Strengthening Social Protection Systems to Manage Disaster and Climate Risk in Asia and Pacific This report summarizes the knowledge shared and issues raised during a conference convened by the World Bank on the above topic held on November 3-5, 2014 in Manila, Philippines. Building on earlier conferences1 on this topic, the conference aimed to raise awareness about, and share good practice on, building a social protection system that integrates disaster risk management and climate change adaptation. It brought together 17 country delegations from Asia and Pacific region comprised of officials from the ministries managing social protection, disaster risk management, and financing and insurance.2 As such, it was both cross-global practice and cross- regional in focus. The conference offered a range of learning opportunities, including presentations from technical experts from the World Bank, country case studies from Asia, Africa and Latin America, roundtable discussions, and group work sessions. The conference was conducted in collaboration with the Government of the Philippines and was financed by the Rapid Social Response Fund and the Global Facility for Disaster Reduction and Recovery. I. Introduction Natural disasters have been increasing in frequency, variability, and severity in recent years. The Asia and Pacific countries3 are among the most risk-prone in the world – the number of disasters that occurred over the last four decades was highest in East Asia, followed by South Asia.4 The Asia and Pacific regions suffer annually from different types of natural disasters such as floods, cyclones, tsunami, drought, landslides, and earthquakes. While rich and poor countries are not spared from disasters, the loss of lives, devastated livelihoods, and financial strain on households, communities, and governments are much more apparent in poor countries. The destruction of physical infrastructure such as roads, telecommunications, school buildings, healthcare facilities, which usually take time to rebuild in poor countries, has long-term effects on human development. A number of countries have managed to build resilience over time, but many more struggle to deal with disasters effectively. Within a country, poorer populations are the most vulnerable to the adverse impacts of disasters. They usually live in more hazardous locations – riverside, coastal areas, earthquake-, landslide- and drought-prone areas. Damage to productive assets and loss of property and jobs create adverse impacts on the poor, who have fewer savings and are less likely to have insurance protection and options for relocation. In addition to these losses, the poor have to bear the brunt of the lingering effects of disasters such as water-borne and contagious diseases that commonly occur in affected areas. Due to lack of alternatives, the poor tend to resort to coping mechanisms that are detrimental to their human capital such as reduction in food intake, withdrawal from school, or working long hours, thereby exposing them to further risks. If not addressed well, natural disasters can reverse the gains in reducing poverty in the Asian region. Social protection programs and systems have the potential to mitigate the adverse impacts of natural disasters ex ante and build resilience among the population. The key elements of social protection programs – from targeting to financing – can incorporate ex ante mechanisms that lower the vulnerability of the population, minimize costs to the governments, and prevent a protracted social and economic crisis. Ex ante mechanisms include tools and activities related to 1 disaster-risk management, climate change adaptability, and disaster-risk financing and insurance, which are critical to ex post delivery of assistance. The key challenge, however, is to integrate these mechanisms into the social protection systems. Governments commonly rely on post-disaster assessments to identify affected populations and their immediate needs, which may take months to complete and do not necessarily provide delivery solutions. By leveraging existing and already well-functioning social protection platforms, disaster-risk social protection programs can help more effectively target the most vulnerable populations and provide immediate assistance following a disaster. Figure 1. Phases of Disaster Response Ex#Ante# Disaster1 Ex#Post# •  Hazard&informa,on& •  Social&protec,on& •  Social&protec,on& •  Social&protec,on& •  Poverty&mapping& programs&to&build& relief&programs& recovery& resilience& programs& •  Targe,ng& •  Insurance&payouts& •  Insurance& •  Social&protec,on& & mechanisms&for& reconstruc.on& governments&&& programs& households& & In the aftermath of a disaster, governments and the international community have tended to cater to the immediate relief phase and, months later, to the reconstruction phase. In the event of a shock, post-disaster efforts put a strain on government resources and, faced with the urgency and limited capacity, governments do not often tap the most cost-effective sources of financing. Due to the limited funds for post-disaster activities, governments tend to turn to disruptive measures to access quick liquidity for humanitarian relief, such as re-allocating funding from other essential investment programs, drawing funds from general contingency budgets, taking out international credit, or seeking relief support from the international community. Moreover, in the absence of accurate and reliable information related to the risks of disasters ex ante, contingent liabilities to natural disasters become difficult to assess and can pose direct impacts on the governments’ fiscal condition ex post. The recovery phase, when the most vulnerable population is dependent on government support, is one that is not properly addressed. This period (roughly defined as two to eight months after a rapid onset event) is when initial relief efforts have slowed down and reconstruction planning has begun but major works have not yet commenced. Those most negatively impacted by this lull in activity are vulnerable populations that are dependent on government support. The lack of ongoing assistance at a time when their homes have yet to be rebuilt and their sources of income yet to be restored means that the impact of any given disaster is prolonged among poorer households. Disaster-risk social protection programs have the potential to support governments in providing timely and focused assistance to affected vulnerable populations in this phase. By providing a safety net to affected individuals, social protection programs can prevent beneficiary households from depleting already-limited savings, cutting expenditure on essential items, and reducing investments in human capital in the face of disasters. Social protection and labor systems, policies, and programs help individuals and societies manage risk and volatility. 2 Significant knowledge gaps exist on “how to” integrate the right elements of disaster risk management, social protection programs, and disaster risk financing in order to build long-term resiliency among vulnerable populations. The social protection systems are usually reactive, rather than responsive. The design and implementation features of most social protection programs have not yet accounted for the new dimension of vulnerability associated with disasters, nor have they been integrated within a broader system of response that involves risk financing. There are commonly major challenges in institutional coordination across SP, DRM and DRFI agencies (within and across levels of government, and between state and non- state actors) which limit the effectiveness of disaster preparedness and response. It is for this reason that the World Bank brought together experts and practitioners to share good practice experiences and identify the roadblocks ahead. Section II summarizes the building blocks of a disaster-risk resilient social protection system; section III presents some instruments to lower vulnerability from disaster-risks; and section IV concludes and presents the way forward. II. Knowledge Shared: Building Blocks of a Disaster-Risk Resilient Social Protection System5 Social protection systems that build resilience and adaptive capacity are better positioned to reduce the vulnerability of poor populations. This involves linking social protection systems (SP) with climate change adaptation (CCA)/disaster risk management (DRM), and disaster risk financing and insurance (DRFI) systems. These three systems often fall under different institutions. CCA and DRM systems are usually under the auspices of the Ministry of Environment for the former, and the Ministry of the Interior, Disaster Management or Civil Protection for the latter, including a potential role for the military; DRFI under the Ministry of Finance; and SP systems under the Ministry of Social Welfare or equivalent. Apart from institutional fragmentation, the different agencies are usually staffed by people with diverse professional backgrounds, ranging from science-oriented experts in the environment area, to social workers and economists in SP, to engineers and logistics specialists in DRM, and to insurance and finance experts in DRFI. In such a setting, common vocabulary and outlook cannot be assured and the demands of coordination may be substantial. There are four main building blocks of a disaster-risk resilient social protection system. These are the following: (1) a comprehensive service delivery system in place ex ante; (2) a social protection program that has sufficient footprint, design flexibility, and contingency plans; (3) financing instrument and resources based on objectively verifiable information; and (4) institutional coordination and capacity infrastructure. The remainder of the section summarizes the good practice in integrating CCA, DRM and DRFI elements into each of these building blocks as presented by technical experts. This includes the lessons learned in integrating these building blocks, as presented by country speakers, and the issues and challenges faced by many countries in the Asia and Pacific regions in adopting the good practice and in applying the lessons from other countries, as presented in roundtable discussions and group work sessions. 3 A comprehensive service delivery system in place ex ante Develop a unified social protection registry. This system stores data about the population that, ideally, includes information about individuals and their households’ level of wealth as well as geospatial information about where they live. It ideally requires each person to be assigned a unique national ID, though this is often absent or SP information systems may not incorporate it. This registry could be linked with information from CCA and DRM systems (remote sensing, hazard and exposure modeling and mapping) to understand the hazard and exposure to disasters of the most vulnerable groups. The registry could then be used to identify potential beneficiaries should a disaster strike, set the parameters for scale-up, and define the contingent liability for financing. In Asia and Pacific region, some countries are more advanced than others in developing such a unified registry. Pakistan has a poverty database and National ID registry while the Philippines has a national household targeting database. In Cambodia and Nepal, an ID system is used for community-based targeting while a PMT model-based database is used to distribute food vouchers in Mongolia. But for other countries, at best, there is a community- identified list of vulnerable households (such as in Myanmar) and there are regular poverty surveys (such as in Vietnam), which can be potentially developed into a database. Collecting information from the population, managing the information system, and allocating funds for it are among the pressing challenges in setting up a national registry. A specific issue in the context of disaster preparedness and response is the scope of population captured in SP registries/databases. Targeting immediately post-disaster is likely to be quite distinct from regular targeting (often area-based rather than household-based), given the importance of covering all affected people and the challenges of thorough beneficiary assessment. To the extent that regular SP beneficiary identification involves collecting information on a wide population (i.e., beyond the ultimate beneficiary population, sometimes on a census basis), it may have information on groups who become target populations post-disaster. Thus, the cost/benefit calculation of what data collection approach to take in the SP system should be made with the disaster perspective in mind for countries which are frequently affected. Employ climate-smart targeting. This targeting strategy involves a combination of geographic, administrative, and community targeting and takes into account the location’s exposure to natural hazard (reinforcing the importance of an integrated registry). In Africa, Ethiopia’s Productive Safety Net Program (PSNP) and Kenya’s Hunger Safety Net Program (HSNP) are among the programs with a built-in mechanism that allows them to expand or contract depending on the occurrence of droughts. In Asia and Pacific, some countries have integrated DRM and CCA elements into their targeting systems (up to a certain extent), but many still struggle to put them together. In Sri Lanka, the government has put in place hazard maps, which can also provide information on the population, poverty profile, and food requirement in case of a disaster. In Nepal, Bangladesh, and Cambodia, a community-based targeting method is used for regular programs while geographic targeting is used for post- disaster programs with the aid of hazard maps. Moreover, some countries are still in the early stages of putting in place a national targeting system such as Mongolia, Fiji, Papua New Guinea, and Timor-Leste. Develop alternative mechanisms to validate identification and targeting criteria ex post. After a disaster, pre-identified beneficiaries may no longer have their ID or documentation to 4 prove their identity. In these cases, there may be a need to employ alternative mechanisms for selecting beneficiaries that are simple yet verifiable. These include using self-declaration with third-party validation (e.g., community leaders), relying other agencies’ standards, such as the World Food Program registration practices for internally displaced persons, or using more straightforward criteria such as place of residence, which is more related to the direct impact of the disaster rather than chronic poverty. Lessons from Pakistan’s Citizen’s Damage Compensation Program (CDCP), which was used to respond to the 2010 flooding, suggest that, after a disaster, a simpler targeting strategy that mixes geographic and socio-economic information may be employed to increase the accuracy in selecting beneficiaries, speed the targeting process, and lower the number of beneficiaries that go through the grievance redress system. The key is balancing the need for speed and ease of use with accuracy understanding that in an emergency scenario the quick provision of benefits is important for mitigating the impact of the disaster. Use pre-existing payment mechanisms to deliver social assistance where possible. In the case of cash transfers, these mechanisms refer to ATM cards, mobile banking, internet banking, postal service, and remittance centers that are already in place prior to disasters. During the flooding in Pakistan, beneficiaries were provided with bank accounts (debit cards) through private banks (with government agreements) and transfers were made electronically with computerized national ID card verification. This strategy allowed for an effective, efficient, and transparent cash transfer system to beneficiaries (97 percent of beneficiaries were able to withdraw cash within 48 hours).6 This experience suggests that having an existing platform ready to execute cash transfers could reduce humanitarian costs brought about by disasters and reduce leakage rates. Kenya’s HSNP is also putting in place a scalable system whereby all registered households are provided with bank accounts to have an already established platform in case a disaster strikes. For some countries, identifying the appropriate delivery mechanism is not always straightforward, even under normal times. Reaching beneficiaries is already challenging, such as in Mongolia, where large masses of land and the population’s high level of mobility make it difficult for the government to locate beneficiaries and deliver assistance. Moreover, for countries with relatively less developed technology, the technology-aided mechanisms are not the preferred means to deliver assistance, as most of the vulnerable populations reside in rural areas where banks and mobile networks are in short supply. In Timor-Leste, for example, an emergency program used banks to deliver cash assistance, but many of the beneficiaries did not have access to banks. One strategy to address this is for banks to open up satellite branches in markets and shopping centers to reach out to beneficiaries, as was the case in Kenya, or use agents such as shopkeepers, as what happens already in some cases with mobile banking and payment systems. Alternative payment system mechanisms should also be considered when standard operating systems are down. After a disaster, the delivery mechanisms under normal times may not work – ATMs, internet, and mobile banking may not function as planned. The affected families may be located in areas that become inaccessible due to damages in road infrastructure. In these cases, there is a need to pre-identify potential channels to deliver assistance. For example, cash grants may need to be hand-delivered to beneficiaries; in-kind assistance (rice, canned food, medicine) may need to be airdropped to remote areas using helicopters as experienced during the typhoon in the Philippines; and transfer payments may need to be 5 processed on-site using hand-held devices as was the case during the flooding in Pakistan. For these alternative mechanisms to work, prior arrangements with public and private service providers must be made (civil defense, private banks, rural banks, and private warehouses to store goods). Community development organizations and volunteer non-government organizations are particularly important partners for countries such as Fiji and Papua New Guinea. Put in place a communication strategy to improve delivery mechanisms post-disaster. Given the massive coordination required in post-disaster response and recovery, there is a need to put in place a mechanism that ensures that the affected population receives the information needed to make decisions and access program benefits and can provide feedback on the program. This include setting up satellite phones in affected areas so that displaced families can communicate with their relatives, or undertaking communications and awareness-building on disbursing the cash grants to some beneficiaries who do not know how to use ATMs. These innovative tools that establish clear communication lines after a disaster depend on the country context, that is, what is available and effective, such as announcements over the radio or at the local mosque; temple or church, or information dissemination through local revenue officers, teachers, or social media. A social protection program that has sufficient footprint, design flexibility, and contingency plans Pre-identify a “backbone” program designated as a core disaster response vehicle. This program should be capable of scaling up post-disaster and allowing for adjustments in some components of the program such as targeting criteria (e.g., residence rather than poverty status), benefit modality (e.g., cash transfer may shift to in-kind transfer due to non-functioning markets), and program co-responsibility (e.g., CCT program may waive conditions after a disaster). Having a pre-identified program is critical to a speedy response to a disaster. One of the recent examples of this is the Philippines’ CCT program, which was used as the main vehicle to deliver social assistance during the typhoon in November 2013. This experience suggests that it is more cost-effective and faster to use pre-existing programs and structures rather than establishing new ones after a disaster. In India, the design of the agriculture insurance program is linked with climate variability so that, in the event of adverse weather conditions, affected farmers can claim for damages to their crops. Identifying a backbone program to respond to a disaster, however, implies that there is a program in place that can be scaled up, which is not the case in many countries in Asia and Pacific. In Fiji, while there is a targeted cash transfer program, there is no formal mechanism to expand during a disaster. Affected individuals usually have to apply for assistance for the government to then validate their eligibility. Integrate disaster-risk reduction and climate change adaptation into the basic program design. In many cases, it may be possible and preferable to go beyond providing assistance to address the beneficiary’s shortfalls in income and consumption. The program could also include activities that build resilience and adaptive capacity of beneficiaries. For instance, cash transfer programs may use mother’s groups or other community organizations to provide messaging and training for disaster preparedness (e.g., Philippines CCT program); public works may focus on 6 building community assets and common facilities for safety that mitigate risks (e.g., Ethiopia); trainings may include skills needed for reconstruction; and programs targeting farmers may encourage the production of weather-resistant crops. In addition, insurance-based programs and community-development programs may include the incorporation of CCA and DRM elements into the program design and implementation (these instruments will be discussed in the next section). Involve communities in program implementation. Communities have an important role to play in the implementation of social protection programs—they can validate the criteria in the selection of beneficiaries, serve as communication links between the government and the beneficiaries, and mobilize community support for disaster preparedness and response. In most countries in Asia and Pacific, community structures remain intact (e.g., wantok system is still strong in Fiji), which can be tapped to address the risks of disasters and climate change. In implementing a social protection program, communities can be empowered by trainings to understand hazard monitoring and early warning systems, public communication, contingency planning, and emergency management systems. In Nepal, for example, the government is looking into institutionalizing DRM trainings to the communities. As will be discussed in the next section, community-driven development programs are also useful in lowering the risks associated with natural disasters and climate change. Ensure contingency plans are developed by communities who are familiar with predictable shocks. In order that a country facing predictable and recurrent shocks can respond effectively, a process of contingency planning should be undertaken. This ensures clarity about what will happen in the event of a shock, where a response may be implemented, what type of response will be implemented, who will be targeted, where resources will come from, the process of decision making, etc. A comprehensive contingency plan can save time, lives, and livelihoods in the event of a shock. In Ethiopia, the preparation of contingency plans was a pre-condition for the triggering of its contingency resources and risk financing mechanism and have proved invaluable in ensuring clear lines of accountability and authority in times of response. While the program design allows for flexibility in rules and procedures post disaster, clear rules should also be established for the recovery phase. Disasters happen in different types and scales that may require different responses. Benefit levels of transfer programs depend on the nature of response, the scale and magnitude of the disaster, and the capacity of governments to provide financing. Programs may need to be adjusted depending on the need of the affected population (e.g., benefit or cash transfers or public works). Kenya’s HSNP has an option to increase the benefit level of the same set of beneficiaries depending on the scale and magnitude of disasters while Ethiopia’s PSNP has the ability to scale up to provide benefits to additional beneficiaries in the face of a disaster. El Salvador’s public works program’s target area was changed from urban to rural after a hurricane hit the country in November 2009. Once the emergency has subsided, the program should also be able to regularize operations, or scale back down in the recovery phase. However, this seems to be the most challenging part of post- disaster response. In Papua New Guinea, for example, rules on which agency is in-charge of rehabilitation are somewhat unclear as volunteer non-government organizations also come to assist affected populations. 7 Financing instruments and resources based on objectively verifiable information Arrange financing schemes in advance so that funding is available for rapid recovery and response. Speedy response after a disaster is very critical to limit humanitarian impacts. Underlying this is the availability of resources to fund this activity. This involves a comprehensive strategy that primarily aims to secure access to post-disaster financing before the disaster strikes to ensure rapid and cost-effective liquidity to finance recovery efforts. Having clearly defined rules ex ante on what the government is going to do ex post can help the government address the challenges to volatility of budget due to shock events (i.e., rules-based approach to scale-up). Panama offers a good case of integrating disaster risk financing into the country’s financial plan. After the massive flooding that hit the country in December 2010, the Panama government realized that it needed to be proactive rather than reactive to disasters. Financing instruments such as the country’s reserves and contingent facilities such as the World Bank’s CAT DDO and Inter-American Development Bank’s parametric loan are available in case of emergency. Another interesting case is the Pacific, where five island states have pooled resources (and risk) to purchase catastrophe risk insurance. In contrast to this, in Mongolia, although the government has contingency funds, they are small in scale and are not sufficient to respond to large-scale disasters. In Nepal, the lack of national-level risk-financing schemes limits the variability of sources of funds because most of the risk planning happens at the community level. Include risk-financing mechanisms in program design. An example of a program that has a built-in risk-financing mechanism is Ethiopia’s PSNP. Built into this program is a continuum of response that allows the program to expand in stages. For low-level and transitory shocks, a contingency budget equivalent to 20 percent of the current program budget is available every year at the local and regional level, which provides temporary, additional resources to affected individuals. If a shock is large enough, as detected by an early warning system, it triggers a risk- financing mechanism response, which leads to the provision of contingent financing through emergency grants held at the federal level, and if those are insufficient to meet the need of the program, resorts to humanitarian assistance. The scale-up feature of Kenya’s HSNP is linked with a risk-financing scheme through the African Risk Capacity (ARC), whereby in the event of a drought, the government has the following options: increase the transfer value while maintaining the same number of beneficiaries; increase the number of beneficiaries while maintaining the same transfer value; or both. Build knowledge on the country’s exposure to risk to support evidence-based financial planning. Information is key to be able to plan ahead. Underestimating the risk could mean huge financial costs to governments while overestimating it could create opportunity costs as funds are locked-in for disaster risk financing. A disaster-risk analysis should be considered in forming the country’s comprehensive financial protection strategy as in the case of Panama, which should be in line with proposed public investments that aim to build resilience among communities, including social protection. In most cases, however, line ministries respond only to their respective mandates and little coordination and communication happen across ministries. Countries that are relatively more exposed to natural disasters tend to have DRM systems that are relatively more advanced compared to their social protection systems. Nepal, Mongolia, Cambodia, and Myanmar, for example, have established hazard monitoring systems 8 that can help their governments evaluate the magnitude of impacts, but these systems are not fully linked with their social protection systems and disaster risk financing. Institutional coordination and capacity Create formal partnership agreements ex ante among national agencies, public and private service providers, and development partners. The need for coordination is essential pre- and post-disaster. Given the large number of players, there should be clear rules on who does what. The “appropriate” institutional set-up seems to vary across countries. In Ethiopia, experience with the PSNP suggests that coordinating safety net support through a single central agency, which works closely with other national agencies and sub-national authorities, is more effective than putting multiple agencies in charge. Whereas, in the Philippines, disaster response is handled by an inter-agency body composed of the ministries of civil defense, social welfare and development, science and technology, and the socio-economic planning agency. This institutional arrangement involving multiple agencies seemed to have worked well in response and recovery stages post disaster. Regardless, while institutional set-ups may vary, having a clear delineation of tasks and functions laid out before the disaster is important, especially clarity on which agency has overall leadership in the disaster response and recovery phases, and on the interface between national agencies and sub-national authorities in high risk areas. The lack of appreciation in linking the systems together seems to prevent the ministries from coordinating with each other. Beyond government, similar coordination challenges exist between state and non-state actors, especially in lower capacity settings. A simple platform for data-sharing across line ministries can lay out the foundation for coordination. As pointed out earlier, information on households and their exposure to disasters and climate change is critical in building a disaster-risk resilient social protection system and risk-financing schemes. This means that the information related to poverty, DRM, and CCA should be consolidated (also related to building an integrated registry as mentioned previously). As mentioned, some countries in the region have established DRM systems such as Nepal, Mongolia, Cambodia, and Myanmar, but the challenge is to focus on how to better integrate them with the social protection system. To achieve this, what seems to have worked, at least for a start, is to use a simple platform that allows multiple agencies to access the data. For example, Sri Lanka has a platform that can host risk information in map format (hazard map) that are available online. The information therein is updated on a real-time basis. Mongolia is also putting in place an accessible “geo portal” that aims to provide information on natural hazards. Build capacity of the relevant institutions to respond, even if they are outside of a ‘typical’ social protection intervention. The success of a scalable response to a shock will stand or fall on the coordination of agencies and the capacity of agencies to respond. Building capacities of agencies in the midst of a response is not possible, although often left until then. Similarly, the role of coordination is often left to insufficiently capacitated and/or politically weak institutions. Coordination cannot be seen as an additional function of an agency when there is a need—it requires a particular set of skills and dedicated resources. Investing in capacity and coordination structures as part of the steady-state operations of a social protection scheme will pay dividends in times of shock. 9 III. Lowering Vulnerability from Disasters and Climate Change: Some Instruments This section presents some of the social protection programs that are commonly used around the world with features that can be modified to integrate DRM and CCA elements into their core design and implementation components. Specifically, this section highlights the roles of these programs as well as the knowledge gaps in implementing them pre and post disaster. Benefit Transfers Benefit transfers can be used as an instrument to assist beneficiaries in meeting their immediate consumption needs. They can take on many forms such as cash, in-kind, near-cash (vouchers); can either be conditional or unconditional; temporary or regular in duration; and can either be targeted according to a criteria or universal. Before a disaster, benefit transfers can build incomes and human capital as a means of building assets that prevent or mitigate the effect of shocks and climate change, and provide insurance to pre-finance potential disaster impacts. After a disaster, benefit transfers can also protect livelihoods in the early stages of a protracted crisis, provide life-saving support in acute crises, contribute to household’s ability to recover from shocks and provide temporary employment in recovery phase. Moreover, through insurance, benefit transfers allow consumption smoothing and reinvestment in productive activities. Given the numerous and often unpredictable factors that arise post-disaster, there is a need to focus on strengthening implementation structures in advance for a range of post- disaster situations. Administering benefit transfer in the aftermath of the disaster can be overwhelming for the implementing agency. For instance, what type of benefit (cash, in-kind, and near-cash) at a certain stage in response is preferred; and what transition and exit strategies are appropriate, from emergency phase through recovery and back to steady-state. Experience from the Philippines indicates that food assistance was important in the early stages of response to address hunger and meet basic needs immediately after a major typhoon, but it was too costly to administer as it required massive manpower for delivery, entailed large storage costs, and was subject to leakages. Soon after the typhoon, the assistance shifted to cash coupons and then to public works. But, so far, there are no clear rules as to the appropriate time to return to normal program operations, nor are there rules within the government as to which agency is in-charge of rehabilitation. Public Works Public works can be used to build infrastructure that supports disaster-preparedness and climate change adaptation, as well as being an effective safety net post-disaster. They address shortfalls in households’ income or consumption by giving transfers in exchange for beneficiaries, creating, maintaining, or rehabilitating assets and infrastructure. Before a disaster, they can be used to build infrastructure that protects the community from adverse impacts (levies, community shelters, polders). After a disaster, they can be used for post-disaster clean- up, to re-build damaged infrastructure (roads, bridges, school buildings) or build new assets that protect the community from future shocks. Public works program with these objectives are undertaken in several countries in the region, which include the irrigation projects conducted in 10 Northern Ethiopia to manage water systems and increase agricultural productivity; community shelters erected in flood-prone areas in Bangladesh; and community projects undertaken by unemployed youth in El Salvador and public works in Mexico. Flexible design features can enhance a program’s effectiveness as a post-disaster response. Bangladesh, for example, has a history of innovation in the design of its safety net programs, particularly public works, which have emerged and evolved in response to disasters. The Employment Generation Program for the Poorest (EGPP), a cash-based workfare program, was implemented by the government as a response to the global food crisis in 2008 and has expanded into a regular program since then. While initially designed as a disaster response mechanism, the program primarily aims to provide short-term employment to the most vulnerable population in rural areas, identified using geographic and household targeting based on poverty statistics. Subprojects implemented under the program are largely focused on small- scale repair and maintenance of community infrastructure, although in years when disasters occur, the emphasis has been on clean up and rehabilitation. Recent innovations in the program include piloting electronic payments through mobile phones and cash cards, and introducing an element of forced savings, both of which serve to ensure that in times of disaster, the most vulnerable would have the means to cope. The government continues to look into ways of increasing flexibility and the scope for scaling up in order for the program to better respond to disasters. There is a need to combine abundant but unskilled labor with some skilled inputs in order to build resilient infrastructure. Because the beneficiaries of public works programs are frequently unskilled, or low-skilled, concerns have been raised regarding the quality of public works outputs, especially because often the infrastructure that they build are meant to protect communities from the adverse impacts of disasters. The Government of Bangladesh has tried to address this by engaging a private firm to undertake random checks on the quality of infrastructure of its public works program. Related to the issue of skills, the Government of Indonesia, through the provincial disaster management agencies, has ensured that the beneficiaries of the Rekompak program, a community-based post-disaster program, are provided training on anti-seismic construction methods so that the houses that the beneficiaries build are earthquake-resistant. Moreover, the institutions in charge of disaster risk reduction and preparedness have supported the Rekompak program throughout its implementation. Insurance-Based Programs Insurance-based programs can help protect beneficiaries and governments from financial risks arising from the loss of productive assets. They especially work for countries with a high proportion of poor households that rely on income from their productive assets. For instance, agriculture insurance could protect farmers, herders, and fishermen from loss. One of the first countries to integrate insurance schemes in social safety nets was Mexico, where the government established the CADENA catastrophe insurance program that is targeted at small- scale vulnerable farmers with no access to credit or insurance. In India, the government adopted risk financing and insurance principles to transition its National Crop Insurance Program from a social crop insurance scheme to a market-based crop insurance program. Insurance-based programs seem to be effective in protecting vulnerable groups from adverse shocks due to 11 climate change, but knowledge gaps exist on how they can be applied to other countries. Nepal, for instance, is looking into how crop insurance can be adopted to the country. Insurance can help in post-disaster recovery (and more generally for slow-onset disasters such as droughts) as the affected populations do not only receive emergency relief goods but also construction materials to help them rebuild their property. An important design consideration in such programs is having a manageable mechanism for verifying and valuing insurable losses. A number of crop and weather insurance programs have simplified this process in recent years through reference to weather indices (e.g., rainfall under a specified level in an area automatically triggers payouts without the need to verify individual agricultural losses). Insurance-based programs have increasingly been used in recent years, but less so in developing countries. Developed countries tend to have insurance so that risks are passed on to private insurance companies. Developing countries have very low insurance penetration, presumably because regulations can be quite weak and there is scant knowledge on what insurances do. There is also an issue of mindset, as people tend to not be attuned to having an insurance, making it difficult to communicate the value of insurance to potential beneficiaries. In the Philippines, it is estimated that only 0.9 to 1.1 percent of the population are covered by general insurance, which is low compared to the average rates of ASEAN countries at 3.9 percent. However, after the typhoon in November 2013, more people have opted to insure their property. This raises the issue of default options and linkages to other programs. Some safety net programs automatically include some form of insurance, as NREGA in India has done in some states by automatic payment of life insurance premia. Community-Driven Development (CDD) Programs CDD programs are a good tool to build capacity among communities and to take the lead in recovery. Indonesia’s Rekompak program offers a good case of an innovative community- based model of a post-disaster program that allows the community to rebuild their homes and community infrastructure and manage resources for reconstruction. This model of community participation has enhances community ownership in the reconstruction process and promotes accountability among the beneficiaries in disaster awareness and mitigation. Experience from the Philippines also emphasizes the importance of mobilizing the communities in post-disaster rehabilitation. Communities were engaged in the bottom-up approach to crafting the local rehabilitation plan, drawing on the communities’ existing capabilities to undertake CDD programs. Recently, CDD projects focusing on climate change adaptation and reducing disaster- risks have been facilitating provision of shelter assistance and relocating families from the no- build zones. IV. Way Forward The above discussions highlight the fact that while there are lessons and good practice experience that countries can learn from, there are also many questions left unanswered. For instance, on targeting, there remains the question on how desirable household-level targeting is and whether or not other considerations trump the usual approaches. Program designs can also differ in the mixture of cash, kind, and near-cash transfer in different disaster 12 response situations and different stages in response. On delivery, there is still little understanding on how best to integrate operations of DR, social welfare, and non-state actors in delivery of transfers post-disaster. Knowledge gaps also exist on the appropriate strategies to transition from emergency phase of SP response through recovery and back to steady state. Moreover, countries struggle to identify the appropriate financing vehicles to facilitate timely post-disaster response for SP programs. They also face different realities with regard to barriers for specific population (e.g., urban/rural), appropriate forms of communication, implementation partnerships, and channel for communication among agencies. To frame the agenda ahead, a country needs to identify at which stage it is in developing a risk resilient social protection system. Ideally, the three systems – DRM/CCA, DRFI and SP – should be strong, both individually as well as collectively. Countries vary in stages of building each of these systems and in the process of putting them all together. For some countries, these systems are strong and well integrated. But for many countries, there are further challenges ahead—the three systems may co-exist but remain independent from each other, one of the systems may be stronger than the others, or some of the systems may not yet be established. In terms of forming a strategy, a country should evaluate where it is at in building these systems and determine the level of coordination required to integrate them all together. The implementation detail involved in the coordination part can be massive, but one can start from a simple point of sharing data and building a common registry, which is the core of a well- performing social protection system. Political will is critical to push the agenda forward. The conference presented good practice on “how to” tackle the issues of disasters and climate variability and showed that there is a wealth of information that countries can learn from. However, finding the solution that is appropriate to a specific country largely depends on the extent to which its government owns the agenda and on having the political will to make things happen. Building each of the systems is the relatively easier part, but the more difficult part is coordinating the ministries to integrate these systems together. Working toward this goal goes beyond theories or building strategies and implementation plans as the real hard work lies in making them happen in practice. For this, technical experts should advocate and make a case for why political will is important. Experience from countries that are able to build such systems highlights the reality that it takes “champions” within the government who are willing to push the agenda forward. The process involves long-term planning, which underscores the need for coordination. Planning for the disaster that may not actually happen is important and can bring discipline into the system. Although planning does not seem to be technically difficult, it is politically challenging. In particular, it is important to get the ministry of finance to own the agenda. Long- term planning involves having a set of parameters to work around with, thinking through available resources, and partnering with external agencies to address the problem. This process involves laying a common ground to understand the issue, wherein one can start with the available tools such as household surveys and hazard maps. Financing the risk involves thinking through various resources that can be available depending on the type and scale of disasters, but information is key in this process. Moreover, while existing budgetary resources can finance small and localized disaster response, there is a need to harness external resources to address the problem. In this regard, it is important to establish partnerships with humanitarian agencies. 13 The World Bank is committed to assisting countries in setting up the technical and coordination mechanisms. The issues that were raised during the conference are global problems that require global solutions. Ultimately, building the three systems and integrating them altogether are about protecting the most vulnerable and improving the lives of the least well-off. As each country struggles to find the right solutions, the role of the World Bank is to partner with them in this process and provide venues to facilitate the sharing of good practice so that countries can learn from each other. While some countries are more aware and better positioned to respond to disasters and climate change, the situation of other countries is more complex. Hence, there is a need for countries to collaborate in dealing with common risks and in reaching for the common goal to help the poorest. Faced with competing development objectives, the task ahead is indeed overwhelming, but this conference highlighted the fact that solutions do exist and can be further developed, refined, and improved. 14 Endnotes                                                                                                                 1 One of the events held by the World Bank related to this topic was the international workshop on “Making Social Protection Work for Pro-poor Disaster Risk Reduction and Climate Change Adaptation” in Addis Ababa (Ethiopia) on March 14-17, 2011. 2 Country delegations are the following: Cambodia, Mongolia, Vietnam, Myanmar, Timor-Leste, Papua New Guinea, Fiji, Sri Lanka, Bangladesh, India, Nepal, Indonesia, Philippines, Ethiopia, Kenya, El Salvador, and Panama. 3 In this document, Asia and Pacific refer to countries in “East Asia and Pacific” and “South Asia” regions as defined by the World Bank. East Asia and Pacific region include the following countries: American Samoa, Cambodia, China, Fiji, Indonesia, Kiribati, Korea Dem. Rep., Lao PDR, Malaysia , Marshall Islands, Micronesia, Mongolia, Myanmar, Palau, Papua New Guinea, Philippines, Samoa, Solomon Islands, Thailand, Timor-Leste, Tonga, Tuvalu, Vanuatu, and Vietnam. South Asia region include the following countries: Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka. 4 Data taken from World Bank and UNDP (2010), Natural Hazards, UnNatural Disasters, Washington, DC. 5 From this point on, these information were gathered from presentations of key speakers and from notes taken during the open forum, round table discussions, and group work sessions. 6 Some withdrew cash from ATM’s or Bank’s directly, others went to centers set-up specifically for the program which had Point of Sale (PoS) machines. 15