected effect, quickly reconstructed infrastructure and housing accelerated the resumption of economic activity. 18 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 19 THE BENEFITS OF TIMELY SUPPORT TO LIVELIHOODS Without Disaster Risk Finance early response does not happen 20 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT The Application of a Probabilistic Catastrophe Risk Modelling Framework to Poverty Outcomes Catherine Porter, Heriot-Watt University Emily White, The World Bank Group Porter and White (2016) explain the power that drought in the context of other household/ probabilistic catastrophe (cat) risk models could community characteristics. The authors have if applied to the assessment of household consider the consumption–drought relationship poverty outcomes. The authors argue that the will be attenuated by certain household and challenge in applying cat risk models in this community characteristics: access to coping way is quantifying the relationship between strategies (such as education); the occurrence hazard and outcome in a poverty context—the of other shocks (such as illness); and access “vulnerability module” in a cat risk model. The to institutional coping strategies (such as authors attempt to derive such vulnerability public safety nets). These factors are examined relationships for the impact of drought on explicitly within the model. The results are households in Ethiopia, asking whether a then tested for robustness using Statistical relationship can be derived between drought Learning Methods to infer applicability of the and household consumption that has internal derived relationships within a cat risk modelling and external validity and, if so, can help (1) framework. model risk (in a probabilistic framework) and (2) understand the benefits of interventions, According to the regression results, the impact including early response. of drought (represented as the LEAP drought variable) is significant across all models Porter and White examine the impact of drought examined, with the baseline result showing that hazard on the welfare of rural households for every 10 percent worsening of the LEAP in Ethiopia using reductions in household drought variable, consumption falls on average consumption as their welfare indicator and crop- by 1.5 percent. The other models typically show yield losses for drought. They use household about a 2 percent fall in consumption per 10 income and consumption expenditure survey percent drought worsening. The results also data as well as welfare monitoring surveys for reveal that access to a safety net (Ethiopia’s 2005 and 2011 to provide representative cross- Public Safety Net Programme, PSNP) mitigates sectional data for rural Ethiopia, including the drought impact by 0.5 percent—that is, household characteristics, consumption households with PSNP access experience a outcomes, and measures of realized shocks 1.5 percent decrease in consumption rather such as illness and unemployment. A view of than a 2 percent decrease. The results of the drought hazard is taken from the World Food Statistical Learning exercise suggest that the Programme’s Livelihood Early Assessment and relationship between drought and consumption Protection (LEAP) data, a drought measure is fairly homogeneous and stable, leading showing expected community crop losses based Porter and White to conclude (with caveats) on water adequacy specific to the respective that the derived drought–poverty relationship crops. demonstrates some level of external and internal validity. Therefore, this relationship A regression model is defined, where the could form the basis of a vulnerability module outcome of interest is the log of household in a catastrophe risk model. consumption per adult, to be examined against DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 21 Effects of Timing of Public Work Payments on Welfare: The Case of Ethiopia’s Productive Safety Net Programme Guush Berhane, International Food Policy Research Institute Mehari Hiluf Abay, International Food Policy Research Institute John Hoddinott, Cornell University The Government of Ethiopia has developed payments reduce the household consumption and implemented its Productive Safety Net expenditure, although the observed effect is Programme (PSNP), covering nearly 8 million weak. In addition, delayed payments reduce Ethiopians. Beneficiary households can engage agricultural income—perhaps because the in public works in the months when labor is households who do not receive timely payments slack in return for income designed to increase are forced to sell their crops when prices are the food security of households during the lean at their lowest. Finally, payment delays are season. The program is also scaling up in times also found to increase the probability that of disaster. children of primary school age will drop out of school, and therefore the delays decrease Berhane, Abay, and Hoddinott (2016) look at educational attainment. These negative welfare whether increasing the ability to provide more effects of payment delays are more pronounced timely support to beneficiaries has an impact on when delays are preceded by bad harvests in welfare. They examine the relationship between the previous season. Conversely, payments For every 10 percent of the existing variation in the timing of payments delayed well into and made in lump sum during drought worsening, to PSNP public works beneficiaries and poverty the lean season are observed to have strong consumption falls outcomes. In doing so, they focus on household positive (unintended) effects on other welfare by between 1.5 and consumption, child-level nutrition, and outcomes, mainly non-food expenditures, 2 percent. schooling outcomes, across four main regions ownership and total value of livestock owned, that have implemented the PSNP since 2005. value of productive assets owned, food gap, and net private transfers. These latter effects suggest The PSNP makes monthly payments for work the lumpy structure of delayed payments likely conducted and is designed to provide income has an investment role albeit at the cost of lost to households in advance of the lean season. welfare in terms of consumption, nutrition, and However, actual implementation varies by schooling during the public works months. region and over time: 57 percent of public works participants reported no delay; 27 percent, In sum, the timing of the payment does affect a small delay of one or two months; and 16 its welfare effect, suggesting that if payments to percent, a delay of three months or more. Very households in need become more timely (that is, few payments are delayed by more than four arrive in advance of the lean season), they would months, and once monthly payments are have a larger impact on welfare. The findings also started, they are likely to continue smoothly. suggest delayed lump sum payments in the lean season may result in unintended positive effects When payments are delayed prior to the start on some outcomes that should be weighed of the lean season (June–September for most of against the welfare losses in the months they are Ethiopia’s highlands), household consumption meant to be paid. and nutrition are negatively affected. Delayed   22 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT Early Warning, Early Action: The Use of Predictive Tools in Drought Response Through Ethiopia’s Productive Safety Net Programme Mareile Drechsler, The World Bank Group Wolter Soer, The World Bank Group Drechsler and Soer (2016) examine how Requirements Satisfaction Index (WRSI). In Ethiopia’s early warning system could be used the process of computing beneficiary numbers, to enable early action to respond to drought. the LEAP tool computes three intermediary And, they find three clearly defined, transparent, outputs: planting dates, WRSI and yield and complementary systems which are already reduction estimates. By combining the WRSI, in place: the Livelihood Early Assessment and past beneficiary numbers, and demographic Protection (LEAP) tool monitors food security data, the LEAP tool estimates beneficiary among the rural population using a water numbers. index; the Livelihood Impact Assessment Sheet (LIAS) captures bottom–up information on LIAS employs a risk-modeled Household local livelihoods and markets; and ad-hoc and Economy Approach (HEA). Its components zone hotspot assessments identify deteriorating food livelihood and market systems geographically; security situations. These tools, when used set wealth categories within the zones; calculate together, could inform early response. the average cash income and food intake for each wealth category (baseline); model the impact of Early response to drought is crucial in droughts on livelihood baselines; analyze coping protecting lives and livelihoods while saving capacities; and estimate beneficiary numbers costs. Response through the humanitarian according to livelihood viability and collapse system requires fundraising and can take up to under duress. eight months. In contrast, response using early- warning data triggers contingency financing and Hotspot and ad-hoc assessments deploy uniform can respond in as little as two months—four food security and nutritional criteria to ensure times faster. comparability across regions. Quarterly, they prioritize the use of scarce resources to enable By using the three tools’ intermediate and final targeted supplementary feedings with the outputs effectively, LEAP, LIAS, and hotspot support of woreda health workers and offices. assessments could be used to detect the onset of a drought and the need to respond. Moreover, These instruments, Drechsler and Soer as each tool is based on a different methodology, conclude, could represent the building blocks it is possible to compare drought predictions to of a well-functioning early action framework. obtain a more accurate view of the severity of a Ethiopia has consistently extended LEAP and drought. LIAS, producing more and better data and further improving their predictive powers. LEAP, based on a basket of primary crops in a geographic area, estimates the number of people in need of food assistance based on the Water DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 23 How to Measure Whether Index Insurance Provides Reliable Protection Karlijn Morsink, Centre for the Study of African Economies, University of Oxford Daniel Clarke, The World Bank Group Shadreck Mapfumo, The World Bank Group Agricultural index insurance has become a 2. Does the insurance provide coverage for common risk management instrument for low- losses that are important, keeping in mind income farmers. For index insurance, payouts that households face many sources of income correlate with the performance of an index, not risk? The reliability of index insurance for actual losses. An imperfect correlation—basis smoothing agricultural production can be risk—means the index may pay out when no evaluated by comparing claim payouts to losses occur and may not pay out when losses actual losses from agricultural production— do occur. The impact on poverty is thus highly that is, going beyond the risks stipulated sensitive to the reliability of the coverage. Until in the contract. For example, if a farmer is now, the lack of an operational and measurable insured against floods, losses from a drought definition of basis risk and the underutilization are not covered. And although her insurance of appropriate statistical techniques have product performs perfectly during droughts, precluded monitoring this reliability. the farmer still suffers a production loss. Morsink, Clarke, and Mapfumo (2016) Morsink et al. devise two indicators to measure discuss the reliability of index insurance and insured-peril and production-smoothing basis propose monitoring indicators that, with basic risk and thus the reliability of index insurance. technical knowledge, can be applied by donors, The first, the probability of catastrophic basis governments, and insurers to any context in risk, assesses the probability of not receiving a which payouts are based on indices correlated claim payout when a farmer has catastrophic with losses. losses from agricultural production. The second, the catastrophic performance Establishing whether index insurance reliably ratio, measures what a farmer receives back protects low-income individuals against losses relative to the premium paid when the farmer from agricultural production requires answering experiences catastrophic losses. two key questions: These indicators are simple to calculate and 1. Does the insurance provide reliable coverage easy to understand, providing proxies for the of the losses it was designed to insure? The reliability of indexed protection. They can be performance of the index can be measured used to compare agricultural insurance products by assessing the basis risk of the insured against a benchmark, to one another, and over peril (such as drought and flood). This time, and to generate the cost and effectiveness assessment will reveal how well the index of alternative disaster risk finance instruments. insurance product actually captures losses They provide invaluable inputs for disaster risk caused by the insured peril—that is, does finance strategies, while improving the quality drought insurance pay out when there is a of products, better protecting consumers, and drought? reducing reputational risk. 24 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT Integrating Social Protection Strategies for Improved Impact: A Comparative Evaluation of Cash Transfers and Index Insurance in Kenya Nathaniel Jensen, Cornell University Munenobu Ikegami, International Livestock Research Institute Andrew Mude, International Livestock Research Institute Social safety nets are an important tool used including the benefits of insurance transfers by policy makers to support and protect versus insurance subsidies and the extensive their constituents. Recently, there has been scaling of cash transfers conditional on a push to increase the sophistication of the environmental conditions. targeting mechanisms and to combine multiple protection tools to improve the efficiency of Kenya, which is currently implementing a set of these programs. This effort is motivated by the social protection schemes, offers an excellent recurring humanitarian interventions in regions opportunity to study the effects of different that have existing social protection programs. protection strategies and simulate the welfare and fiscal gains of integration. The authors use Although a number of countries have or are data from five annual rounds of panel household planning to implement integrated social survey data and instrumental variables to protection programs, there is little existing identify the observed impacts of a cash transfer empirical evidence on the welfare outcomes program, the Hunger Safety Net Program, and from these programs. Not only might the added insurance, the Index Based Livestock Insurance logistical burden of implementation offset (IBLI) product, on pastoralist households in potential gains, but it is still not clear that northern Kenya. Those parameter estimates are welfare dynamics are sufficiently homogeneous then used to simulate the impacts of a menu of to make sophisticated targeting around specific hypothetical social protection policies, which inflection points in wealth a realistic objective. would each have the same cost. Jensen, Ikegami, and Mude (2016) examine Jensen et al. find for the most part very few how to develop evidence-based policy differences in the poverty outcomes associated recommendations by studying the impacts of with the various targeted approaches. Although various social protection programs on poverty. this apparent ambiguity may seem disappointing, They test whether there are poverty benefits it frees policy makers to develop their social to using an integrated approach to social protection strategies with additional objectives protection, providing differential programs in mind—for example, to pursue strategies for the ultra-poor and those who are less poor with the lowest overhead or to support the but still vulnerable. Of particular interest is development of a robust insurance market. the use of cash to help the poorest and the use of insurance to protect the vulnerable from This study is not without its own shortcomings, falling into poverty during large shocks. They which the authors discuss and use to highlight also examine other policy-relevant questions, the need for further research. DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 25 Weather Index Insurance and Shock Coping: Evidence from Mexico’s CADENA Program Alain de Janvry, University of California–Berkeley Elizabeth Ramirez Ritchie, University of California–Berkeley Elisabeth Sadoulet, University of California–Berkeley Few tools have been successfully implemented payments increase the hectares sowed of insured at scale for the rural poor to manage the risks of crops by about 17 percent relative to those in weather shocks and to cope in their aftermath. municipalities that do not receive payments. Mexico is an exception. Its government pioneered a weather index insurance program The impact of drought index insurance when in 2003 that by 2013 insured more than 6 million comparing a municipality that receives no hectares of cropland. CADENA grew from a payment with one that receives payment is an drought index insurance for maize in one state increase of about 27 percent in expenditure per to a near-national insurance program for many capita and 38 percent in income per capita. This perils and crops. increase corresponds to about 6,000–8,000 pesos in additional income. However, results De Janvry, Ramirez Ritchie, and Sadoulet (2016), suggest that the insurance transfer induces recognizing the expansive coverage and tenure a reduction in remittances sent by migrants, of the CADENA program, use the unique setting lowering the net income effect of the payouts. to evaluate index insurance’s effects on ex post CADENA insurance production decisions and coping mechanisms. Turning to the cost side, premium payments payments to They focus on the program’s largest component exceed indemnity payments in all but two municipalities increase the historically, drought index insurance. years, resulting in an overall loading factor of hectares of insured 73 percent. Although these results suggest that crops by about CADENA insures farmers growing staple crops the cost of insurance is high relative to the 17 percent. on less than 20 hectares of rain-fed land. If payouts received, a cost–benefit analysis using precipitation as measured by the corresponding the increase in household income implied by weather station falls below the designated the regression estimates finds that the benefits threshold in any of the three phases, the insurer of the program exceed the costs for a wide makes a payment to the state. The state then range of estimates. Moreover, CADENA, by transfers payments to eligible farmers in the design, makes government expenditures more insured area in time for the next growing season. predictable and disciplines the responses of state Overall, results comparing municipalities governments to weather shocks. Recognizing that receive payments with those that do not these benefits, the federal government provides show the federal government–funded program subsidies of up to 90 percent for CADENA helps sustain rural livelihoods, mitigating while simultaneously increasing the required the losses from drought without the need to contributions of uninsured states seeking funds assess individual damage. Another finding was for ex post relief. farmers in municipalities that receive insurance 26 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 27 SAVING MONEY THROUGH DISASTER RISK FINANCE Without Disaster Risk Finance disaster response is too expensive 28 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT Evaluating Sovereign Disaster Risk Finance Strategies: A Framework Daniel Clarke, The World Bank Group Olivier Mahul, The World Bank Group Richard Poulter, The World Bank Group Tse-Ling Teh, London School of Economics Ministries of Finance of disaster-prone methodology builds on insights from actuarial countries, along with donor partners who are science and financial economics to divide the also facing rising costs from disasters, are problem in a way that makes ex ante evaluation increasingly asking such questions as: of the financing side of the problem possible. • Should we set aside funds in a reserve fund, Specifically, the paper considers the case in which and how large should this reserve fund be? a government has chosen a set of fixed responses • How much reliance should be placed on for possible future disasters and wishes to emergency reallocations of funds away understand the costs and benefits of financing from other parts of our budget to finance these responses through various combinations of disaster losses? financing instruments (see figure). • Should we seek to establish a line of credit on which we can immediately draw if a 15 disaster were to occur? • How can we evaluate proposals for risk 12 transfer products such as disaster insurance or catastrophe bonds? 9 US$ millions These questions, and others about how to 6 decide on the details of disaster risk finance strategies, have been difficult to answer, in 3 part because of lack of a robust methodology that allows a full range of budgetary and 0 financial instruments to be compared side by side in a consistent, comprehensive way. This -3 Average 1 in 5 year 1 in 15 year 1 in 50 year methodological limitation has meant that disaster disaster disaster strategies may be chosen and implemented Cost saving of A Cost saving of B Cost saving of C without systematic analysis of whether the over Base Case over Base Case over Base Case programs and financial strategies being Source: Clarke, Mahul, Poulter, and Teh 2016. employed are appropriate and cost-effective, bearing in mind the risks faced. The framework is flexible enough to be useful To begin to solve the problem of this for decision makers concerned with different analytical gap, Clarke et al. (2016) develop a aspects of disaster. For example, it could be robust, comprehensive methodology to allow used to calculate the long-run average cost quantitative analysis of the full economic or the cost for specific potential extreme cost of these financial instruments. The disasters.  DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 29 Evaluating Sovereign Disaster Risk Finance Strategies: Guidance and Case Studies Daniel Clarke, The World Bank Group Naomi Cooney, The World Bank Group Anna Edwards, U.K. Government Actuary’s Department Andrew Jinks, U.K. Government Actuary’s Department Clarke, Cooney, Edwards, and Jinks (2016) 3. Set base assumptions. Assumptions are set apply the framework developed in Clarke, in reference to the economic and political Mahul, Poulter, and Teh (2016) to five conditions of the underlying country, practical case studies, and present a guidance simplified to avoid identification of the note on how the framework can be applied countries. in practice. In doing so, they illustrate the 4. Calculate the opportunity cost of each strategy. flexibility of the framework and its ability to For each strategy, an analysis is presented be used by governments to systematically for the financing cost both on an average determine whether their financial strategies basis and for different shock severities using are appropriate and cost-effective in view of the previous assumptions made about the the risks they face. economic environment and the probability and magnitude of the events. Application of the framework to the five anonymized, simplified, real-world countries 5. Consider sensitivity and scenario testing. Each involves the following steps: case study includes a sensitivity analysis in which assumptions and specifications 1. Define the contingent liability. To enable are varied to illustrate how the costs might quantitative analysis, implementers have to change. decide on a set of rules that would trigger expenditures. The contingent liabilities The report does not make any generalized considered in this report range from conclusions about which instruments or country-wide response costs arising from strategies are cheapest. Instead, for each drought to an insurance program covering country, results and sensitivities are presented. public emergency losses in multiple regions The most cost-effective strategy for each case of a country arising from earthquake and study depends on the risk tolerance of the tropical cyclone events. relevant policy makers.   2. Specify the choice of financing strategies. Each case study compares at least three alternative financing strategies, including a base strategy against which alternative strategies can be compared. 30 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT A Methodology to Assess Indicative Costs of Risk Financing Strategies for Scaling Up Ethiopia’s Productive Safety Net Programme Daniel Clarke, The World Bank Group Sarah Coll-Black, The World Bank Group Naomi Cooney, The World Bank Group Anna Edwards, U.K. Government Actuary’s Department Rural safety nets in low-income countries additional beneficiaries) ranges from US$175 remain a challenge to develop, yet the to $230 million. Government of Ethiopia has developed and The average implemented the Productive Safety Net Both the FCB and budget reallocation are cost of financing Programme (PSNP), providing nearly 8 million depleted in a significant proportion of the the US$139 Ethiopians with the means to work their way 5,000 simulated scenarios. Because of the million average out of chronic poverty. assumed layer of insurance available and the liability ranges relatively low pricing multiple (1.35 compared from US$175 to US$230 million, Clarke, Coll-Black, Cooney, and Edwards with an opportunity cost of 2.0 applied to the depending on (2016) adapt the framework of Clarke, Mahul, HRD), strategy B is the cheapest on average. As the disaster risk Poulter, and Teh (2016) to comparatively the figure shows, the cost savings of insurance finance strategy. analyze potential risk finance structures that also rises for more severe droughts; the results support drought response through the PSNP. of the costs of a one-in-five-year and a one-in- They define a hypothetical version of the 30-year event demonstrate this clearly. PSNP in which woredas (districts) receive automatic financing based on an early warning Finally, the paper highlights that financial cost system that is tied to a water deficit index. is only one component of the risk financing Under these hypothetical “rules,” the PSNP decision, and that other aspects need to be scale-up supports annually, on average, 2.9 considered for any practical recommendation. million transitory poor, requiring an average 1000 expenditure of US$139 million per year. Average cost 1 in 5 year cost 1 in 30 year cost Humanitarian response Three primary hypothetical risk strategies are 800 then considered to finance these expenditures. The initial instrument in all strategies is the Budget reallocation Opportunity Cost federal contingency budget (FCB), which must 600 be exhausted before other instruments can be applied. Unlimited humanitarian response Insurance 400 (HRD) is always assumed to be a last resort. The base case, strategy A, includes the FCB Federal and HRD only; strategies B and C consider a 200 Contingency layer of insurance and budget reallocation, Budget respectively, between the two (see figure). 0 Strategy A Strategy A Strategy A Strategy B Strategy C Strategy B Strategy C Strategy B Strategy C Under the best estimate assumptions, the average cost of financing the US$139 million average liability (average of 2.9 million Source: Clarke, Coll-Black, Cooney, Edward 2016. DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 31 Financing the Reconstruction of Public Capital after a Natural Disaster David Bevan, University of Oxford Christopher Adam, University of Oxford Using a macroeconomic model, Bevan and fall in the range of 12–15 percent, higher than Adam (2016) simulate the effects of alternative the tax alternative, but much lower than the post-disaster financing mechanisms when reallocation. Overall, direct tax financing increased foreign borrowing is impractical. appears to be the most attractive option. They examine sovereign disaster risk insurance, Yet it requires that raising taxes be a feasible increased taxation, and budget reallocation option. In the chosen model application, tax- Post-disaster as alternative financing mechanisms for financed reconstruction would require a tax budget rebuilding public capital. The model measures increase of 1.5–2 percent over a full decade. reallocations away costs and benefits in terms of real household The researchers propose that for high-damage from operations consumption and decomposes the expected rare events insurance could be a better choice and maintenance net cost of the public finance responses into because of the faster pace of reconstruction. result in an opportunity gross benefits from the rebuilding of the public In any case, budget reallocation is associated cost as high as capital stock and the gross cost of mobilizing with the most costly and substantially slower 37–44 percent. the required fiscal resources (that is, the recovery in all scenarios. These results span opportunity cost). a limited range of financing options, and it is straightforward to extend the analysis Although their model can be adjusted to fit to examine “blended” financing packages, various financing and disaster scenarios, they including ones incorporating debt financing. use the risk profile and national accounts data from Jamaica as an example and simulate a This methodology does not adequately capture number of possible outcomes for a cyclone some aspects of disaster risk finance that merit disaster. In addition to doing nothing, the further research. First, this analysis assumes government has three choices: (1) to take that the risk transfer products perfectly out full insurance with premiums financed match the government’s expenditure rules. from taxation; (2) to reallocate public It would be useful to extend the analysis to expenditure away from recurrent operations accommodate the case in which payments and maintenance expenditures; or (3) to raise from risk transfer instruments do not taxes. Their calculations estimate the internal synchronize with expenditures. Second, many rate of return of faster reconstruction to be in of the assumptions required for practical the range of 11–17 percent. The opportunity implementation of this methodology, such as cost of funds is more variable, depending on the opportunity cost of budget reallocations the financing choice. For the tax financing or the cost of delayed response, are based on regime, the opportunity costs are lowest, at limited evidence. The analysis would benefit 6–9 percent, measured on an internal rate from further work addressing these empirical of return basis. Reallocations away from limitations. Finally, this paper addresses only operations and maintenance, by contrast, responses to the destruction of public capital. result in an opportunity cost as high as 37–44 A natural extension would be to incorporate percent. For insurance, the equivalent numbers the loss and reconstruction of private capital. 32 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT Using Probabilistic Models to Appraise and Decide on Sovereign Disaster Risk Financing and Insurance Roberto Ley-Borrás, Consultoría en Decisiones Benjamin D. Fox, The World Bank Group Ley-Borras and Fox (2015) provide an overview Strategy generation tables can help anyone of how catastrophe (cat) risk models can be seeking to design a small number of coherent, used to appraise disaster risk finance strategies effective, and affordable strategies. The tool and identify the impacts of natural disasters on is easy to use and can be scaled to the desired the poor and vulnerable. Probabilistic cat risk level of detail, with the type of financial models typically comprise four modules: instrument heading the columns and the specific alternatives in the body of the table. 1. Hazard module: a catalog of simulated natural hazards. Integral decision analysis brings all of the interested 2. Exposure module: a database of the physical parties to the process, helps them understand the characteristics of assets and households pros and cons, and moves the parties forward in that are at risk from specific events in the unison through climatic events. Four valuable hazard module. steps are: (1) define the scope, (2) structure 3. Vulnerability module: a database of generated the objectives, (3) generate alternatives and damage estimates and uncertainty strategies, and (4) identify and measure uncertain parameters. (including manmade) events. 4. Loss module: conversion of the damage estimates into direct and indirect economic In concluding, Ley-Borrás and Fox provide losses. eight suggestions to build catastrophe risk models that facilitate decisions on sovereign Probabilistic cat risk models usually contain disaster risk finance. tens of thousands of event scenarios. As a result, the model produces more risk Infrastructure Population Human estimates (magnitudes) for rarely occurring Characteristics Distribution Impact natural disasters than can be found in the historical records. Cat risk models also provide probabilistic estimates of costs and DRM consequences—key inputs for any formal Strategy Fiscal Overall decision-making process on disaster risk Severe Cost Impact Natural finance instruments. Event Influence diagrams—showing the relationships SDRFI Strategy Economic among decisions, uncertain events, and Impact consequences—support the linked decision making of disaster risk management and disaster risk finance instruments (see figure). Available Response Funds Effectiveness Such diagrams can provide insights and are valuable in gaining consensus and buy-in. Source: Ley-Borrás and Fox 2016. DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT 33 REFERENCES Berhane, G., M. H. Abay, and J. Hoddinott. 2016. “Effects of Timing of Public Work Payments on Welfare: The Case of Ethiopia’s Productive Safety Net Programme.” The World Bank Group, Washington, DC. Bevan, D., and C. Adam. 2016. “Financing the Reconstruction of Public Capital after a Natural Disaster.” The World Bank Group, Washington, DC. Boudreau, L. 2016. “Disasters and Discipline: The Political Economy of Natural Disasters and of Sovereign Disaster Risk Financing and Insurance in Mexico.” The World Bank Group, Washington, DC. Clarke, D. J., S. Coll-Black, N. Cooney, and A. Edwards. 2016. “A Methodology to Assess Indicative Costs of Risk Financing Strategies for Scaling Up Ethiopia’s Productive Safety Net Programme.” The World Bank Group, Washington, DC. Clarke, D. J., N. Cooney, A. Edwards, and A. Jinks. 2016. “Evaluating Sovereign Disaster Risk Finance Strategies: Guidance and Case Studies.” The World Bank Group, Washington, DC. Clarke, D. J., and S. Dercon. 2016. Dull Disasters? How Planning Ahead Will Make a Difference. Oxford, U.K.: Oxford University Press. Clarke, D. J., O. Mahul, R. Poulter, and T.-L. Teh. 2016. “Evaluating Sovereign Disaster Risk Finance Strategies: A Framework.” The World Bank Group, Washington, DC. Clarke, D. J., and L. Wren-Lewis. 2016. “Solving Commitment Problems in Disaster Risk Finance.” The World Bank Group, Washington, DC. De Janvry, A. 2015. “Quantifying Through Ex-post Assessments the Micro-level Impacts of Sovereign Disaster Risk Financing and Insurance Programs.” Development Economics Working Paper Research Series No. 7356. The World Bank Group, Washington, DC. De Janvry, A., A. del Valle, and E. Sadoulet. 2016. “Insuring Growth: The Impact of Disaster Funds on Economic Development in Mexico.” The World Bank Group, Washington, DC. De Janvry, A., E. Ramirez Ritchie, and E. Sadoulet. 2016. “Weather Indexed Insurance and Shock Coping: Evidence from Mexico’s CADENA Program.” The World Bank Group, Washington, DC. Drechsler, M., and W. Soer. 2016. “Early Warning, Early Action: The Use of Predictive Tools in Drought Response Through Ethiopia’s Productive Safety Net Programme.” The World Bank Group, Washington, DC. 34 DISASTER RISK FINANCE AS A TOOL FOR DEVELOPMENT Hallegatte, S. 2016. “The Indirect Cost of Natural Disasters and an Economic Definition of Macroeconomic Resilience.” Development Economics Working Paper Series No. 7357, The World Bank Group, Washington, DC. Hill, R., and H. Neda Fuje. 2016. “Meso-Level Impacts of Disasters: A Case Study Using Ethiopian Data.” The World Bank Group, Washington, DC. Jensen, N., M. Ikegami, and A. Mude. 2016. “Integrating Social Protection Strategies for Improved Impact: A Comparative Evaluation of Cash Transfers and Index Insurance in Kenya.” The World Bank Group, Washington, DC. Ley-Borrás, R., and B. D. Fox. 2015. “Using Probabilistic Models to Appraise and Decide on Sovereign Disaster Risk Financing and Insurance.” Development Economics Working Paper Research Series No. 7358, The World Bank Group, Washington, DC. Morsink, K., D. J. Clarke, and S. Mapfumo. 2106. “How to Measure Whether Index Insurance Provides Reliable Protection” The World Bank Group, Washington, DC. Porter, C., and E. White. 2016. “The Application of a Probabilistic Catastrophe Risk Modelling Framework to Poverty Outcomes.” The World Bank Group, Washington, DC. © 2016 International Bank for Reconstruction and Development / International Development Association or The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. RIGHTS AND PERMISSIONS The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. PHOTO CREDITS If not indicated otherwise, photos used in this publication have been sourced from the following locations with full rights: World Bank Flickr Website The World Bank About DRFIP: The Disaster Risk Financing and Insurance Program (DRFIP) 1818 H Street NW is a joint program of The World Bank Group’s Finance & Markets Global Washington DC 20433 Practice and the Global Facility for Disaster Reduction and Recovery Telephone: 202-473-1000 (GDFRR). DRFIP has helped governments, businesses, and households bring together the financial, analytical & advisory, and convening services of Internet: www.worldbank.org the Bank in more than 60 countries. DRFIP leads the dialogue on financial resilience as a component of the Bank’s support to vulnerable countries on better-managing disasters and climate shocks.