SEPTEMBER 20 22 Towards Adaptive Social Protection in Vanuatu TECHNICAL REPORT Disclaimer This work is a product of the staff of The World Bank and the Global Facility for Disaster Reduction and Recovery (GFDRR) with external contributions. The findings, analysis and conclusions expressed in this document do not necessarily reflect the views of any individual partner organization of The World Bank, its Board of Directors, or the governments they represent. Although the World Bank and GFDRR make reasonable efforts to ensure all the information presented in this document is correct, its accuracy and integrity cannot be guaranteed. Use of any data or information from this document is at the user’s own risk and under no circumstances shall the World Bank, GFDRR or any of its partners be liable for any loss, damage, liability or expense incurred or suffered which is claimed to result from reliance on the data contained in this document. The boundaries, colors, denomination, and other information shown in 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. Copyright Statement The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other 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. Photography: ©  World Bank Please cite this work as follows: Stein, Wiebke, Kenia Parsons, and Kirstie Petrou. (2022). Towards Adaptive Social Protection in Vanuatu: Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19. Washington, D.C.: World Bank Towards Adaptive Social Protection in Vanuatu Acknowledgements This report was written by Wiebke Stein, Social support during the survey design and sampling, Protection and Jobs Consultant, Kenia Parsons, as well as the respondents who graciously Social Protection Specialist, and Kirstie Petrou, participated in this survey, shedding light on how Social Protection and Jobs Consultant, based on cash transfers have impacted their livelihoods in survey instruments and sampling designed by Vanuatu, post Tropical Cyclone (TC) Harold and Rita Fernandes, Social Protection Consultant, throughout COVID-19. We are also grateful for under the coordination of Yasser El-Gammal, the Government of Vanuatu ministries and line Practice Manager. The team would like to agencies for the inputs and comments provided acknowledge the contributions provided by during the preparation of this report. This report Alejandro González Arreola, Dung Doan, Margaret has been made possible by the generous support Daniel, Matthew Dornan, Jesse Doyle, Sandor of the Global Facility for Disaster Reduction and Karacsony, Yasuhiro Kawasoe, Annette Leith, Tevi Recovery (GFDRR), the Australian Government Obed, and Leisande Otto. Department for Foreign Affairs and Trade (DFAT), and USAID (US Agency for International We would like to thank Tebbutt Research, who Development). The report was copy edited led the field work, data collection, and data by Angela Takats and graphically designed by cleaning. The team would like to extend our Félix Genêt Laframboise. sincere gratitude to Oxfam, especially to Sandra Hart and Vivian Fischer who have provided ii Towards Adaptive Social Protection in Vanuatu Table of Contents List of Tables.........................................................................................................................................................................................v List of Figures.......................................................................................................................................................................................v Abbreviations.......................................................................................................................................................................................vi Executive Summary..........................................................................................................................................................................2 1. Introduction.................................................................................................................................................................................. 6 2. The Importance of Cash Transfers and Adaptive Social Protection during Natural Disasters and Economic Shocks........................................................................................................................... 10 3. The Cash Transfer Program in Sanma: Context, Methods, and Sampling...............................................................13 3.1 Program Description............................................................................................................................................................13 3.2 Study Design and Sampling.............................................................................................................................................. 14 3.3 Characteristics of Surveyed Households.......................................................................................................................15 4. Role of the CTP in Facilitating Household Recovery..................................................................................................... 18 4.1 Impacts of TC Harold on Housing and Agricultural Activities................................................................................. 18 4.2 The Role of the CTP in Repairing Homes and Restoring Agricultural Assets..................................................... 19 4.3 Impacts of the Dual Shocks on Jobs, Food Security, and Health.......................................................................... 22 4.4 The Role of CTP in Deterring Negative Coping Strategies and Supporting Human Capital......................... 25 5. Informal and Formal Social Safety Nets.............................................................................................................................28 6. Perceptions on the Implementation of the CTP in Sanma..........................................................................................30 7. Key Findings................................................................................................................................................................................. 32 8. Recommendations....................................................................................................................................................................34 References........................................................................................................................................................................................36 Appendix A: Additional Figures and Tables...............................................................................................................................40 Appendix B: Additional Descriptions.......................................................................................................................................... 47 Appendix C: Examples of Social Protection Responses Following COVID-19 in the Pacific.................................... 52 iv Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 List of Tables Table A. 1: Vendor to beneficiary ratio in Sanma............................................................................................................... 41 Table A. 2: Household composition by vulnerable group............................................................................................... 41 Table B. 1: Total respondents by age groups in percent..................................................................................................48 Table B. 2: Total respondents by sex in percent.................................................................................................................48 Table B. 3: Unweighted sample distribution by vulnerable group................................................................................48 Table B. 4: Comparison sample attrition..............................................................................................................................49 Table B. 5: Regression results attrition..................................................................................................................................50 List of Figures Figure 1: Country map..............................................................................................................................................................7 Figure 2: Social protection programs: Vertical and horizontal expansion...............................................................12 Figure 3: Main income sources of surveyed households............................................................................................. 16 Figure 4: Status of housing reconstruction by vulnerable group............................................................................... 19 Figure 5: Use of cash transfers by vulnerable group following TC Harold..............................................................20 Figure 6: Percentage of job losses experienced by vulnerable groups due to dual shocks............................... 22 Figure 7: Coping strategies of households between March–October 2020.........................................................24 Figure 8: Percentage of households that ran out of food before and after the CTP...........................................26 Figure 9: CT usage with respect to the impacts of the dual shocks on people’s livelihood..............................26 Figure 10: Overview of non-CTP assistance in the immediate aftermath of the shocks......................................29 Figure 11: Overview of information sources......................................................................................................................30 Figure 12: Mode of transport between home and local stores.....................................................................................31 Figure A. 1: Timeline of events..................................................................................................................................................40 Figure A. 2: Households by area council...............................................................................................................................42 Figure A. 3: Main income sources by vulnerable group....................................................................................................42 Figure A. 4: Income derived from agriculture by vulnerable group...............................................................................43 Figure A. 5: Average monthly household income at baseline by vulnerable group..................................................43 Figure A. 6: Breakdown of status of cyclone induced damage to agricultural land six months after TC Harold (prior to CT) by vulnerable group..................................................................44 Figure A. 7: Percentage of households that lost access to the ocean or river because of TC Harold by vulnerable group.....................................................................................................................44 Figure A. 8: Breakdown of CT usage for repairs to dwelling............................................................................................45 Figure A. 9: Breakdown of average travel time from home to the registration site...................................................45 Figure A. 10: Mode of transport between home and registration point.........................................................................46 v Towards Adaptive Social Protection in Vanuatu Abbreviations ADB Asian Development Bank ASP Adaptive Social Protection CaLP The Cash Learning Partnership CPS Child Protection Scheme CT Cash Transfer CTP Cash Transfer Program DAS Disability Allowance Scheme DFAT Department of Foreign Affairs and Trade – Australian Government DRM Disaster Risk Management FAO Food and Agriculture Organization ILO International Labour Organization NGO Non-Governmental Organization OCHA United Nations Office for the Coordination of Humanitarian Affairs OPM Oxford Policy Management PBS Poverty Benefit Scheme RBV Reserve Bank of Vanuatu SAT Samoan Tala TC Tropical Cyclone VNPF Vanuatu National Provident Fund VNSO Vanuatu National Statistics Office VT Vatu vi Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 11 Towards Adaptive Social Protection in Vanuatu Executive Summary Due to its geographical location on the Pacific Ring of Fire, Vanuatu ranks as the country with the highest disaster risk in the world.1 Enhancing community resilience and adaptive capacity to climate change and natural disasters is a key goal of the Government’s Vanuatu 2030: The People’s Plan. The ni-Vanuatu culture is deeply rooted in traditional values and shared prosperity, and, as such, informal social protection mechanisms – such as support from family or the church – are important for local communities. However, natural disasters put pressure on both informal social protection and formal, government-led social protection systems. Adaptive Social Protection (ASP)2 systems can enhance the capacity of communities to prepare for, respond to, and cope with shocks, reducing and alleviating the negative impacts on their wellbeing. To this end, other social registry of those most in need, humanitarian countries in the Pacific region (such as Fiji and assistance was fundamental in providing support Tonga) are increasingly investing in ASP. This report and relief to affected households. aims to highlight how cash transfers are a social As part of this humanitarian assistance, a protection tool that can enhance the capacity of multi-purpose Cash Transfer Program (CTP) was local communities in Vanuatu to cope with large established to assist 3,745  vulnerable families and wide-spread shocks. in Sanma, Shefa, and Tafea, disbursing VT In April 2020, just two weeks after COVID-19 262,150,000 (approx. US$2,364,907). The CTP caused the closure of its borders, Vanuatu was implemented by local and international was hit by Tropical Cyclone (TC) Harold.3 This Non-Governmental Organizations (NGOs), and Category 5  cyclone affected 159,000  people, led by Oxfam Vanuatu between October 2020 and damaged 17,000  homes, and left 87,000  people May 2021. Households spent the transfer primarily without shelter.4 The province of Sanma was the on food and water (62  percent), sanitation and worst affected, with 53,344  people impacted. In hygiene products (13  percent), and hardware the absence of a social protection program with a materials (11 percent). Behlert et al. (2020). World Risk Report. 1. 2. Adaptive Social Protection “…helps to build the resilience of poor and vulnerable households to the impacts of large, covariate shocks, such as natural disasters, economic crises, pandemics, conflict, and forced displacement. Through the provision of transfers and services directly to these households, ASP supports their capacity to prepare for, cope with, and adapt to the shocks they face – before, during, and after these shocks occur. Over the long term, by supporting these three capacities, ASP can provide a pathway to a more resilient state for households that may otherwise lack the resources to move out of chronically vulnerable situations”. (Bowen et al., 2020) 3. Due to strict international travel restrictions, tourism numbers dropped to zero, which affected thousands of formal and informal jobs in the tourism sector and associated industries. Vanuatu’s economy was severely affected and the country’s GDP contracted by 6.8 percent in 2020 (World Bank, 2022). 4. DFAT. (2020). Tropical Cyclone Harold https://www.dfat.gov.au/crisis-hub/Pages/tropical-cyclone-harold 2 The CTP became an important source of livelihood due to the impacts of the dual shocks of TC Harold and COVID-19 on vulnerable households5 in Vanuatu. The CTP supported vulnerable households with a transfer of VT 70,000  (approx. US$631)6 in six monthly installments. These cash transfers were provided to program beneficiaries in the form of an e-voucher card, which was used to buy products at local registered stores or vendors. In total, 2,530  beneficiary households and 204  vendors were registered in Sanma. household and business assets. Around 30 percent of surveyed households included at least one This study demonstrates that vulnerable house- household member who lost their job as a holds in the province of Sanma were highly direct result of the crises. The majority of these affected by the dual shocks of TC Harold and households ran out of food in the months after TC COVID-19. A ‘before and after comparison’ of a Harold. Households received informal assistance subsample of 194  households showed that the through friends, family, or the community, mainly majority of surveyed households experienced in the form of food, water, or clothing; this informal cyclone-related damage to their dwelling and social protection provided invaluable support in agricultural land. TC Harold also destroyed the immediate aftermath of the shocks. 5. Households were defined as vulnerable if they included at least one vulnerable member, i.e., the beneficiary. Vulnerability was defined according to one of five criteria: elderly (above 60 years), people living with a disability, single mothers, widows/widowers, and households displaced by TC Harold. The program identified one beneficiary per household. The targeting was designed with the intention to ensure gender equality among the group of program beneficiaries. 6. All monetary values refer to the following exchange rate: 1 US Dollar = 110.85 Vatu (https://xe.com/, accessed on 5th November 2021). 3 Towards Adaptive Social Protection in Vanuatu Many households adopted negative coping food consumption (68  percent), delayed paying strategies in the immediate aftermath of the bills (27 percent), or removed children from school shocks. In order to manage the prolonged (9 percent). These negative coping strategies likely economic impacts, surveyed households reduced increased the risk of falling into hardship. Key Findings: The CTP complemented existing informal social protection. Around 75 percent of The CTP was an effective tool to help surveyed households shared their purchased vulnerable households accelerate their goods with others, mainly with relatives but recovery from both the physical and also with friends or religious institutions. economic impacts of the dual shocks. Around 57 percent of surveyed households used the CTP to repair their dwelling, for The CTP not only increased access to example to repair roofs, floors, and cooking medical care but also increased access areas. Households also used the money to varied food and thus supported food to regenerate or buy agricultural land security and health among surveyed (27 percent), regain access the ocean or households, which in turn, contributes to river for the purpose of fishing (8 percent), preserving human capital gains. Households or restore livelihoods (35 percent). switched to purchasing more varied food items and fewer households ran out of food during the period of the CTP as compared to The report indicates that households the six months before the program. Around found the program accommodated their 47 percent of surveyed households used needs. The majority of surveyed households the CTP to access medical treatment. (96 percent) reported positive experiences of the program. Complaints related to the transfer not being enough to cover While experiences of the CTP were household needs or to problems with generally positive, the report also the e-voucher payment card. highlights challenges faced by households and potential areas for improvement. For example, some 45 percent of beneficiaries remained food insecure at the end of the CTP and roughly 34 percent of recipients experienced difficulties in getting the goods they needed at local stores. 4 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 This pilot demonstrates there is a potential role Recovery Strategy for 2020–2023 following for social protection in the immediate recovery COVID-19 and TC Harold. after shocks and highlights the importance of ongoing social protection programs. The Based on the findings from the pilot and the report findings could be used to inform the Government of Vanuatu’s high level policy implementation of the National Sustainable documents mentioned above, the report puts Development Plan: Vanuatu 2030 (SOC1, SOC4, forward the following recommendations for SOC6, ENV1, ENV3, ENV4, ECO3, ECO4) and the Government, development partners, and the medium-term goals set out in the Vanuatu civil society organizations: Key Recommendations: Establish an ongoing formal social protection system. This would complement Focus on ‘building back better’ in order to existing informal social safety nets, in order strengthen the resilience of households. to ensure that vulnerable households have This should be done by supporting continued access to essential services. households’ capacity to adapt to future It would also enhance the preparedness, shocks through ASP and Disaster Risk response, and coping capacities of vulnerable Management (DRM) systems. households and communities given the high exposure of the country to natural hazards. Implement pre-disaster ASP programs and interventions with resilience-building impacts Establish an ASP strategy. This would result on both basic community infrastructure and in more strategic, integrated, and streamlined the wellbeing of households. This could include, approach to disaster recovery, contributing to for instance, seasonal public works programs a more responsive social protection system oriented to repair or build basic community and a quicker and more inclusive recovery infrastructure before the rainy season which process in particular for the poor and would help to reduce flooding risks while most vulnerable people. providing income to poor and vulnerable households. 5 Towards Adaptive Social Protection in Vanuatu 1. Introduction Vanuatu is a small island state situated in the Pacific with roughly 83  islands. Agriculture is a critical source of income for the three-quarters of the population that live in rural areas and for a good share of the urban population. Around 63 of the nation’s islands are inhabited. With a population of 299,882  inhabitants (as of 2019), Vanuatu is the fourth largest Pacific Island Country in terms of population. The country is divided into six provinces and most ni-Vanuatu live in the three provinces of Malampa, Sanma, and Shefa. The three largest islands are Malekula, Santo, and Efate (Figure 1) (World Bank, 2011). The island of Efate is home to the country’s capital, Port Vila. Luganville, which is located on the island of Santo, is Vanuatu’s other urban center. However, the majority of the population (around 75  percent) live in rural areas, which explains the country’s reliance on agriculture; around 80  percent of Vanuatu’s population work in the agricultural sector. Employment in the farming sector is pre­ do­­ minantly informal and only around 20  percent of Vanuatu’s labor force is engaged in formal jobs (Fischer, 2020; UNDP, 2018). In Vanuatu, 15.9  percent of the population live in hardship and around 96.7  percent of those experiencing hardship live in Vanuatu’s rural areas (VNSO, 2021b; World Bank, 2021a).7 events. On average, Vanuatu has experienced Due to its geographical location, Vanuatu faces 2.6 cyclones annually over the past 40 years. This exposure to natural disasters and ranks as the is more than any other Pacific country (World country with the highest disaster risk worldwide Bank, 2021b). The agricultural sector, which most (Behlert et al., 2020). The country is located on ni-Vanuatu rely on for food and income, is highly the Pacific Ring of Fire, where two tectonic plates susceptible to the effects of natural disasters, meet and this exposes it to frequent earthquakes which can destroy agricultural land, crops, and and volcanic eruptions (OCHA, 2021). This natural assets (World Bank, 2021c). Thus, the increasing exposure is further amplified by the consequences effects of natural disasters will likely pose a of global warming, with rising sea levels and an challenge to Vanuatu’s population, especially increase in the frequency of extreme weather those already facing hardship. Hardship is defined as the rate of people living below the ‘National Poverty Line’ and was calculated with data from the 2019–2020 Vanuatu National 7. Sustainable Development Plan (NSDP) Baseline Survey (VNSO 2021b). 6 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Figure 1: Country map Source: Nations Online Project (2021) 7 Towards Adaptive Social Protection in Vanuatu In 2020, Vanuatu suffered substantial economic provide only limited support when communities losses due to the dual shocks of the global are faced with aggregate shocks or in communities COVID-19 pandemic and a devastating Category with already limited resources (World Bank, 2013). 5  Tropical Cyclone (TC) in April 2020. The dual Vanuatu’s main formal social protection shocks resulted in a significant reduction in mechanism is the Vanuatu National Provident income, consumption, and livelihoods across the Fund (VNPF), a retirement savings fund for the country, especially among the most vulnerable formally employed. In the aftermath of TC Pam in populations in rural areas (Government of Vanuatu, 2015, VNPF members were able to withdraw up to 2020). Consequently, it was projected that the 20  percent of their total contributions. Other proportion of Vanuatu’s population living below formal social protection measures have included the poverty line of US$1.90 per day would increase school fee waivers and the Employment from approximately 13.9 percent to 18.3 percent. Stabilisation Payment (ESP), a temporary payment Consumption was projected to decline by for the formally employed introduced during the 10 percent as a result of COVID-19 (ADB, 2022). As COVID-19 pandemic. However, the country does a result of the strict travel restrictions, which were not have an established social protection system implemented to avoid a COVID-19 outbreak in that targets those facing hardship. Vanuatu Vanuatu, international tourism numbers dropped therefore relies mostly on humanitarian assistance to zero. Hence, businesses drastically reduced in the event of large, aggregated shocks, such as their staff and unemployment increased (Arahan et natural disasters. This can make it difficult for the al., 2020). The number of full-time employees in Government to quickly respond to shocks as the tourism industry reduced by 70 percent during humanitarian assistance can sometimes be the crisis (Vanuatu Tourism Office, 2020). Overall, delayed. For example, in the aftermath of TC around 2,000 formal jobs and tens of thousands of Harold, reconstruction projects were delayed and jobs in the informal sector (such as bus and taxi international aid responses were severely restricted drivers, handicraft artisans, and market vendors) because of the ongoing pandemic. This further were affected. In addition, the country was hit by a impacted Vanuatu’s economy and the wellbeing severe tropical cyclone in early April 2020. of its population (Australian Red Cross, 2020; GoV, TC Harold affected 159,000 people in Vanuatu and 2020, RBV, 2020; DFAT, 2021). left three dead. More than 17,000  homes were damaged or destroyed, leaving 87,000  people Formal and Adaptive Social Protection (ASP) without shelter. TC Harold damaged schools and programs, such as cash transfer programs, can health facilities, and led to an increase in be extremely beneficial in the aftermath of a communicable diseases, such as conjunctivitis shock and this report aims to shed light on the and diarrhea. The provinces of Sanma, Penama, potential opportunities of such measures for and Malampa were the worst affected. The Vanuatu. The next chapter of this report agricultural sector in these areas suffered severely (Chapter  2) provides information about ASP as TC Harold damaged crops and agricultural systems. Furthermore, this report analyzesthe produce, putting food security at risk. In the consequences of TC Harold and COVID-19 on affected areas, around 60  percent of croplands the livelihoods of vulnerable households in Sanma, were severely damaged (FAO, 2020). and the role of a humanitarian cash transfer program, implemented between October 2020 Vanuatu’s social protection system is predomi­ and August 2021 as a response to thesedual nantly informal (ILO, 2015) and the country shocks. It presents findings from a survey, which therefore relied on humanitarian assistance in was conducted among a subset of pro- the aftermath of TC Harold. Informal social gram beneficiaries in Sanma, and provides protection, defined as the care and support recommendations to complement informal safety provided to family and community members nets towards the implementation of a broader through social networks and social structures, is social protection framework in Vanuatu. an essential part of ni-Vanuatu culture. Informal social protection can play an important role in community wellbeing, however, these systems 8 Towards Adaptive Social Protection in Vanuatu 2. The Importance of Cash Transfers and Adaptive Social Protection during Natural Disasters and Economic Shocks Cash Transfer Programs can reduce hardship, improve human capital, such as childhood health, improve human capital and asset accumulation, schooling, and psychological wellbeing (Kremer as well as increase agricultural productivity. et al., Haushofer and Shapiro, 2013; Dietrich et al, A large body of evidence demonstrates the 2019). In addition, CTPs can also support asset importance of cash transfers in reducing poverty. accumulation, such as agricultural machinery or CTPs8 have reduced national inequality and business equipment, allowing households to poverty levels, for example, in Brazil, South Africa, increase productivity (Taaffe et al., 2016). In and Mexico (Arnold et al., 2011; Samson et al., Zambia, for example, maize production increased 2004; Fiszbein et al., 2009). In Mexico, for example, by 8 percent and livestock ownership by 21 percent the nationwide transfer program PROGESA/ as a result of the Social Cash Transfer Program. Oportunidades led to a reduction in the country’s The share of land that households farmed poverty gap by around 29 percent (Fiszbein et al., increased by 18  percent (World Bank, 2021d). 2009). Moreover, evidence suggests that CTPs Similarly in South Africa, households that received 8 Cash Transfer Programs (CTPs) can be designed as unconditional cash transfers, conditional cash transfers (i.e., the transfer is conditional on the recipient meeting certain prerequisites, such as children being enrolled in school), cash for work (or public works), cash plus (e.g., nutrition-sensitive services, agricultural training), vouchers (quasi-cash), and others. 10 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 more income from a Child Support Grant were ASP is the use of social protection programs and more likely to invest in productive assets; engage systems to address covariate shocks,9 such as in poultry, staple crop, and vegetable production; natural disasters and economic crises, to build and grow a larger variety of crops, even 14 years the resilience of poor and vulnerable households. after the intervention (Hajdu et al., 2020). Social protection systems and policies “help individuals and societies manage risk and volatility During the COVID-19 pandemic, CTPs have been and protect them from poverty and destitution – particularly important in supporting vulnerable through instruments that improve resilience, groups and became one of the main social equity, and opportunity” (World Bank, 2012, p.3). protection responses to the crisis around the Social protection thus acts as a buffer to protect world. Globally, 782 cash transfer programs have households by building resilience prior to natural been implemented following the pandemic, as disasters, and supporting quick recovery and countries were able to quickly adapt and expand reconstruction after disasters (Pelham et al., 2011). their existing programs. This represents 23 percent Adaptive social protection systems enable policy of global social protection responses. Data from makers to target support to specific segments of an ongoing study of social protection responses the population that were made poor or vulnerable following the COVID-19 crisis demonstrate the due to disasters.10 importance of pre-existing social protection databases. Countries that could rely on adminis- In countries with established social protection trative data about their population, and had programs, ASP builds on the existing programs identification systems in place, were able to quickly and systems to help the poorest and most identify beneficiaries and implement ASP responses vulnerable individuals build resilience to prepare, to the crisis (Gentilini et al., 2020). Appendix C cope, and adapt along the pre- and post-shock provides an overview of some specific cash continuum (World Bank, 2018). Social protection transfer interventions that were implemented by can build resilience by the promotion of human governments in the Pacific in response to capital gains, job opportunities, accumulation of COVID-19. assets, and diversification of livelihoods Households are, therefore, better prepared to face shocks Vertical and horizontal ASP expansion is generally when they occur. For this reason, over the past deployed through cash transfers which are decade, many countries started to invest in social considered an efficient and timely tool to protection systems. On average, developing and promptly support those in need, particularly transitioning countries spend 1.5  percent of GDP when local markets are functioning and payment on such systems (World Bank, 2018). systems are operational. Cash transfer programs have the advantage of having in place a ASP is a government-led initiative and has four government-to-person delivery system and, in building blocks: social protection programs; data some countries, the pre-identification of poor and and information; institutional arrangements and vulnerable groups. As such, cash and voucher partnerships; and financing. These building programs accounted for 17.9  percent of inter­ blocks emphasize the importance of existing national humanitarian assistance programs in 2019 social protection systems that are cornerstones globally, totaling approximately US$5.6  billion for building household resilience, as well as (CaLP, 2020). In addition, cash has the advantage additional priorities required for systems to be ex- that recipients can choose what they need, in ante prepared. ASP systems allow the government terms of quantity and requirements. to coordinate efforts from multiple agencies, humanitarian actors, and development partners, 9 According to OPM (2019), covariate shocks are understood as shocks that affect a considerable proportion of the population simultaneously, such as natural hazards (hurricanes, floods, droughts, etc.), conflict, or economic shocks. Covariate shocks can be differentiated by type, onset, size, and recurrence. 10 Adaptive Social Protection “…helps to build the resilience of poor and vulnerable households to the impacts of large, covariate shocks, such as natural disasters, economic crises, pandemics, conflict, and forced displacement. Through the provision of transfers and services directly to these households, ASP supports their capacity to prepare for, cope with, and adapt to the shocks they face – before, during, and after these shocks occur. Over the long term, by supporting these three capacities, ASP can provide a pathway to a more resilient state for households that may otherwise lack the resources to move out of chronically vulnerable situations”. (Bowen et al., 2020) 11 Towards Adaptive Social Protection in Vanuatu so that funds are disbursed effectively to priority areas and needs (Bowen et al., 2020). ASP is characterized by programs and delivery systems designed to intervene and scale-up operations when a shock occurs. In countries with mature social protection systems, there are various options for making use of their social protection programs and systems for preparedness and response. A program can be expanded vertically (scaling-up) to increase benefit amounts or the duration of an existing program or system to current beneficiaries at the time of a shock, or expanded horizontally (scaling-out) to temporarily extend support to new households affected by a shock (Figure  2). Programs can have ‘design tweaks’ to adjust the design of routine social protection interventions and systems during a possibility (O’Brien et al., 2018; OPM, 2019; World crisis. Another option is ‘piggybacking’ – which is Bank, 2019). Transfers and services could be used using part of an established system or program in various ways to support households in need, while delivering something new when an emer- although cash transfers were the most widely gency occurs. Alignment of social protection with used during COVID-19. the humanitarian system deployment is also a Figure 2: Social protection programs: Vertical and horizontal expansion Source: Bowen et al., (2020). 12 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 3. The Cash Transfer Program in Sanma: Context, Methods, and Sampling 3.1 Program Description The program targeted households that included at least one vulnerable member. In 2020, Oxfam piloted the UnBlocked Cash Vulnerability was defined according to one of Project, an e-voucher cash transfer program the following criteria: (i) widow/widower, i.e., (CTP)11 in Vanuatu, designed to support partner, husband, or wife is deceased; (ii) vulnerable households affected by TC single mother, i.e., the person has at least one Harold and COVID-19. Oxfam designed the child, lives alone with the child(ren) and has CTP in close collaboration with local NGOs no husband or partner; (iii) people living and development partners.12 The program with a disability (whether a person was aimed to improve consumption levels and disabled was assessed via the Washington livelihoods in the aftermath of the dual shocks questionnaire, which is a standard tool that is in the three provinces of Sanma, Shefa, and used in many surveys to assess whether a Tafea. Oxfam was the technical lead for the person lives with a disability);13 (iv) individuals design and implementation of the transfers – displaced by TC Harold, i.e., those whose identifying, selecting, and registering bene- dwellings were destroyed by TC Harold; or (v) ficiary households for the program. 11 For this report, we are referring to the Unblocked Cash Project as the Cash Transfer Program (CTP), even though technically this is an e-voucher program and there are a small number of restrictions on goods that can be purchased (e.g., alcohol). The program, however, resembles broadly the terms of a traditional cash transfer because the money could be used at any time and at any registered vendor. 12 The CTP is officially called the UnBlocked Cash: TC Harold and COVID-19 Recovery & Response Program. Implementing partners on the ground were World Vision, Red Cross, Adventist Development and Relief Agency (ADRA), Vanuatu Christian Council (VCC), Conference of Churches of Christ in Vanuatu (CCCV), Vanuatu Disability and Promotion Association (VDPA), Vanuatu Society for People with Disabilities (VSPD), Save the Children, and Vanuatu Business Resilience Council (VBRC) (Oxfam 2020). The UnBlocked Cash Project was funded by the Australian Government, Australian Department of Foreign Affairs and Trade (DFAT) through the Australian Humanitarian Partnership (AHP), by the Government of New Zealand, Ministry of Foreign Affairs and Trade (MFAT) and by the International Organization for Migration (IOM). 13 Two of Oxfam’s implementing partners were organizations focused in this area: the Vanuatu Society for People with Disability (VSPD) and the Vanuatu Disability Promotion & Advocacy Association (VDPA). Both organizations have client databases in most of the target areas. If the client database was not available for a certain area or if the implementing partner identified additional participants living with a disability, a team member from either VSPD or VDPA would accompany the partners to verify people living with disabilities. In addition, all field partners received training from either of the partners to understand how to identify people living with disabilities. 13 Towards Adaptive Social Protection in Vanuatu the elderly (60+ years). One vulnerable beneficiary Juden, 2017; Berryhill et al., 2018). Beneficiaries, per household was selected based on these criteria although benefiting from options to receive and thus the beneficiary was not necessarily the payments and transact in specific stores, are still household head.14 financially excluded from the country’s banking system. Depending on the province, the cash transfer was rolled out at different times but all 3.2 Study Design and Sampling beneficiaries received the same payment amount in the end. The payouts started at different points This report evaluates data from only one of the in time depending on the province. The program three CTP provinces – Sanma, where a total of started in Sanma and payouts were made between 2,530 beneficiary households were registered. October 2020 and March 2021. In Shefa, the The two main groups were the elderly and those program started in April 2021 and the last living with disabilities, with 38  percent of installments were made in July 2021. In Tafea, the beneficiaries being above the age of 60  and payouts started in April 2021 and the last 37  percent having a disability. Around 12  percent installments were made in May 2021. Overall, all of beneficiaries were widowed, 10 percent were households received a total CT of VT 70,000 single mothers, and 3  percent were individuals (approx. US$631). In Sanma, the money was displaced by TC Harold.15 transferred monthly, however, the last installment In Sanma, a total of 204 local vendors, including was a double payment of VT 20,000. The program community stores, market vendors, and only ran for one and three months in the other two transportation providers, were registered – provinces and hence, the amount and frequencies variations of shop ownership and distribution of the payments were adjusted accordingly. across areas were observed.16 The number of The CTP program relied on the use of blockchain male vendors was slightly higher than female technology to electronically disburse money to vendors, because the program required beneficiaries, which could be spent at an participating vendors to have a bank account, and authorized local vendor. The money was bank account ownership in Vanuatu is higher transferred to an e-voucher card, which CTP among men (VNSO, 2017; Oxfam, 2021). Moreover, beneficiaries were provided with after they some geographic differences in the prevalence of registered. Beneficiary households could purchase registered vendors were observed. This was mainly goods such as food, hygiene products, and due to the damage of markets and local food agricultural supplies using the e-voucher card at production by TC Harold, which were not fully authorized vendors. The feasibility of the restored when the program started. As a result, the blockchain approach was tested in a previous vendor to beneficiary ratio is lower in some areas; study that took place in Vanuatu in 2019 by Oxfam the area councils of East Malo and West Coast had prior to its implementation (ConsenSys, 2019). the lowest vendor to beneficiary ratio (Appendix A, While there are advantages to blockchain Table A. 1) (Oxfam, 2021). technology, such as lower transaction costs and Preliminary results from an Oxfam CTP higher efficiency in delivering social assistance monitoring survey demonstrate the use of the through direct transfers (Daniels, 2019), there are cash transfer for essential items had a positive also downsides. Governments, in particular, face effect on the local economy. The majority challenges with respect to data privacy laws, (58  percent) of CTP beneficiaries in Sanma used capacity to deliver social protection programs, and the money to purchase food items. The cash the governance of the blockchain system (Pisa and transfers were also used to buy sanitation products, 14 Oxfam worked closely with local partners to ensure gender equity, e.g., through selecting single mothers as a target group. In Sanma and Tafea more than half of beneficiaries were female (55 percent) (Oxfam, 2021). 15 Note that households may include more than one vulnerable household member. However, there is only one beneficiary per household. Thus, the groups are based on the selection criterion of the beneficiary and evaluated based on just one vulnerability group. 16 The network of vendors was diverse and registered vendors included: community stores (53 percent), transportation providers (25 percent), market vendors (12 percent), hardware stores (6 percent), grocery stores (2 percent), energy providers (1 percent), agriculture suppliers (1 percent), bakeries (1 percent), butchers (1 percent), and pharmacies (1 percent). 14 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 clothes, water, and to pay bills. The number of CTP) survey took place in mid-October and beneficiary households considered (near-) food started shortly after the program registration was insecure reduced by 27 percentage points during completed.19 The data collection for the baseline the time of the CTP. In addition, the CTP had survey took about a week and was finished around positive effects on the local economy, as the the time that beneficiaries received their first money was spent in local stores. Around 85 percent payments. The endline (after CTP) data collection of authorized vendors in Sanma reported a positive started a week after the last payouts were made in impact of the CTP on their business, with an mid-March 2021 and ended in mid-April 2021. increase in the volume of customers by 69 percent Although only one beneficiary per household (Oxfam, 2021). was registered, the intention of the CTP was to benefit the whole household. It was assumed Data for this study were collected among a members would share the purchased goods subgroup of the CTP beneficiaries in Sanma and among household members, which was confirmed the final sample includes 194  households. For through the survey results.20 Surveyed respondents this subsample, beneficiary households were thus answered questions representing the randomly selected from the overall CTP registry household as a whole and hence the analysis was from Sanma. Survey weights were applied done at the household level. throughout the analysis to ensure similarity of the subsample drawn for this study to the overall 3.3 Characteristics of Surveyed group of CTP beneficiaries in Sanma based on their vulnerability status. The effects of the Households CTP were analyzed through a ‘before and Most surveyed households were located in rural after’ comparison among the same surveyed areas and were larger than the average household households.17 It was not feasible to identify a con- size in Sanma. The province of Sanma includes trol group for this study, as the CTP was a twelve islands, Santo being the largest, and has a humanitarian response program that targeted all population of about 58,917 (VNSO, 2018). It vulnerable households in Sanma province. consists of 10 area councils, of which eight were Consequently, this study analyzes the impacts of included in the survey. Luganville is the biggest the CTP via a ‘before and after’ assessment among area council in Sanma, however, the majority of the same households – that is, households that surveyed households (73 percent) for this study participated in both the baseline and the endline lived in rural areas outside of Luganville (Figure A. survey.18 The sample of this study is not represen- 2). The average household size for all households tative at the national or local level but only for the in Sanma was 4.6 for the year 2020 (VNSO, 2021a). selected sample in Sanma province. In contrast, surveyed households had on average nine family members but household size ranged Data collection for the baseline and endline from one member to 21 household members surveys among the subgroup of selected (Appendix A, Table A. 2).21 One potential explanation beneficiary households took place in 2020 and for the larger household size among surveyed 2021. The surveyed households were interviewed households is that the CTP specifically targeted via telephone with phone numbers provided vulnerable households, which are often larger. through the CTP registry. The baseline (before 17 The survey tried to reach the same respondents for both waves, however, in a number of cases (around 4 percent) the respondent varied between the baseline and the endline survey. 18 This panel sample excludes households that did not participate in the endline survey and includes only those households that were interviewed both in the baseline and the endline survey. Results throughout the report are based on this same sample of 194 households. However, the results in Section 4.1. and 4.3, which are based on questions from the baseline survey, are comparable when using the full baseline sample (see also Appendix B for more information). A detailed explanation of the sampling procedure and the construction of the survey weights and the final sample is provided in Appendix B. 19 Figure A. 1 in Appendix A provides a timeline of important study dates. 20 Around 98 percent of respondents stated that at least some of the money was shared within the household. Within 47 percent of households, the beneficiary decided on his/her own about how to spend the money. Around 43 percent of households stated that the household head decided on how to use the money and in 10 percent of cases the decision was made jointly between the beneficiary and other household members. 21 Please note that the data collection methods between the VNSO survey and this survey may differ and that the numbers may not be comparable. However, this description still gives some insights into the sample population and how they compare to the general population in Sanma. 15 Towards Adaptive Social Protection in Vanuatu Vanuatu’s population relies heavily on agriculture On average, surveyed households earned less for income generation, which was reflected in than VT 17,275, which is the average monthly the sample. Around 74  percent of surveyed household income in Vanuatu (VNSO, 2021a). The households relied on agriculture as a source of average monthly household income among income and for around 61 percent of households, surveyed households was VT 14,680 (at baseline), agriculture was their main source of income which is approximately US$130. Households with (Figure 3). Similarly, almost all sampled households people living with disabilities had the highest owned some agricultural land (90  percent). Only average monthly household income among the 15  percent of the households relied on income five groups, at VT 15,981, and widowed house- from regular wages and salaries.22 Other income holds had the lowest income at VT 12,771 (Appendix sources included formal businesses, casual wage A, Figure A. 5). Furthermore, house ownership income, and pensions. Around two percent of was high among the sample households, with households did not have any source of income.23 96 percent of households owning their dwelling. Figure 3: Main income sources of surveyed households Source: World Bank staff calculations based on survey data, baseline, 2020. N = 194. 22 Surveyed households without agricultural land live in Luganville (74%), South East (12%), East Malo (8%), Canal Fanafo (1%), and South Santo 1 (4%). Those that earn wages and salary predominantly live in Luganville (48%) and South East (48%). 23 Figure A. 3 and Figure A. 4 in Appendix A provide additional information on income sources by vulnerable group. 16 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 17 Towards Adaptive Social Protection in Vanuatu 4. Role of the CTP in Facilitating Household Recovery 4.1 Impacts of TC Harold on da­mage to their dwelling had not fixed it comple­ tely (Figure  4). The percentage of households Housing and Agricultural that had not fixed their dwelling was particularly Activities high among displaced and widowed households, at 37  percent and 38  percent, respectively. In TC Harold caused severe damage to housing addition, the cyclone destroyed household and household assets in the province of Sanma, assets such as furniture, TVs and radios, home which is reflected in the survey results. Almost all appliances, personal computers, and vehicles. surveyed households in Sanma (97 percent)24 expe­ Around 77  percent of surveyed households rienced at least some damage to their dwelling experienced damage to at least one of these as a result of the cyclone and around 9  percent assets. Losses were again highest among displaced of households did not live in the same dwelling and widowed households. as before TC Harold. Six months after TC Harold, 72  percent of households that experienced 24 Note that all results relate to the surveyed households if not stated otherwise. 18 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19  tatus of housing reconstruction by vulnerable group Figure 4: S Source:World Bank staff calculations based on survey data, baseline, 2020. N = 189. Note: This graph relates to the 96.5 percent of households that experienced damage to their dwelling. The bars by vulnerability group relate to relatively small group sizes and findings are only indicative. TC Harold damaged agricultural land and assets, direct result of TC Harold, which restricted their and hence affected agricultural productivity fishing activities (Appendix A, Figure A. 7). and households’ income generating activities. The majority of households (98  percent) with 4.2 The Role of the CTP in access to agricultural land reported cyclone- Repairing Homes and Restoring related damage.25 Around 77 percent of households experienced a lot of damage to their agricultural Agricultural Assets land or had it completely destroyed. The Cash transfers helped households to repair their percentage of affected households was highest houses and purchase assets that were destroyed among widowed (85  percent). As would be by the cyclone. Houses in Vanuatu are generally expected, displaced households were also made with a mix of traditional building materials significantly affected (79 percent). A large share of (such as those used in natangura houses)26 and households (70 percent) had not completely fixed purchased materials such as wood, concrete, and the damage to their agricultural land at the time of metal. Census data from 2009 indicate that non- the baseline survey (Appendix A, Figure A. 6). traditional housing – that is, housing constructed Moreover, households did not just experience from wood, concrete, and metal – is particularly damage to their dwellings and agricultural land, common in urban areas such as Luganville where but also to crops (83  percent), agricultural many of the surveyed households were located. tools (43  percent), and livestock (36  percent). While material for natangura houses is often Furthermore, around 24  percent of households available for free in rural areas, it is likely that reported that they lost access to the ocean or river damage caused by the cyclone reduced its (due to blocked or damaged roads or paths) as a availability. As a result, many households needed Around 90 percent of surveyed households reported having agricultural land. 25 Housing constructed from traditional materials. These houses typically have that ched roofs woven from the natangura palm (Metroxylon warburgii) 26 and bamboo walls. 19 Towards Adaptive Social Protection in Vanuatu to purchase materials in order to conduct repairs. used the transfer to repair their dwelling. Those Figure 5 presents an overview of how cash transfers households that used the CTP to fix damage were used to cope with the impacts of TC Harold. mainly used the money to repair roofs (70 percent), Around 57 percent of households used the transfer walls and windows (31 percent), floors (27 percent), to repair damage to their dwelling. The percentages and cooking areas (23  percent) (Appendix A, are especially high for elderly, widowed, and Figure A. 8). displaced households where more than 60 percent  se of cash transfers by vulnerable group following TC Harold Figure 5: U Source: World Bank staff calculations based on survey data, endline, 2021. Note: The graph comprises answers from multiple questions by vulnerable groups and for the total sample (outer right bars). Each bar shows the percentage of households that answered th e question with ‘yes’ by groups. For example, a total of 57 percent of households used some of the money to repair their dwelling, while the remaining 43 percent did not. Please note that the total case numbers by question are 194 except for the answer to the question ‘Repair dwelling’, which totals to 193 responses. Also note that the bars by vulnerable group relate to relatively small group sizes and findings are only indicative. In addition, the CTP was used to restore agricul- of widowed households used the CTP for repairs tural land, access to the ocean and river, and to their agricultural land, while the numbers to purchase agricultural tools and machinery. were slightly lower for the other three groups. Twenty-seven percent of surveyed households Of those households that used cash transfers reported that they used cash transfers to repair to repair damage to their land, 61  percent said damage to their agricultural land (Figure 5). Around that they used the money to buy capital inputs, 31 percent of elderly households and 29 percent such as machines and farm tools, and 43 percent 20 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 bought consumable inputs like seed and fertilizer. likely to use the money for this purpose. Furthermore, 8 percent of households stated they Consequently, as the CTP was used to restore used cash transfers to gain access to the ocean or agricultural land and access to aquatic resources, river by buying tools and machinery to restore it likely increased the wellbeing of recipient road access. Single mothers (17  percent) and households.27 widows/widowers (12  percent) were the most In Vanuatu, access to aquatic resources and kastom (traditional) land on which food gardens can be planted is closely linked to wellbeing (VNSO, 2012). 27 Studies show that extreme climate events such as storms and floods can decrease individual subjective wellbeing (Maddison and Rehdanz, 2011; Sekulova and van den Bergh, 2016; von Möllendorff and Hirschfeld, 2016). Thus, TC Harold induced damage to aquatic resources and land likely reduced the wellbeing of impacted households. 21 Towards Adaptive Social Protection in Vanuatu 4.3 Impacts of the Dual Shocks Under normal circumstances, 20.9  percent of ni-Vanuatu experience moderate levels of food on Jobs, Food Security, and insecurity, while 2.4  percent of the population Health experiences severe food insecurity (VNSO, 2021c). After TC Harold, 74 percent of surveyed households In addition to the physical losses resulting from reported that they had run out of food between TC Harold and the pandemic, the dual shocks March and October 2020 (Figure  8) and around also impacted livelihoods and health. Of the 95 percent of households stated that this was due households surveyed, 30 percent included at least to the combined impacts of TC Harold and one member who had lost their job as a direct COVID-19. The number of households that ran out result of TC Harold or COVID-19. Households of food was highest among widows/widowers, with widows/widowers, displaced individuals, and displaced people, and those living with disabilities, those living with disabilities reported higher levels with more than 80  percent of this group having of job losses (Figure  6).28 Around 18  percent of run out of food. Moreover, households reduced surveyed households included a member that their food consumption or switched to less costly needed medical treatment as a direct result of food items to cope with the effects of TC Harold TC Harold. and COVID-19. In fact, these were reported as the The dual shocks impacted household food most important coping strategies following the security and consumption patterns. The damage to shocks (Figure  7). Around 68  percent of house- agricultural produce and the impact of the shocks holds reported reducing their food consumption on income generating activities led to a decrease between March and October 2020, and 69 percent in food security among vulne­ rable house­ holds. reported they switched to less costly food items. Figure 6: Percentage of job losses experienced by vulnerable groups due to dual shocks Source: World Bank staff calculations based on survey data, baseline, 2020. N = 194. Note: The figure shows the percentage of households that include at least one member, who lost his/her job as a consequence of the dual shocks. Please note that the bars by vulnerable groups relate to relatively small group sizes and findings are only indicative. 28 While the survey did not collect data on the number of working members within a household, it is likely that those households have more working members than elderly or single mothers, which might be an explanation for the higher numbers of job losses among these three vulnerable groups. 22 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 In order to cope with the impacts of the shocks, bills (27  percent), or stopped sending children to households adopted a variety of strategies. In school (9  percent) (Figure  7). Some of these addition to reducing their food consumption, strategies can increase households’ vulnerability households used some of their savings (43 percent), and the risk of falling into hardship. In addition, reduced church and community donations or around 85  percent of surveyed households attending ceremonies (36 percent), sold crops or employed two or more coping strategies to deal livestock in advance (32 percent), delayed paying with the consequences of the dual shocks. 23 Towards Adaptive Social Protection in Vanuatu  oping strategies of households between March–October 2020 Figure 7: C Source: World Bank staff calculations based on survey data, baseline, 2020. N = 194. Note: Multiple answers possible. 24 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 4.4 The Role of CTP in Deterring CTP, the number of households that ran out of food was 29 percentage points lower than before Negative Coping Strategies and the CTP (Figure  8). This increase in the level of Supporting Human Capital food security is most likely a combination of the CTP providing money to buy food and a result of The CTP helped households to restore livelihoods, improved access to food over time.29 Importantly, and the money was used for activities such as starting informal income generating activities or around 68  percent of households used cash looking for jobs. Thirty-five percent of households transfers to increase the variety of food items. used the money from the program to restore their Elderly households, in particular, used cash livelihoods and engage in varied income generating transfers to change their food consumption activities (Figure  9). Cash transfers were mainly (75  percent). These findings are significant given used to buy equipment and materials for new that one in ten ni-Vanuatu are undernourished informal income-generating activities but were (VNSO, 2021c), and suggest that longer-term also used to pay for transportation to work or to CTPs might have a role to play in improving food attend job interviews. In the groups of elderly, security. single mother households, and those living with Households also used the CTP to access medical disabilities, more than 30  percent used cash treatment (Figure 9). Almost half of the surveyed transfers for this purpose. households (47  percent) used cash transfers to In addition, cash transfers helped households access medical treatment. This was particularly to switch to more varied food items and thus true of households with elderly members and reduced food insecurity. During the time of the those living with disabilities (around 50 percent). No causal inferences can be made here because of the absence of a control group. 29 25 Towards Adaptive Social Protection in Vanuatu  ercentage of households that ran out of food before and after the CTP Figure 8: P Source: World Bank staff calculations based on survey data, baseline, 2020 and endline, 2021. N = 194. Note: Baseline refers to the period between March to October 2020 and endline refers to the period between October 2020 and March 2021. The bars show the percentage of households that ran out of food. Please note that the bars by vulnerability group relate to relatively small group sizes and findingsare only indicative.  T usage with respect to the impacts of the dual shocks on people’s livelihood Figure 9: C Source: World Bank staff calculations based on survey data, endline, 2021. Note: The graph comprises answers from multiple questions by vulnerable groups and for the total sample (outer right bars). Each bar shows the percentage of households that answered the question with ‘yes’ by groups. For example, a total of 35 percent of households used some of the money to restore livelihoods, while the remaining 65 percent did not. Please note that the total case numbers by question are 194 except for the answer to the question ‘Restoring livelihoods’, which totals to 193 responses. Also note that the bars by vulnerable group relate to relatively small group sizes and findings are only indicative. 26 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 27 Towards Adaptive Social Protection in Vanuatu 5. Informal and Formal Social Safety Nets Informal social safety nets are an important in-kind and cash assistance (Figure 10). Six percent source of support for all ni-Vanuatu, which is of households received only cash. In-kind support reflected in the survey results. The majority of was predominantly received in the form of food surveyed households (72  percent) received some (100 percent), water (49 percent), clothing shelter form of assistance to cope with the effects of TC or housing (25  percent), house repairs and Harold and COVID-19 (Figure 10). This assistance improvements (14 percent), medicine (9 percent), was provided via informal channels or through and agricultural tools (8 percent).31 It was provided formal government sources. Around 66  percent by NGOs (around 86  percent), the Government of surveyed households received assistance (28 percent), family members (21 percent), religious from family, friends, NGOs, or churches. Some institutions (13  percent), and friends (11  percent). 21  percent of households received additional Cash assistance came from family members assistance from the Government.30 (76 percent), NGOs (28 percent), the Government (20  percent),32 remittances (14  percent), or from Assistance was mainly provided in-kind and a friends (13  percent) and was mainly used to buy small percentage of households received cash. food or water. Around 60 percent of households received in-kind assistance and another 6  percent received both 30 Most of those households (92  percent) received government assistance in addition to the assistance received from informal sources. For around 8 percent of those households the assistance from the government was the only assistance they received. 31 Households also received other in-kind assistance such as hygiene products or school supplies but those items made up less than 5 percent of the assistance. 32 In the immediate aftermath of TC Harold, the Government of Vanuatu provided assistance in the form of food and non-food items via its National Disaster Management Office to affected areas (OCHA, 2020). 28 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Figure 10: Overview of non-CTP assistance in the immediate aftermath of the shocks Source: World Bank staff calculations based on survey data, baseline, 2020. N = 193. Note: The figure provides an overview of assistance received between March and October 2020 (i.e., in the months prior to the CTP). Lastly, the survey findings contribute to the (11  percent). This suggests that rather than growing body of evidence indicating that formal compromising or replacing informal social safety social protection systems – such as the CTP – nets, the CTP was incorporated into these existing support existing informal social safety nets. After systems of support. As a result, the benefits of the receiving the CTP, around 75 percent of surveyed CTP were spread beyond participating households households shared their purchased goods with and were able to assist families and communities others, mainly with relatives (84 percent) but also more widely. with friends (23  percent) or religious institutions 29 Towards Adaptive Social Protection in Vanuatu 6. Perceptions on the Implementation of the CTP in Sanma The program assessed in this report is the first The registration for the CTP started in early large-scale CTP implemented in Vanuatu in October 2020 with several registration points response to a natural disaster. The findings provided throughout Sanma. Most beneficiary demonstrate that the CTP greatly assisted the households surveyed were made aware of the surveyed households in Sanma to cope with the program via friends or family (50  percent), local impacts of the dual shocks. The following section authorities (36  percent), or NGOs (19  percent) provides an overview of the CTP implementation (Figure 11). Those that were eligible for the program and how it was perceived by surveyed households. then had to register themselves. The majority Topics discussed include how surveyed households of households (90  percent) travelled to the were informed about the program, means of registration site and experienced relatively short transport and travel time to the registration site travel times of around 30  minutes or less. They and registered vendors, and whether households mainly reached the registration point by walking experienced any difficulties with the program. (47  percent) or by taking the bus (27  percent). About 10 percent of households were registered at home, these were mostly elderly beneficiaries. Figure 11: Overview of information sources Source: World Bank staff calculations based on survey data, endline, 2021. N = 194. Note: Households could select multiple answers. 30 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 The majority of surveyed households reported households walked (45  percent) or took the bus smooth program registration and quick payouts. (39 percent) to reach local stores (Figure 12). The Overall, complaints about the registration process majority of surveyed households (55  percent) and the program itself were low. Only 9 percent of spent some money to reach local stores over the beneficiary households experienced difficulties time of the CTP. A total 34  percent of surveyed with the program registration, relating mainly to households reported difficulties in getting what problems understanding the program content, was needed and this was mainly because the accessing the registration point, or difficulties goods or services were not available (84 percent) relating to a disability or providing the required but also because they did not have enough money information and documents. Some 4  percent of (26  percent). This may indicate that the payment households made a formal complaint about the mode (i.e., households could buy goods only at CTP. These complaints centered on issues around registered vendors) limited access to certain the amount of money (some reported that the products or shops. However, it may also be the transfer was not enough to cover household case that availability of goods was still limited needs) or problems with the e-payment card. due to the prolonged consequences of TC Harold Around 62  percent of surveyed households (i.e., damage of agricultural produce). received their first payment within a week of registration. In 90  percent of cases, the first Most surveyed households reported that the program payouts were provided within 2–3 weeks money from the CTP lasted two weeks or longer. after registration. Around 8 percent of surveyed households reported that the money lasted only one week, 56 percent Households interviewed were able to purchase stated that it lasted about two weeks, and around goods at any registered local vendor with the 36  percent reported that the money lasted e-voucher card and most could do so within a three weeks or longer. For those households travel time of 30  minutes or less. The average for which cash transfers lasted two or three weeks, travel time between the beneficiaries’ homes and they were able to give some of the purchased registered vendors was relatively short. Around goods away and thus support other community 70  percent of surveyed households reported a members. travel time of less than 30 minutes. Most surveyed Figure 12: Mode of transport between home and local stores Source: World Bank staff calculations based on survey data, endline, 2021. N = 192. 31 Towards Adaptive Social Protection in Vanuatu 7. Key Findings After assessing the impacts of the dual shocks 4. The CTP was an effective tool to help of COVID-19 and TC Harold on a subset of vulnerable households to accelerate their households located in Sanma province and how recovery from both the physical and the CTP was used to assist these households, economic impacts of the dual shocks. here are the findings of this report: Surveyed households used the money from the CTP to repair damage to their dwelling 1. Vulnerable households in the province of (57  percent), regenerate or buy agricultural Sanma were highly affected by the dual land (27 percent), regain access to the ocean shocks of TC Harold and COVID-19. Almost or river for the purpose of fishing (8 percent), all surveyed households experienced at least or restore livelihoods (35 percent). In so doing, some damage to their dwelling (97  percent), it is likely that the CTP contributed to increased household assets (77  percent), agricultural wellbeing among participating households. land (98  percent), and agricultural machinery (43  percent). Around 30  percent of surveyed 5. The CTP not only increased access to medical households included at least one member care but also increased access to varied food who lost a job as a direct result of either and thus supported food security and health TC Harold or the impacts of COVID-19. among surveyed households which, in turn, contributes to preserving human capital 2. Many households adopted negative coping gains. At baseline, 74  percent of surveyed strategies in the immediate aftermath of the households reported to have run out of food. shocks. Prior to the CTP, the majority of During the six months of the CTP, fewer households (68  percent) reduced their food households reported running out of food consumption to cope with the impacts of the (45  percent). Surveyed households used the dual shocks of TC Harold and COVID-19. money to increase the consumption of more Households also delayed paying bills varied food items (68  percent). In addition, (27 percent) or removed children from school around 47  percent of surveyed households (9  percent). Negative coping strategies can used the CTP to access medical treatment for increase the risk of households falling deeper some of their household members. into hardship. 6. The report indicates that households found 3. The CTP complemented existing informal the program accommodated their needs. social protection. Informal social safety nets Most surveyed households were able to reach are an important part of ni-Vanuatu culture the registration point of the CTP and stores in and the survey results show that formal social less than 30  minutes, by foot or bus. Only protection programs, like the CTP, can support a small group of surveyed households, around and complement these informal systems. In 9  percent, experienced some difficulties with the immediate aftermath of the shocks, the program registration. Official complaints households received informal assistance, like about the program in general were even food, water, or clothing from friends, family, or lower; only 4 percent of surveyed households the community. During the CTP program, the filed a complaint about the program. Those majority of surveyed beneficiary households households reported that the transfer was not (75  percent) shared their purchased goods enough to cover their needs or that they had with others (e.g., relatives, friends or religious problems with the e-payment card. institutions) and hence supported the community. 32 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 7. While experiences of the CTP were generally the use of other cash transfers modalities positive, the report also highlights challenges rather than an e-payment may therefore faced by households and potential areas for be preferable from a financial inclusion improvement. Firstly, the survey indicates perspective. A careful approach is needed to that the benefit amount may not have been avoid negative unintended consequences on sufficient, as 45  percent of beneficiaries beneficiary households. Finally, beneficiary remained food insecure at the end of the CTP. households may continue to face challenges Secondly, roughly 34  percent of recipients such as increased food insecurity after the experienced difficulties in getting the goods CTP program ends. A longer-term system of they needed. This may hint at the fact that social protection may be useful in alleviating the use of an e-payment card limited the such challenges. availability of shops or that markets had not been fully restored after TC Harold, which limited the availability of goods. In the future, 33 Towards Adaptive Social Protection in Vanuatu 8. Recommendations The following recommendations are put forward 3. Establish an ASP strategy, as this would for the Government of Vanuatu, development result in a more strategic, integrated and partners, and civil society organizations to streamlined approach to disaster recovery, consider in relation to the role of social protec- contributing to a more responsive social tion measures, such as the CTP, to support protection system and a quicker and more vulnerable and poor households in the aftermath inclusive recovery process in particular for of shocks: the poor and most vulnerable people. This should be incorporated into the National 1. Establish an ongoing formal social protection Disaster Recovery Framework and action plan system that complements existing informal (specifically relating to Recovery Objective 1.5 social safety nets, in order to enhance the ‘Strengthen disaster preparedness, response preparedness, response, and coping capacities and recovery mecha­ nisms and structures’). An of vulnerable households and communities. ASP strategy would allow the Government to The people of Vanuatu face recurring disasters better plan, prepare, and manage disaster and due to the country’s high disaster risk. A formal emergency responses – as well as support the social protection system would therefore: (i) most vulnerable. This would enable the support people across the country to build Government to take the lead in disaster resilience prior to disasters, (ii) support response. Establishing an ASP strategy would the population during and after disasters involve elements from existing SP programs to prevent the use of negative coping such as disaster risk financing, databases and strategies, and (iii) provide continued support systems, and government-led coordination to to essential services to vulnerable house- maximize the impact and reach during crises. holds. Humanitarian programs are by nature Placing the Government at the center of the temporary and there is a high risk that social protection response will ensure data households might return to negative coping protection, accountability, and transparency strategies once this support ends. In contrast, of funds. The Government should consider a formal SP system could provide sustainable establishing a steering com­ mittee to assist in and ongoing support as well as the ability leading the development of ASP and SP to build government data and information programs, policies, and systems. This would systems to quickly identify and deliver help to ensure policy coherence for ASP and assistance to those in need. The best way to formalize existing humanitarian interventions. ensure a complimentary approach would be to build on the existing experience and local 4. Focus on ‘building back better’ in order to knowledge already in place by collaborating strengthen the resilience of households by with various partners to support the delivery supporting their capacity to adapt to future of social protection programs and related shocks through ASP and DRM systems. services. Such an inclusive approach would Although this report provides evidence that also underpin SOC 1.4 in the Vanuatu 2030 Plan, households used the CTP to repair damage to which is to strengthen the link between their housing and help them with their recovery traditional and informal governance systems. and reconstruction effort, there is no evidence A formal social protection system could to demonstrate that houses were built stronger support Vanuatu’s population in building than before or that livelihoods were improved resilience to disaster risks. during the recovery process. On the contrary, the temporary nature of the CTP could 2. arguably indicate the prioritization of short- term needs over longer-term considerations. 34 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Government leadership in disaster response, include: (i) temporary but regular ex-ante through Adaptive Social Protection, is crucial labor intensive public works programs aimed to support households to ‘build back better’ at infrastructure requirements for communities and quickly rebuild their livelihoods so that no to withstand shocks, (ii) social protection one is left behind in the recovery process. programs, like cash transfers, (iii) interventions This strategy would align with the pillars of such as behavioral change communications, the Vanuatu 2030 Plan which are: society, financial services, social services, community ronment, disaster resilience, and productive envi­ sensitization, or life skills training aimed at economy (jobs). households reliant on subsistence farming or fishing to improve food security, increase 4. Implement pre-disaster ASP programs and productivity, and/or improve nutrition, and interventions with resilience-building impacts (iv) targeted/seasonal training in livelihood on both basic community infrastructure and diversification. wellbeing of households. These efforts could 35 Towards Adaptive Social Protection in Vanuatu References Arahan, R. J. V., Doan, D. T. T., Dornan, M.S., Munoz, Department of Foreign Affairs and Trade (DFAT). A.G., Silva Parsons, K. H., Yi, S., Vergara, H., and (2021). Tropical Cyclone Harold. Australian Damai, E. (2020). Pacific Island Countries in the Government. https://www.dfat.gov.au/crisis- Era of COVID-19: Macroeconomic Impacts hub/Pages/tropical-cyclone-harold [Accessed and Job Prospects. Washington, D.C.: World online: 25th May 2021] Bank Group. Fiji Village. (2021). More than 20,000 applications Arnold, C., Conway, T., and Greenslade, M. (2011). received for government’s food assistance. DFID Cash Transfers Literature Review. https://www.fijivillage.com/news/More-than- Department for International Development 20000-applications-received-for-governments- (DFID, Policy Division). food-assistance-r5fx48/ [Accessed online: 15th August 2021]. Asian Development Bank (ADB). (2022). Proposed Grant and Administration of Grants Republic of Fischer, V. (2020). Call-based Survey: Vulnerable Vanuatu: COVID-19 Response for Affected Poor Livelihoods and Income Impact – Covid-19 and and Vulnerable Groups Project. Reports and TC Harold. Survey Report. Oxfam. Recommendations of the President. Fiszbein, A., Schady, N., Ferreira, F.H.G., Niall, K., Australian Red Cross. (2020). Local response in a Grosh, M., Olinot, P., and Skoufias, E. (2009). global pandemic: a case study of the Red Cross Conditional Cash Transfers - Reducing Present response to Tropical Cyclone Harold during and Future Poverty. A World Bank Policy COVID-19 in Vanuatu and Fiji. Research Report. The World Bank, Washington, D.C. Bowen, T., del Ninno, C., Andrews, C., Coll-Black, S., Gentilini, U., Johnson, K., Kawasoe, Y., Kryeziu, Food and Agriculture Organization of the United A., Maher, B., and Williams, A. (2020). Adaptive Nations (FAO). (2020). The Republic of Vanuatu Social Protection. Building Resilience to Shocks. - Tropical Cyclone Harold caused widespread International Development in Focus. Washington, damage to the agriculture sector. GIEWS Update. D.C.: World Bank. https://reliefweb.int/report/vanuatu/republic- vanuatu-­t ropical­- cyclone-harold-caused- Behlert, B., Diekjobst, R., Felgentreff, C., Mucke, P., wide­spread-damage-agriculture-sector Pries, L., Radtke, K., and Weller, D. (2020). World Risk Report 2020 – Focus: forced Displacement Gentilini, U., Almenfi, M., Orton, I., Dale, P. (2020). and Migration. Büdnis Entwicklung Hilft. Social Protection and Jobs Responses to COVID-19: A Real-Time Review of Country ConsenSys. (2019). Revolutionising Humanitarian Measures. World Bank, Washington, D.C.: World Cash Transfers in Vanuatu. https://consensys. Bank. https://openknowledge.worldbank.org/ net/blockchain-use-cases/social-impact/ handle/10986/33635. project-unblocked-cash-case-study/ [Accessed online: 15th August, 2021]. Daniels, R. (2019). The pros and cons of blockchain in social assistance, Socialpolicy.org, Online, https://socialprotection.org/discover/blog/ pros-and-cons-blockchain-social-assistance [Accessed, 12th September, 2021]. 36 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Government of Vanuatu. (2020). Vanuatu Recovery Reserve Bank of Vanuatu (RBV). (2020). Quarterly Strategy 2020 - 2030. TC Harold & Covid-19, Economics Review – September 2020. Vol 138, Vanuatu. Yumi Evriwan Tugeta. Department of No. 3. https://www.rbv.gov.vu/index.php/en/ Strategic Planning and Aid Coordination research-publications/311-2020-quarterly- (DSPPAC), Port Vila, Vanuatu. economic-reviews Hajdu, F., Granlund, S., Newves, D., Hochfeld, T., Samson, M., Lee, U., Ndlebe, A., Quene, M., van Amuakwa-Mensah, F., and Sandström, E. (2020). Niekerk, K., Gandhi, I., Harigaya, V., Tomoko, Cash transfers for sustainable rural livelihoods? and Abrahams, C. (2004). The social and eco- Examining the long-term productive effects of nomic impact of South Africa’s social secu- the Child Support Grant in South Africa. World rity system. EPRI research paper. http://alla- Development Perspectives, 16, 2-12. frica.com/download/resource/main/main/ idatcs/00010352:3ca37b223f2ad1b0dc- International Labour Organization (ILO). (2015). 6479ccca726034.pdf. Supporting the definition and extension of a social protection floor (SPF) in Vanuatu and the Sekulova. F., van den Bergh, J.C.J.M. (2016). Floods Solomon Islands (‘the project’). Final Report. Fiji: and happiness: Empirical evidence from Bulgaria, International Labour Organization. Ecological Economics, 126, 51-57. Maddison, D., and Rehdanz, K. (2011). The impact Taaffe, J.E., Longosz, A. F., Wilson, D. (2016). of climate on life satisfaction, Ecological The impact of cash transfers on livelihoods, Economics, 70, 2437-2445. education, health and HIV – What’s the evidence? Development Policy Review, 35, 601-619. Nations Online Project. (2021). Administrative map of Vanuatu. Retrieved from https://www. The Cash Learning Partnership (CaLP). (2020). The nationsonline.org/oneworld/map/vanuatu- State of the World´s Cash 2020: Cash and map.htm [Accessed, 5th May 2022] Voucher Assistance in Humanitarian Aid. Report. O’Brien, C., Scott, Z., Smith, G., Barca, V., Kardan, The Fijian Government. (2021). Applications for A., Holmes, R., Watson, C., and Congrave, J. government-funded unemployment assistance (2018). Shock-Responsive Social Protection open from 9th August 2021. Ministry of Commu­ Systems Research: Synthesis Report, Oxford: nications. https://www.fiji.gov.fj/Media-Centre/ OPM. News/APPLICATIONS-FOR-GOVERN­M ENT­- FUNDED-UNEMPLOYMENT-AS [Accessed online: Oxfam. (2021). UnBlocked Cash: TC Harold and 15th August 2021] COVID-19 Recovery & Response Program – Quarterly Monitoring Report, October- The Office of the Ministry of Internal Affairs. (2020). December. One off Welfare Support, https://www.intaff. gov.ck/covid19-response-package/family- Oxford Policy Management (OPM). (2019). Shock- elderly-children/one-off-support-to-our- Responsive Social Protection in the Caribbean: existing-beneficiaries-on-welfare-support/. Literature Review, Oxford: Oxford Policy [Accessed online: 7th April 2021]. Management. United Nations Office for the Coordination of Pelham, L., Clay, E., and Braunholz, T. (2011). Humanitarian Affairs (OCHA). (2020). Asia and Natural Disasters: What is the Role for Social the Pacific: Weekly Regional Humanitarian Safety Nets? Social Safety Nets Primer Notes; Snapshot (12-18 May 2020). https://reliefweb. No. 32. The World Bank, Washington, D.C.: int/report/philippines/asia-and-pacific-weekly- https://openknowledge.worldbank.org/handle regional-humanitarian-snapshot-12-18-may /10986/11686 -2020 [Accessed online: 15th August 2021] 37 Towards Adaptive Social Protection in Vanuatu United Nations Office for the Coordination of Von Möllendorff, C., and Hirschfeld, J. (2016). Humanitarian Affairs (OCHA). (2021). Vanuatu. Measuring impacts of extreme weather events https://www.unocha.org/office-pacific-islands/ using the life satisfaction approach, Ecological vanuatu [Accessed online 5th June 2021] Economics, vol. 121, pp. 108-116. United Nations Development Program (UNDP). World Bank. (2011). Climate Risk and Adaptation (2018). VNPF to study micro-pensions for Country Profile – Vanuatu. https://climateknow­ Vanuatu’s informal sector. https://www.pacific. ledgeportal.worldbank.org/country/vanuatu undp.org/content/pacific/en/home/presscenter/ pressreleases/2018/vnpf-study-micropensions- World Bank. (2012). Resilience, Equity, and for-Vanuatus-informal-sector.html [Accessed on­ Opportunity – The World Bank´s Social line 8th June 2021] protection and Labor Strategy 2012-2022. Washington, D.C.: World Bank. Vanuatu National Statistics Office (VNSO). (2012). Alternative Indicators of Wellbeing for Melanesia: World Bank. (2013). Building Resilience to Disaster Pilot Study Report, Port Vila, Vanuatu.https:// and Climate Change through Social Protection. vnso.gov.vu/images/Special_Report/Altinative_ Washington, D.C.: World Bank. Indicator_of_Wellbeing/Alternative_Indicators_ World Bank. (2018). The State of Social Safety Nets of_Well-being.pdf 2018, Washington, D.C.: World Bank. Vanuatu National Statistics Office (VNSO). (2018). World Bank. (2019). Social Protection and Disaster Provincial Statistical Indicators – Sanma Recovery, Disaster Recovery Guidance Series, Province, Port Vila, Vanuatu. https://vnso.gov. Washington, D.C.: World Bank and GFDRR. vu/index.php/en/provincial-statistical-indicator- psi#sanma-psi World Bank. (2020a). Tonga - Supporting Recovery after Dual Shocks Development Policy Operation Vanuatu National Statistics Office (VNSO). (2020). (English). Washington, D.C.: World Bank Group. Statistics Release: Gross Domestic Product 2018. http://documents.worldbank.org/curated/en/ Ministry of Finance and Economic Management, 433611606738946310/Tonga-Supporting-Reco­ Port Vila, Vanuatu. very-After-Dual-Shocks-Development-Policy- Vanuatu National Statistics Office (VNSO). (2021a). Operation Well-being in Vanuatu: NSDP 2019-2020 World Bank. (2020b). Samoa - First Response, Baseline Survey, Port Vila, Vanuatu. Recovery and Resilience Development Policy Vanuatu National Statistics Office (VNSO). (2021b). Operation with a Catastrophe-Deferred Draw­ Hardship in Vanuatu: NSDP 2019-2020 Baseline down Option (Cat-DDO) Project (English). Survey, Port Vila, Vanuatu. Washington, D.C.: World Bank Group. http:// documents.worldbank.org/curated/en/106­ Vanuatu National Statistics Office (VNSO). (2021c). ­­3 41606728391675/Samoa-First-Response- Food Security in Vanuatu: NSDP 2019-2020 Reco­very-and-Resilience-Development-Policy- Baseline Survey, Port Vila, Vanuatu. Operation-with-a-Catastrophe-Deferred- Drawdown-Option-Cat-DDO-Project Vanuatu Tourism Office and Department of Tourism. (2020). Survey Results: National World Bank. (2021a). Dealing with Disasters: Tourism Business Impacts Survey: TC Harold and Analyzing Vanuatu´s economy and public COVID-19 pandemic, Port Vila, Vanuatu. https:// finances through the lens of disaster resilience. images.impartmedia.com/vanuatu.travel/docu­ Republic of Vanuatu: Country Economic ments/National_Tourism_Business_Impacts_ Memorandum and Public Expenditure Review Survey_TCHarold_Covid-19.pdf (forthcoming). 38 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 World Bank. (2021b). Climate Change Knowledge Portal – Vulnerability. Vanuatu. https://climate­ know­­­ledgeportal.worldbank.org/country/vanuatu/ vulnerability World Bank. (2021c). Prosperity and Resilience in the Pacific - Pacific Poverty Assessment 2021. National Volume (forthcoming). World Bank. (2021d). In Zambia, Regular Social Cash Transfers Play a Direct Role in Reducing Poverty – Feature Story. https://www.worldbank. org/en/news/feature/2021/05/27/in-zambia- regular-social-cash-transfers-play-a-direct-role- in-reducing-poverty [Accessed online 5th August 2021] World Bank. 2022. Global Economic Prospects, June 2022. Washington, D.C.: World Bank. 39 Towards Adaptive Social Protection in Vanuatu Appendix A: Additional Figures and Tables Figure A. 1: Timeline of events Source: World Bank staff presentation of events. 40 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Table A. 1: Vendor to beneficiary ratio in Sanma Vendor to Vendor ratio Beneficiary beneficiary Area Council Vendor by area Beneficiary ratio by area ratio by area council council council Luganville 41 22% 741 29 % 6 % South East 45 25% 431 17 % 10 % Canal Fanafo 19 10% 400 16 % 5 % South Santo 2 40 22% 270 11 % 15 % South Santo 1 15 8% 253 10 % 6 % East Malo 3 2% 160 6 % 2 % West Coast 2 1% 145 6 % 1 % West Malo 18 10% 126 5 % 14 % Total 183 2,526 Source: Oxfam 2021. Table A. 2: Household composition by vulnerable group Average number Average number Average number Average number Vulnerable group of household of children of adults of children members attending school Elderly 8 5 4 2 People living with 10 5 4 3 disability Widowed 9 5 4 3 Single mother 9 5 4 2 Displaced 10 5 4 3 Total 9 5 4 3 Source: World Bank staff calculations based on survey data, baseline, 2020. N = 194. 41 Towards Adaptive Social Protection in Vanuatu Figure A. 2: Households by area council Source: World Bank staff calculations based on survey data, baseline, 2020. N = 194. Figure A. 3: Main income sources by vulnerable group Source: World Bank staff calculations based on survey data, baseline, 2020. N = 194. Note: Please note that the bars by vulnerable group relate to relatively small group sizes and findings are only indicative. 42 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Figure A. 4: Income derived from agriculture by vulnerable group Yes No Source: World Bank staff calculations based on survey data, baseline, 2020. N =194. Note: The graph shows the percentage of households deriving at least some income from agriculture. Please note that the bars by vulnerability grouprelate to relatively small group sizes and numbers are only indicative. Figure A. 5: Average monthly household income at baseline by vulnerable group Source: World Bank staff calculations based on survey data, baseline, 2020. N = 148. Note: Please note that the bars by vulnerability group relate to relatively small group sizes and findings are only indicative. 43 Towards Adaptive Social Protection in Vanuatu Figure A. 6: Breakdown of status of cyclone induced damage to agricultural land six months after TC Harold (prior to CT) by vulnerable group Source: World Bank staff calculations based on survey data, baseline, 2020. N = 173. Note: This graph relates to the 98 percent households that experienced damage to their agricultural land and reported the status of the damage. Please note that the bars by vulnerability group relate to relatively small group sizes and numbers are only indicative. Figure A. 7: Percentage of households that lost access to the oceanor river because of TC Harold by vulnerable group Source: World Bank staff calculations based on survey data, baseline, 2020. N = 194. Note: Please note that the bars by vulnerability group relate to relatively small group sizes and numbers are only indicative. 44 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Figure A. 8: Breakdown of CT usage for repairs to dwelling Source: World Bank staff calculations based on survey data, endline, 2021. N = 110. Note: Multiple answers possible. This graph relates only to households that used the CT to repair damage to their dwelling. Please note that the bars by vulnerability group relate to relatively small group sizes and numbers are only indicative. Figure A. 9: Breakdown of average travel time from home to the registration site Source: World Bank staff calculations based on survey data, endline, 2021. N = 175. Note: This graph is a breakdown of cash assistance sources among the 89 percent of households that had to travel to register for the CTP. 45 Towards Adaptive Social Protection in Vanuatu Figure A. 10: Mode of transport between home and registration point Source: World Bank staff calculations based on survey data, endline, 2021. N = 193. Note: Households could select multiple answers. 46 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Appendix B: Additional of eligible beneficiaries were identified in Sanma, and a lower number in Tafea than initially identified Descriptions prior to the verification process. Information on the Cash Transfer Program This Report Sample Selection Oxfam and partners distributed the CTP based The sample for this report is based on the on the results from their ‘Vulnerable Livelihood registered households of the CTP in Sanma. In and Income Impact Survey’. This survey was run total, 2,530 beneficiary households were registered prior to the CTP to identify vulnerable groups and in Sanma. During registration, households were geographical areas, which were severely affected asked whether they would be willing to participate by TC Harold and COVID-19 (Oxfam, 2021). It took in an independent survey to assess the program place between May and June 2020. The survey impacts. Around 900  households agreed to collected data among some 1,117 respondents via participate in the survey. For the purpose of this call-based surveys across 52  area councils in subsample, beneficiary households were selected Vanuatu (Fischer, 2020). The results from this from the list of households willing to participate in survey were then used to identify beneficiaries for this survey. It was prespecified that the sample the CTP. Stratified and proportional sampling should include a certain number of beneficiaries methods were used to identify beneficiaries and from each group to ensure a large enough sample vendors in Sanma, Tafea, and Shefa province. For per vulnerable group. Based on these group sizes, the stratification, respondents were grouped into beneficiaries were then randomly selected from six categories: people living with disability, elderly, the list of households willing to participate in the widowed persons, displaced persons, single survey. mothers, and vendors. Moreover, based on the population of targeted group in each strata, the At baseline, 311  households were interviewed. sampling was allocated proportionally across all However, 18 households showed discrepancies in area councils in the three target provinces (Oxfam, their vulnerability assessment (i.e., the reported 2021). Data from the 2010 micro census were vulnerability assessment collected during this used to calculate percentages of each vulnerable survey did not match the Oxfam vulnerability group in the population. Additionally, data from assessment). Those 18 households were excluded the International Migration Organization were from the sample. used to identify displaced households. All households that participated in the baseline Based on these assessments, Oxfam and partners survey were contacted again for the endline identified 2,000  eligible beneficiaries in Sanma survey. A total of 205 households were contacted and Shefa provinces, and another 1,000 in Tafea for the endline survey and the remaining province for the CTP. During the program 106  households could not be reached by the registration, the eligibility of beneficiaries was survey team.33 Another 11  households that were verified via two steps. First, beneficiaries and excluded from the baseline survey due to vendors were verified via a desk verification where discrepancies in their vulnerability profile were vendor ID’s and bank account information were again excluded from the endline sample. Therefore, cross-checked and the eligibility of beneficiaries the endline sample includes 194 households. Since was verified. In addition, eligible beneficiaries were the analysis is based on a before and after verified via an on-spot verification, which was comparison, we only use households that used to rule out any errors in the initial data participated in both surveys for our analysis. We collection (Oxfam, 2021). In the end, 2,530 eligible are aware that the baseline sample has more households were identified in Sanma and observations, however to ensure consistency, we 398 households in Tafea province. The discrepancy ran the analysis with only the 194 households from in numbers is due to the fact that a higher number both surveys. Several attempts were made to reach these households. Since the survey was conducted via telephone it was harder to follow-up with households 33 when phone numbers changed or the phone was not answered. 47 Towards Adaptive Social Protection in Vanuatu An analysis of sample characteristics between Table B. 2: Total respondents by sex in percent households that participated in both surveys and those that did not participate in the endline Sex Percent survey, revealed no significant differences in Male 46.43 % terms of the vulnerability groups and basic Female 53.57 % household characteristics. One difference was found with respect to the area councils. A Source: World Bank staff calculations based on survey data, baseline, 2020. N = 194. significantly higher share of households that did not participate in the endline survey was from the Sample Weights area of Canal Fanafo. A more detailed analysis and explanation of the comparison between those The sample used for this report was drawn with households in the final sample and those that the intention to represent the structure of the dropped out is provided below. Please also note CTP registry in Sanma. Of the 2,530 beneficiaries that results for the baseline survey as reported in registered for the CTP, 38 percent were elderly, Sections 4.1 and 4.3 are similar when using the full 37  percent were people living with a disability, baseline sample. Results for the total sample vary 12 percent were widowed, 10 were single mothers, between 1–3 percentage points. However, we and 3 percent are displaced. The sample for this observe some larger differences for the analysis by survey included higher numbers per vulnerability vulnerable groups due to the smaller sample size. group to ensure that the sample by groups was Results vary by up to 10  percentage points for not too small. Table B. 3 shows the unweighted some questions and vulnerable groups when using sample distribution of households in this report’s the full vs. the restricted baseline sample. Yet, the sample by vulnerability group. To adjust for these overall picture that emerges when using the differences in group sizes between this sample household panel sample, is comparable to using and the CTP registry, sampling weights were the full baseline sample. applied throughout the analysis. The survey was conducted on the household Table B. 3: Unweighted sample level, meaning that one respondent, who was not distribution by vulnerable group necessarily the CT beneficiary, answered the questions for the household as a whole.34 Vulnerability Group size in Respondents had to be 18 years or older. Around Distribution group numbers 57 percent of the respondents were middle aged Elderly 24.74% 48 (i.e., between 30  and 60  years old) (Table B. 1). People living 22.68% 44 Another 27  percent of respondents were elderly with a disability people above 60 years of age. The sample included slightly more female respondents than male Widowed 17.53% 34 respondents, with women comprising roughly Single mother 21.13% 41 54 percent of respondents (Table B. 2). Displaced 13.92% 27 Table B. 1: Total respondents Total 100% 194 by age groups in percent Source: World Bank staff calculations based on survey data. Age groups Percent 18-30 Years 15.28 % 31-40 Years 16.80 % 41-50 Years 14.29 % 51-60 Years 25.98 % 61-70 Years 15.42 % >70 Years 12.23 % Source: World Bank staff calculations based on survey data, baseline, 2020. N = 194. 34 In 96 percent of households surveyed, the respondent was the same person in the baseline and endline survey. 48 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Attrition Analysis South East. In the group of non-participants, a significantly higher share of households is from Table B. 4 presents an overview of sample the area of Canal Fanafo and a significantly lower charac­­teristics between households that parti­ share from South East. A simple regression on the cipated in the endline survey (participants) and likelihood to participate reveals that only Canal those households that did not participate in the Fanafo is significantly correlated with participation endline survey (non-participants). Participants status (Table B. 5). However, there is only a weak and non-participants of the endline survey only significant correlation. Most importantly, there differ with respect to the area a household lives are no significant differences in terms of the in. For two area councils, the proportions of parti- vulnerability groups. cipants and non-participants differs significantly, namely for the area councils of Canal Fanafo and Table B. 4: Comparison sample attrition Variable Full sample Participants Non-participants Difference Area council 3.52 3.55 3.47 0.07 Luganville 0.24 0.26 0.21 0.05 Canal Fanafo 0.2 0.16 0.27 -0.11** East Malo 0.06 0.06 0.07 -0.01 South East 0.18 0.22 0.12 0.10** South Santo Area 1 0.08 0.06 0.11 -0.05 South Santo Area 2 0.12 0.14 0.09 0.05 West Coast 0.05 0.05 0.05 0 West Malo 0.06 0.05 0.07 -0.02 Vulnerable 3.16 3.11 3.27 -0.16 household type Elderly 0.26 0.25 0.27 -0.03 People living with a 0.22 0.23 0.21 0.01 disability Widowed 0.19 0.18 0.21 -0.04 Single mother 0.2 0.21 0.18 0.03 Displaced 0.13 0.14 0.12 0.02 Respondent age 47.26 47.3 47.18 0.12 Respondent sex 1.62 1.63 1.6 0.03 Level of affectedness TC 1.54 1.53 1.58 -0.05 Harold Household size 6.54 6.69 6.26 0.42 Number of adults 4.89 5 4.67 0.33 Income (in VT) 14,388 14,524 14,077 447 Same dwelling 1.1 1.11 1.08 0.03 Observations 293 194 99 293 World Bank staff calculations based in survey data, baseline 2020 and endline, 2021. Note: *, **, and *** denote significance at the 10, 5, and 1 percent levels. 49 Towards Adaptive Social Protection in Vanuatu Table B. 5: Regression results attrition Participation endline 0.153* Canal Fanafo (0.090) 0.046 East Malo (0.131) -0.096 South East (0.086) 0.158 South Santo Area 1 (0.127) -0.068 South Santo Area 2 (0.097) 0.017 West Coast (0.140) 0.083 West Malo (0.140) 0.031 Elderly (0.105) -0.018 People living with a disability (0.099) 0.057 Widowed (0.104) -0.033 Single mother (0.103) -0.001 Respondent age (0.002) -0.024 Respondent sex (0.064) -0.012 Level of affectedness TC Harold (0.037) -0.008 Household size (0.018) -0.008 Number of adults (0.022) -0.095 Same dwelling (0.092) 0.606** Constant (0.256) Observations 293 Source: World Bank staff calculations based on survey data, baseline, 2020 and endline, 2021. Note: *, **, and *** denote significance at the 10, 5, and 1 percent levels. Standard errors in parentheses. 50 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Areas for future research 2. Assess administrative capacity at different government levels to identify areas for future The CTP, which was implemented in the province interventions and to support the delivery of of Sanma, supported vulnerable households, social protection programs, which could be restored livelihoods, and assisted the broader used during natural disasters. community. This reports’ analysis highlighted some potential areas for future research to identify 3. Explore how women, the elderly, and people if, and how, ongoing social protection programs with disabilities are adequately resourced could be built in Vanuatu. Specifically, how future and programmed at each stage of the policy, research could be conducted to: program design, and implementation to sus­ tainably reduce poverty and achieve gender 1. Assess and estimate the level of investment equality. (costing) required for a gradual introduction of social protection interventions as well as providing options on how to sustainably afford social protection programs (fiscal space and social policy analyses), considering coverage and adequacy analyses. 51 Towards Adaptive Social Protection in Vanuatu Appendix C: Examples of Social Protection Responses Following COVID-19 in the Pacific Fiji Tonga The Government of Fiji was able to upscale Following the COVID-19 crisis, the its existing social protection measures to Tongan Government provided top-up support its citizens during the COVID-19 payments to beneficiaries of the Social crisis. The Government provided top-up Welfare Scheme for the Elderly and the payments to recipients of the Poverty Disability Welfare Scheme. Beneficiaries Benefit Scheme (PBS), Child Protection received a one-time payments of T$100 Scheme (CPS), and the Disability Allowance (approx. US$44), which was paid in Scheme (DAS). In addition to their regular addition to the regular benefits. The monthly payment, beneficiaries of the DAS Government also provided payments received two payments of F$50 (approx. to recipients of the secondary school US$24) in August and September 2020. PBS conditional cash transfer program. and CPS recipients received two additional Beneficiaries received a payment of T$200 payments of F$100 (approx. US$48) in (approx. US$89) (Gentilini et al., 2020). August and September 2020. In addition, In addition, wage subsidies were provided CTs were provided to 19,000 Fijian families to workers affected by the COVID-19 most severely affected by COVID-19. crisis. The scheme helped prevent Families received a CT of F$100 (approx. unemployment, especially in the tourism US$48) per month for four months (Gentilini industry, because it lowered the cost et al., 2020; World Bank, 2021g). In May of wages for employers. In total, 2021, the Government distributed food 5,326 affected workers received the assistance to more than 20,000 families wage subsidies, representing over in Suva and Nasori, where supermarkets 13 percent of the labor force. The were closed temporarily because of local subsidies provided a one-time payment lockdowns (Fiji Village, 2021). In August of T$535 (approx. US$238) to workers 2021, the Government announced further and equaled around two-thirds of the support to formal and informal workers average monthly expenditure of a family affected by COVID-19 through monthly of five in the bottom quintile of the payments of F$120 (approx. US$57) to income distribution (World Bank, 2020a). affected workers for a time period of six months (Fijian Government, 2021). 35 Exchange rates were calculated according to: https://xe.com/ (accessed on 27th August 2021). 52 Lessons from a Humanitarian Cash Transfer Program in Sanma Province in Response to Tropical Cyclone Harold and COVID-19 Samoa over 70, to $A 200 (approx. US$147) per month for all citizens over 60. The 2020 The Samoan Government disbursed a payout budget also made the Disability Support of SAT 50 (approx. US$19) per citizen at the Allowance more generous, with the range of end of 2020 (Gentilini et al., 2020). benefits, which depends on the level of Additionally, the Government provided: disability, raised from $A 30–50 (approx. (i) top-up payments to the elderly pension US$22-37) per month to $A 50–80 scheme, (ii) an unemployment benefit to at (approx. US$37-59) per month. least 2,000 workers who lost their jobs as a result of COVID-19, (iii) a short-term paid training scheme for unemployed hospitality workers to at least 200 individuals, and (iv) Cook Islands rent subsidies to vendors at government- The Government of the Cook Islands owned markets (World Bank, 2020b). provided a one-off cash transfer to vulnerable households following the COVID-19 pandemic. In addition to the health risks posed by the virus, vulnerable Kiribati households were impacted by the economic In Kiribati, several social protection measures consequences, such as job losses, were introduced in 2020. These measures that followed from it (Gentilini et al., 2020). helped address the impacts related to the Beneficiary households comprised of those COVID-19 crisis. In September 2020, the that were already on the country’s welfare list. Government of Kiribati introduced the Support The cash transfer was disbursed as a one-off Fund to the Unemployed (SFU). All citizens payment and beneficiaries received a aged 18–59 who are not formally employed or payment of US$400 (The Office of the in a government-sponsored training program Ministry of Internal Affairs, 2020). are eligible to receive an unemployment benefit of $A 50 (approx. US$19) per month. By the end of May 2021, there were more than 52,000 registered beneficiaries, representing nearly three-quarters of the working age Tuvalu population (15–64 years old). In addition, the In April 2020, the Government of Tuvalu Senior Citizen’s Benefit was expanded in two launched a universal CT program to ease important ways in 2020. First, the eligibility the economic impacts of the pandemic. age was lowered from 65 to 60, leading to an All citizens received a monthly payment of increase in the number of eligible beneficiaries US$40 for the first three months of the crisis. to 7,700 people as opposed to 3,000 under In June, when the State of Emergency was the previous age restriction. Second, the lifted, the scheme was altered so that only benefit amount more than tripled from $A persons affected by the pandemic receive 50 per month for age 65–69 and $A 60 for the transfer (Gentilini et al., 2020). 36 According to preliminary data from the 2020 Census of Population and Housing. 53 Towards Adaptive Social Protection in Vanuatu 54