## wpiea2019152

## Source details

**Canonical URL:** [wpiea2019152](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019152.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019152.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019152.pdf.json)

---

### Natural disaster frequency and trends in PICs
- Number of severe disasters by decade and annual totals are presented for 1980-2016 (figures shown in source).
- Total occurrence of natural disasters (1980-2016) and its 3-year moving average indicate variation in frequency over time.
- For Tonga and Tuvalu, severe disasters have typically been much less frequent over the period as a whole, but the interval between natural disasters appears to be shortening for these countries.
- Important implication: countries in the region will typically have less time to recover from greater damage than in the past, underscoring the importance of adaptation and rapid recovery capacity.

### Intervals between natural disasters (1980-2016)
- Country observations noted: WSM, VUT, SLB, MHL, FJI, FSM, TON, TUV (labels: MHL – Marshall Islands, WSM – Samoa, VUT – Vanuatu, SLB – Solomon Islands, FJI – Fiji, TON – Tonga, FSM – Micronesia).
- FSM, TUV and TON have only one disaster interval observation.
- Mean is calculated using total number of disasters divided by the sample years.
- Displayed statistics include: Max years without a severe disaster, Min years in between severe disasters, Median, Mean (as shown in figures in source).

### Impact of severe natural disasters on GDP growth
- Lee et al. (2018) estimate: a severe natural disaster tends to reduce GDP growth by an average of 1.8 percentage points in the Pacific region.
- Country-specific estimated reductions in growth (historical probability differences across PICs):
  - Vanuatu: 0.5-0.6 percentage points
  - Samoa: 0.2-0.4 percentage points

### Role of fiscal buffers
- After a natural disaster, governments need funding for emergency relief and reconstruction while revenues may be adversely affected.
- Typical government responses include reallocating budget spending (often cutting planned capital spending) and drawing down deposits in the central bank or commercial banks.
- External support may include grants and aid from other governments, international organizations, NGOs and the private sector, though amounts and timing are uncertain.
- External insurance, including catastrophe bonds, is an option but may be costly and have limited triggers for payouts.
- Borrowing from domestic or external sources can finance recovery but may harm debt sustainability, especially for highly indebted countries.
- Without sufficient funding for reconstruction, much of the financial burden falls on the private sector, and government resources are redirected away from planned investments and current expenditures.
- Adequate fiscal buffers:
  - reduce dependence on uncertain foreign donations,
  - ensure quick disbursement for recovery and relief,
  - minimize disruption of pre-existing spending plans on health, education, and infrastructure,
  - provide flexibility in addressing post-disaster liquidity needs.

### Definition of fiscal buffers for natural disasters (as used in this paper)
- Fiscal buffers are defined as government financial assets usable to address relief and reconstruction needs, helping to insulate budgeted government spending from major disruption.
- Included elements:
  - Government deposits and funds that can be disbursed expeditiously in the event of natural disaster shocks without much restriction.
  - Funds that can be disbursed if withdrawal rules are changed.
- Excluded elements:
  - Funds established for other purposes and not intended for use in dealing with substantial fiscal shocks.
  - Sovereign wealth funds (SWFs) may be large but may not be able to be used to cover natural disaster costs (see Appendix referenced in source).
- Note: Government cash buffers have great flexibility but would normally be too small to be considered a buffer for natural disasters.

### Existing contingency and disaster funds in PICs
- Currently six PICs have contingency funds for natural disasters (Table 1 summary from source):
  - Fiji: National Disaster Relief and Rehabilitation Fund (known as the Prime Minister's Fund), established 2004; receives an annual contribution of F$ 1 million (US$ 0.5 million).
  - Marshall Islands: Disaster Assistance Emergency Fund, established 2004 (part of Compact of Free Association with the U.S.).
  - Micronesia: Disaster Assistance Emergency Fund, established 2004 (part of Compact of Free Association with the U.S.).
  - Solomon Islands: Natural Disaster Council Fund, established 1989; fund was misused and often diverted away from disaster response activities and has not received an appropriation since 2008.
  - Tonga: National Emergency Fund, established 2008; around US$ 7.8 million.
  - Tuvalu: Climate Change and Disaster Survival Fund, established 2015 with initial funding of AUD 5 million.
  - Tuvalu: Consolidated Investment Fund (CIF) can be considered a fiscal buffer as the government can freely draw down the CIF to finance budget spending.
- Funds that do not fit the paper’s definition of fiscal buffers:
  - Compact Trust Funds (CTFs) in Marshall Islands and Micronesia, and the Climate Change Trust Fund (CCTF) in Tonga — CTFs disbursed as grants from 2024 onwards and no disbursement is allowed prior to 2024; the CCTF’s purpose is to finance small, community-based climate adaptation and mitigation projects and the climate component of non-community-based projects.

### ADB support and contingent financing
- The ADB has worked with several PICs under the Pacific Disaster Resilience Program by establishing the Pacific Disaster Contingent Savings Facility and strengthening disaster risk management frameworks.
- In December 2017, the ADB decided to provide a total of US$ 15 million in financing to Samoa, Tonga and Tuvalu.
- After Cyclone Gita hit Tonga in February 2018, the ADB provided US$ 6 million to the Government of Tonga to help fund priority early recovery activities.
- Footnote detail on assistance composition:
  - For both Samoa and Tonga, total assistance is US$ 6 million ($3.1 million for loans and $2.9 million for grants).
  - For Tuvalu, only grant financing worth US$ 3 million is provided.

### Institutional framework and design characteristics for disaster fiscal buffers
- IMF (2016) recommended characteristics for a well-designed framework:
  - The fund should be consolidated with budget information to allow assessment of the overall fiscal situation; at a minimum, the fund balance should appear in financial statements, and drawdowns should appear in budget execution reports.
  - There should be a standing appropriation that allows spending immediately after a certain trigger event, such as a declaration of a disaster emergency by the executive.
  - The fund should generally apply normal PFM rules, including:
    - Clear rules governing use of resources.
    - Following normal government accounting standards.
    - Preparing and publishing audited financial statements.
    - Defining governance rules.
  - Procurement rules for immediate disaster response should be adjusted to allow for quicker procurement.

### Box 1 — Disaster Assistance Emergency Fund (DAEF) in Marshall Islands: establishment and rules
- The Disaster Assistance Emergency Fund (DAEF) was established in 2004 under the Compact of Free Association with the U.S.
- The agreement prescribes financial contributions from the governments of the Marshall Islands and the U.S., and withdrawal rules and limits.
- From Fiscal Year (FY) 2005 to 2023, the governments of the Marshall Islands and the U.S. need to deposit US$ 200,000 annually in the DAEF.
- After the government declares a state of national emergency, it can withdraw from the fund up to US$ 50,000 per state of emergency.
- No more than US$ 100,000 could be withdrawn in any one calendar year.
- At the end of FY2023, full ownership of the DAEF will pass to the government.

### Box 1 — DAEF: operational guidance and limits
- The fund should be limited to respond to disasters with large fiscal impacts: drawdowns should only start above a minimum fiscal cost; smaller expenditure needs should be covered through budget contingencies.
- The fund should not be allowed to get too large if:
  - (i) its primary purpose is to cover immediate relief and recovery expenses while longer-term rebuilding is financed by borrowing, or
  - (ii) it may generate pressures to tap it for other purposes or discourage international donors from providing grants or loans.
- The financial investment strategy of the fund should be to maintain a high degree of liquidity, given the potential urgency of relief expenditures.
- For PICs, these assets may be better kept offshore because:
  - domestic financial markets may be under stress after disasters;
  - relief and recovery efforts may involve substantial import content;
  - post-disaster repatriation of offshore investments would also strengthen the balance of payments at a time of economic weakness.

### Box 1 — Opportunity cost of holding fiscal buffers and financing alternatives
- Tradeoff: accumulating public savings instead of investing in productive infrastructure forgoes the rate of the social return on the associated public investment; saved funds will eventually be used for relief and recovery, imposing a cost to waiting to use the buffer.
- In the absence of a buffer, costs arise from delayed recovery owing to lack of resources, or from borrowing to rebuild.
- Investing in resilient infrastructure can reduce recovery costs and acts as a substitute with financial buffers up to a point; beyond that point, financial and physical resilience become complements.
- Paying down debt can be better than accumulating fiscal buffers if the cost of public debt is higher than the benefits of holding buffers; however, buffers provide immediate liquidity post-disaster while lower debt only implies more room to borrow and arranging loans may take time.
- Authorities could consider a combination: an asset buffer to cover immediate needs and sufficient borrowing capacity to finance longer-term recovery needs.
- Pre-arranged lines of credit (contingent financing) can reduce borrowing delay; example cited: World Bank development policy loan with a Catastrophe Deferred Drawdown Option (Cat DDO).
  - Cat DDO: contingent financing line providing immediate liquidity after a natural disaster; funds disbursed after drawdown trigger (typically declaration of a state of emergency).
  - At approval, a country must have an adequate macroeconomic policy framework and a satisfactory disaster risk management program in place (or under preparation).
  - Some PICs, such as Vanuatu, are discussing the Cat DDO with the World Bank.
- Choice between lines of credit and own-fund buffers:
  - For infrequent disasters, a line of credit may be preferred as it does not tie up capital.
  - For frequent disasters, a line of credit is likely to be drawn and must be repaid, making budgetary implications similar to saving in a buffer fund.
  - A buffer of own funds has fewer strings attached and can be as large as desired, but may be more prone to being tapped for unintended purposes; governance considerations therefore matter.

### Box 1 — Empirical findings on fiscal costs of severe natural disasters (regional and Marshall Islands specifics)
- Definition: Fiscal cost of natural disasters = Increase in government expenditure + Decrease in government revenue, excluding grants (central government only).
- Estimation approach: panel VAR for 12 PICs using EM-DAT disaster data (1980–2016) and fiscal indicators from World Economic Outlook and World Development Indicators. Severe disasters defined as above the 75th percentile of an intensity measure combining damage-to-GDP and population affected.
- Key results (regional averages):
  - The impact of natural disasters on government expenditure extends to the third year after the initial disaster year; the impact on revenue excluding grants is not statistically significant.
  - On average, a severe natural disaster is likely to increase government expenditure by 13.8-20.6 percent of GDP over a three-year period.
    - In the disaster year, government expenditure is likely to increase by 6.4 percent.
    - In the first year post-disaster, government expenditure tends to increase by 6.8 percent.
    - In the second year post-disaster, government expenditure tends to increase by 7.4 percent.
  - The cumulative fiscal cost for one disaster at the regional level: Disaster year 6.4; 1st yr post-disaster 6.8; 2nd yr post-disaster 7.4; Cumulative fiscal cost for one disaster 13.8 ~ 20.6.
- Marshall Islands (selected figures from Table 4):
  - Total Fiscal Cost (in percent of GDP): 8.1
  - Average Probability: 12.3
  - Implied Average Interval: 12.5
  - Actual Median Interval: 8.0
  - Implied probability: 1.1 ~ 1.6
  - Annual Fiscal Cost (based on median interval): 1.1 ~ 1.7

### Key findings and headline estimates
- The analysis estimates an average cumulative fiscal cost of one severe natural disaster that strikes PICs of 13.8-20.6 percent of GDP.
- Considering a median severe natural disaster interval of 13.8 years, the annual fiscal cost is estimated to be around 1-1.5 percent of GDP.
- Samoa, Vanuatu, and Solomon Islands have the highest annual fiscal cost, reflecting high frequency of disasters and less time to rebuild buffers.
- The fiscal costs shown in Table 4 should probably be regarded as lower bound estimates.

### Data limitations / Caveats
- Estimation uses pooled data to estimate a “typical” disaster fiscal cost in the region; figures are indicative rather than precise.
- Intervals between disasters are based on data since 1980; no allowance made for impact of climate change which may shorten intervals between disasters.
- Fiscal costs of smaller natural disasters are excluded; estimated costs focus only on relatively severe natural disasters.
- EM-DAT is the only source used for natural disaster data; for some disasters EM-DAT damage or population affected numbers differ from other sources (e.g., World Bank), so damage might be under- or over-estimated in certain cases.
- Both timing of disasters (stochastic) and magnitude of impact can vary greatly; even saving at rates indicated would not guarantee coverage for all major disasters—sometimes savings would be more than sufficient, and other times not.

### Determining the appropriate size of fiscal buffers for natural disasters
- The appropriate size of fiscal buffers should account for:
  - (i) external financing (loans and grants, including insurance payments),
  - (ii) domestic financing,
  - (iii) contingency credit or provision in the annual budget and budget reallocation.
- The fiscal cost borne by government can be reduced by involvement of international donors, non-governmental organizations, local business community, and remittances from relatives abroad.
- Fiscal buffer formula presented:
  - Fiscal buffer for natural disasters = Fiscal cost of natural disasters – External financing – (Domestic Financing + Contingent provisions / credits)

### A. External financing (observed experience and ranges)
- Humanitarian aid after severe disasters in PICs ranged from US$ 2.8 million to US$ 43.7 million, depending on severity.
- Insurance payouts under the Pacific DRFI Program: Vanuatu received US$ 1.9 million (about 0.25 percent of GDP) after Tropical Cyclone Pam; Tonga received US$ 3.5 million (about 0.8 percent of GDP) after Tropical Cyclone Gita.
- IMF provided emergency loans to Vanuatu of US$ 23.8 million (about 3.0 percent of GDP) in June 2015 with no interest payment.
- ADB and WB expected to provide Vanuatu around US$ 40 million (about 3.8 percent of GDP) for reconstruction during 2017–2022, with grant elements more than 50 percent.
- For Fiji, ADB and WB each provided US$ 50 million emergency assistance loans.
- Total external support (humanitarian aid in disaster year plus loan disbursement from disaster year onwards) is expected to range from 1.7 to 18.5 percent of GDP, depending on type and severity of disasters.
- Policymakers should be conservative in assuming post-disaster external budget support because much assistance may go directly to the private sector.

### B. Domestic financing (observed experience and constraints)
- Domestic borrowing scope is very limited in PICs and cost of borrowing is generally high; domestic borrowing risks crowding out private sector recovery financing.
- Vanuatu did not borrow domestically after Cyclone Pam for recovery and reconstruction.
- Fiji increased domestic borrowing partly to fund recovery; in FY16-17 infrastructure bonds amounting to almost 2 percent of GDP were sold, largely to the Fiji National Provident Fund (FNPF), commercial banks, and insurance companies.
- Central bank financing is a potential source but should be strictly limited due to risks of fiscal dominance and higher inflation.
- National Provident Funds (NPFs) can support private sector rebuilding and have been used as disaster relief instruments:
  - Fiji allowed FNPF members to withdraw up to 30 percent of their pension account balance after Cyclone Winston.
  - Vanuatu’s NPF allowed withdrawals up to 20 percent of retirement savings, with total drawdowns amounting to VUV 1.7 million (2.1 percent of GDP) for three months after Cyclone Pam.
- Early pension withdrawals can help household recovery but, without measures to strengthen NPF financial position (e.g., adjusting contribution rates), contingent government liabilities for under-saving for retirement could materialize.
- Conclusion: governments should not plan to draw heavily on domestic borrowing as a significant part of public post-disaster financing.

### C. Contingent credit / provision and budget reallocation
- Some PICs have contingency provisions in annual budgets:
  - Solomon Islands sets aside SI$ 26 million (about 0.2 percent of GDP and 0.7 percent of total expenditure).
  - Vanuatu reserves VUV 50 million (about 0.05 percent of GDP and 0.2 percent of total expenditure, excluding grants-funded expenditure) in the 2018 budget.
- These contingency funds are very limited relative to disaster costs and cannot be considered a significant source of disaster financing.
- Governments may use own resources via supplementary budgets and reallocation:
  - Vanuatu spent VUV 402 million (about 0.5 percent of GDP) for Cyclone Pam.
- Budget reallocation allows immediate relief spending but may complicate budget execution and fiscal policy and can be disruptive or costly if investment projects are halted and later restarted.
- Budget reallocation is effectively a means of limiting additional spending rather than providing new financing.

### Case examples (selected numbers)
- Cyclone Pam (Vanuatu, March 2015):
  - Overall damages more than 60 percent of GDP; affected almost 72 percent of the population (more than 188,000 inhabitants).
  - International grants: around US$ 40.2 (4.6 percent of GDP) from bilateral and multilateral donors; around 85 percent disbursed as of end-2017.
  - IMF financial assistance: US$ 23.8 million (3.0 percent of GDP) in June 2015.
  - IDA expected US$ 50 million (about 4.7 percent of GDP) during 2017–2022 (50 percent loans, 50 percent grants).
  - ADB expected about US$ 29 million (about 3 percent of GDP) during 2017–2019 (50 percent loans, 50 percent grants).
  - Government used VUV 401.6 million (about 0.5 percent of GDP) for relief and recovery.
  - Table 4 aggregates Grants total US$ 40.2 (4.6%); Loans total US$ 63.3 (6.8%); Combined totals shown in context (Total 79.7 / 8.5% and Total 63.3 / 6.8% as presented).
- Cyclone Winston (Fiji, early 2016):
  - Damage and losses: US$ 1.42 million (about 31 percent of GDP) and affected more than 60 percent of population.
  - DRF estimated recovery costs at F$ 730 million for fiscal 2016–2018.
  - Government allocated F$ 136 million from own sources and received F$ 22 million from donors, leaving a financing gap of F$ 575 million.
  - Fiji received about US$ 42.1 million humanitarian aid; US$ 21.5 million from UN Flash Appeal.
  - Major bilateral donors included Australia (US$ 27 million), New Zealand (US$ 10.3 million), EU (US$ 4.8 million); about 75 percent recovery support in-kind.
  - Fiji signed US$ 50 million loan agreements with both World Bank and ADB, totaling US$ 100 million.
  - FNPF allowed members to withdraw up to F$ 1000 (US$ 465) plus additional F$ 5,000 (US$ 2,325) with proof of damaged housing; within two months FNPF approved 170,000 withdrawal applications and disbursed about F$ 250 million cash.
  - Projected costs of recovery by priority (in F$ millions): Priority 1: 183.94 (Government 72.14, Donors 0, Gap 111.8); Priority 2: 169.65 (Government 36.17, Donors 9.89, Gap 123.7); Priority 3: 353.39 (Government 25.79, Donors 12.04, Gap 315.57); Priority 4: 23.88 (Government 2, Donors 0, Gap 23.88); Total 730.87 (Government 135.99, Donors 21.93, Gap 574.95). In percent of GDP: Total 7.41 (Government 1.38, Donors 0.22, Gap 5.83).

### Policy implications and recommendations
- PICs are highly vulnerable to natural disasters; frequency of severe disasters has increased sharply in the past decade and climate change may make this the new normal.
- Fiscal costs of disasters (mainly higher expenditures) are substantial, yet fiscal risks from natural disasters are not systematically accounted for in fiscal planning and budgetary policies of PICs.
- In absence of significant self-insurance, PIC governments are highly dependent on foreign aid and budgetary reallocations; uncertainty of external assistance and disruption of budget plans may inhibit full recovery and slow long-term growth.
- PIC governments should start to build natural disaster risk into fiscal and budgetary planning, including budgetary funding of natural disaster buffers as self-insurance in the medium- and long-term.
- The appropriate size and pace of buffer accumulation should be determined case-by-case, taking into account:
  - estimated frequency and fiscal costs of natural disasters,
  - current level of fiscal buffers,
  - availability of alternative financing options, including external loans and grants.

### Box 4 — Contingency Provision and Exceptional Borrowing in Solomon Islands
- The 2018 budget strategy budgets SI$ 26 million (about 0.2 percent of GDP and 0.7 percent of total expenditure, excluding grants-funded expenditure) as a contingency warrant provision for the first time.
- Public Financial Management Act (PFMA) 2013 requirements:
  - Use of contingent warrants should be tabled in the next sitting of National Parliament and disclosed within 30 days after it is tabled.
  - PFMA 2013 allows short-term borrowing to deal with exceptional circumstances like a major economic shock or a natural disaster.
- In the 2017 budget, the government can undertake short-term borrowing up to SI$ 100 million (about 1.0 percent of GDP).
- Concern: given a small domestic market, it is not clear whether the government can secure the funding.

### Box 4 — Considerations for the appropriate scale of fiscal buffers (enumerated)
- (i) Investment in resilient infrastructure:
  - Reduces damage from natural disasters and could lessen the need for fiscal buffers.
  - Building resilient infrastructure and building financial resilience should normally be considered as complements rather than substitutes.
- (ii) Reliance on external donor finance following a disaster:
  - The extent to which governments can rely on external donor finance is uncertain.
  - Much donor assistance goes directly to the private sector rather than to budget support.
- (iii) Risk transfer mechanisms:
  - Include issuance of catastrophe bonds, insurance of physical assets, and regional risk transfer mechanisms, such as the Pacific Disaster Risk Financing and Insurance Program.
  - Currently, the cost of such insurance in the Pacific is high, and its coverage quite limited, so that payouts meet only a small fraction of needs.
- (iv) Scope for domestic borrowing to finance rebuilding:
  - Domestic borrowing should probably be kept to a minimum so as not to crowd out private sector borrowing to support recovery.
- (v) Budget contingency provisions:
  - Funds should be set aside in the annual budget to prepare for frequent shocks, such as relatively small natural disasters.
  - Authorities with a contingency provision in place should assess the level of the provision and raise it when necessary to take into account actual disbursements for previous disaster events and the internationally common practice of setting aside up to 3 percent of spending.
  - Unused contingency provision should be transferred to a precautionary saving fund, such as a natural disaster fund, which could be used during a disaster in the future.
- (vi) Scope for external borrowing, especially contingent lines of credit:
  - Examples include World Bank’s Cat DDO and ADB’s Pacific Disaster Contingent Savings Facility.
  - Such borrowing is a potential substitute for building independent fiscal buffers.
  - Governance properties surrounding access are somewhat different, and the scale of borrowing will likely be limited by the country’s debt sustainability profile.
  - Consequently, even if the country makes use of such facilities, it may well need to build and replenish its own buffers.

*Source: wpiea2019152*

### 7.0 percent of GDP and 7.5 percent of total population, respectively.

### 7.0 percent of GDP and 7.5 percent of total population, respectively.

### Natural disaster frequency and trends in PICs
- Number of severe disasters by decade and annual totals are presented for 1980-2016 (figures shown in source).
- Total occurrence of natural disasters (1980-2016) and its 3-year moving average indicate variation in frequency over time.
- For Tonga and Tuvalu, severe disasters have typically been much less frequent over the period as a whole, but the interval between natural disasters appears to be shortening for these countries.
- Important implication: countries in the region will typically have less time to recover from greater damage than in the past, underscoring the importance of adaptation and rapid recovery capacity.

### Intervals between natural disasters (1980-2016)
- Country observations noted: WSM, VUT, SLB, MHL, FJI, FSM, TON, TUV (labels: MHL – Marshall Islands, WSM – Samoa, VUT – Vanuatu, SLB – Solomon Islands, FJI – Fiji, TON – Tonga, FSM – Micronesia).
- FSM, TUV and TON have only one disaster interval observation.
- Mean is calculated using total number of disasters divided by the sample years.
- Displayed statistics include: Max years without a severe disaster, Min years in between severe disasters, Median, Mean (as shown in figures in source).

### Impact of severe natural disasters on GDP growth
- Lee et al. (2018) estimate: a severe natural disaster tends to reduce GDP growth by an average of 1.8 percentage points in the Pacific region.
- Country-specific estimated reductions in growth (historical probability differences across PICs):
  - Vanuatu: 0.5-0.6 percentage points
  - Samoa: 0.2-0.4 percentage points

### Role of fiscal buffers
- After a natural disaster, governments need funding for emergency relief and reconstruction while revenues may be adversely affected.
- Typical government responses include reallocating budget spending (often cutting planned capital spending) and drawing down deposits in the central bank or commercial banks.
- External support may include grants and aid from other governments, international organizations, NGOs and the private sector, though amounts and timing are uncertain.
- External insurance, including catastrophe bonds, is an option but may be costly and have limited triggers for payouts.
- Borrowing from domestic or external sources can finance recovery but may harm debt sustainability, especially for highly indebted countries.
- Without sufficient funding for reconstruction, much of the financial burden falls on the private sector, and government resources are redirected away from planned investments and current expenditures.
- Adequate fiscal buffers reduce dependence on uncertain foreign donations, ensure quick disbursement for recovery and relief, minimize disruption of pre-existing spending plans on health, education, and infrastructure, and provide flexibility in addressing post-disaster liquidity needs.

### Definition of fiscal buffers for natural disasters (as used in this paper)
- Fiscal buffers are defined as government financial assets usable to address relief and reconstruction needs, helping to insulate budgeted government spending from major disruption.
- Included elements:
  - Government deposits and funds that can be disbursed expeditiously in the event of natural disaster shocks without much restriction.
  - Funds that can be disbursed if withdrawal rules are changed.
- Excluded elements:
  - Funds established for other purposes and not intended for use in dealing with substantial fiscal shocks.
  - Sovereign wealth funds (SWFs) may be large but may not be able to be used to cover natural disaster costs (see Appendix referenced in source).
- Note: Government cash buffers have great flexibility but would normally be too small to be considered a buffer for natural disasters.

### Existing contingency and disaster funds in PICs
- Currently six PICs have contingency funds for natural disasters (Table 1 summary from source):
  - Fiji: National Disaster Relief and Rehabilitation Fund (known as the Prime Minister's Fund), established 2004; receives an annual contribution of F$ 1 million (US$ 0.5 million).
  - Marshall Islands: Disaster Assistance Emergency Fund, established 2004 (part of Compact of Free Association with the U.S.).
  - Micronesia: Disaster Assistance Emergency Fund, established 2004 (part of Compact of Free Association with the U.S.).
  - Solomon Islands: Natural Disaster Council Fund, established 1989; fund was misused and often diverted away from disaster response activities and has not received an appropriation since 2008.
  - Tonga: National Emergency Fund, established 2008; around US$ 7.8 million.
  - Tuvalu: Climate Change and Disaster Survival Fund, established 2015 with initial funding of AUD 5 million.
  - Tuvalu: Consolidated Investment Fund (CIF) can be considered a fiscal buffer as the government can freely draw down the CIF to finance budget spending.
- Funds that do not fit the paper’s definition of fiscal buffers:
  - Compact Trust Funds (CTFs) in Marshall Islands and Micronesia, and the Climate Change Trust Fund (CCTF) in Tonga — CTFs disbursed as grants from 2024 onwards and no disbursement is allowed prior to 2024; the CCTF’s purpose is to finance small, community-based climate adaptation and mitigation projects and the climate component of non-community-based projects.

### ADB support and contingent financing
- The ADB has worked with several PICs under the Pacific Disaster Resilience Program by establishing the Pacific Disaster Contingent Savings Facility and strengthening disaster risk management frameworks.
- In December 2017, the ADB decided to provide a total of US$ 15 million in financing to Samoa, Tonga and Tuvalu.
- After Cyclone Gita hit Tonga in February 2018, the ADB provided US$ 6 million to the Government of Tonga to help fund priority early recovery activities.
- Footnote detail on assistance composition:
  - For both Samoa and Tonga, total assistance is US$ 6 million ($3.1 million for loans and $2.9 million for grants).
  - For Tuvalu, only grant financing worth US$ 3 million is provided.

### Institutional framework and design characteristics for disaster fiscal buffers
- IMF (2016) recommended characteristics for a well-designed framework:
  - The fund should be consolidated with budget information to allow assessment of the overall fiscal situation; at a minimum, the fund balance should appear in financial statements, and drawdowns should appear in budget execution reports.
  - There should be a standing appropriation that allows spending immediately after a certain trigger event, such as a declaration of a disaster emergency by the executive.
  - The fund should generally apply normal PFM rules, including:
    - Clear rules governing use of resources.
    - Following normal government accounting standards.
    - Preparing and publishing audited financial statements.
    - Defining governance rules.
  - Procurement rules for immediate disaster response should be adjusted to allow for quicker procurement.

*Sources: EM-DAT and IMF staff calculations; Lee et al. (2018); IMF (2016).*

### Box 1. Disaster Assistance Emergency Fund (DAEF) in Marshall Islands

### Box 1. Disaster Assistance Emergency Fund (DAEF) in Marshall Islands

### Establishment and legal rules
- The Disaster Assistance Emergency Fund (DAEF) was established in 2004 under the Compact of Free Association with the U.S.
- The agreement prescribes financial contributions from the governments of the Marshall Islands and the U.S., and withdrawal rules and limits.
- From Fiscal Year (FY) 2005 to 2023, the governments of the Marshall Islands and the U.S. need to deposit US$ 200,000 annually in the DAEF.
- After the government declares a state of national emergency, it can withdraw from the fund up to US$ 50,000 per state of emergency.
- No more than US$ 100,000 could be withdrawn in any one calendar year.
- At the end of FY2023, full ownership of the DAEF will pass to the government.

### Operational guidance and limits (policy considerations)
- The fund should be limited to respond to disasters with large fiscal impacts: drawdowns should only start above a minimum fiscal cost; smaller expenditure needs should be covered through budget contingencies.
- The fund should not be allowed to get too large if:
  - (i) its primary purpose is to cover immediate relief and recovery expenses while longer-term rebuilding is financed by borrowing, or
  - (ii) it may generate pressures to tap it for other purposes or discourage international donors from providing grants or loans.
- The financial investment strategy of the fund should be to maintain a high degree of liquidity, given the potential urgency of relief expenditures.
- For PICs, these assets may be better kept offshore because:
  - domestic financial markets may be under stress after disasters;
  - relief and recovery efforts may involve substantial import content;
  - post-disaster repatriation of offshore investments would also strengthen the balance of payments at a time of economic weakness.

### Opportunity cost of holding fiscal buffers and financing alternatives
- Tradeoff: accumulating public savings instead of investing in productive infrastructure forgoes the rate of the social return on the associated public investment; saved funds will eventually be used for relief and recovery, imposing a cost to waiting to use the buffer.
- In the absence of a buffer, costs arise from delayed recovery owing to lack of resources, or from borrowing to rebuild.
- Investing in resilient infrastructure can reduce recovery costs and acts as a substitute with financial buffers up to a point; beyond that point, financial and physical resilience become complements.
- Paying down debt can be better than accumulating fiscal buffers if the cost of public debt is higher than the benefits of holding buffers; however, buffers provide immediate liquidity post-disaster while lower debt only implies more room to borrow and arranging loans may take time.
- Authorities could consider a combination: an asset buffer to cover immediate needs and sufficient borrowing capacity to finance longer-term recovery needs.
- Pre-arranged lines of credit (contingent financing) can reduce borrowing delay; example cited: World Bank development policy loan with a Catastrophe Deferred Drawdown Option (Cat DDO).
  - Cat DDO: contingent financing line providing immediate liquidity after a natural disaster; funds disbursed after drawdown trigger (typically declaration of a state of emergency).
  - At approval, a country must have an adequate macroeconomic policy framework and a satisfactory disaster risk management program in place (or under preparation).
  - Some PICs, such as Vanuatu, are discussing the Cat DDO with the World Bank.
- Choice between lines of credit and own-fund buffers:
  - For infrequent disasters, a line of credit may be preferred as it does not tie up capital.
  - For frequent disasters, a line of credit is likely to be drawn and must be repaid, making budgetary implications similar to saving in a buffer fund.
  - A buffer of own funds has fewer strings attached and can be as large as desired, but may be more prone to being tapped for unintended purposes; governance considerations therefore matter.

### Empirical findings on fiscal costs of severe natural disasters (regional and Marshall Islands specifics)
- Definition: Fiscal cost of natural disasters = Increase in government expenditure + Decrease in government revenue, excluding grants (central government only).
- Estimation approach: panel VAR for 12 PICs using EM-DAT disaster data (1980–2016) and fiscal indicators from World Economic Outlook and World Development Indicators. Severe disasters defined as above the 75th percentile of an intensity measure combining damage-to-GDP and population affected.
- Key results (regional averages):
  - The impact of natural disasters on government expenditure extends to the third year after the initial disaster year; the impact on revenue excluding grants is not statistically significant.
  - On average, a severe natural disaster is likely to increase government expenditure by 13.8-20.6 percent of GDP over a three-year period.
    - In the disaster year, government expenditure is likely to increase by 6.4 percent.
    - In the first year post-disaster, government expenditure tends to increase by 6.8 percent.
    - In the second year post-disaster, government expenditure tends to increase by 7.4 percent.
  - The cumulative fiscal cost for one disaster at the regional level: Disaster year 6.4; 1st yr post-disaster 6.8; 2nd yr post-disaster 7.4; Cumulative fiscal cost for one disaster 13.8 ~ 20.6.
- Marshall Islands (selected figures from Table 4):
  - Total Fiscal Cost (in percent of GDP): 8.1
  - Average Probability: 12.3
  - Implied Average Interval: 12.5
  - Actual Median Interval: 8.0
  - Implied probability: 1.1 ~ 1.6
  - Annual Fiscal Cost (based on median interval): 1.1 ~ 1.7

*Source: Box 1 and surrounding sections on DAEF and fiscal costs of natural disasters in the supplied IMF content.*

### 1.5 percent of  GDP. We use the  median  interval  as a more robust  estimator  than  the sample

### wpiea2019152 - 1.5 percent of  GDP. We use the  median  interval  as a more robust  estimator  than  the sample

### Key findings and headline estimates
- The analysis estimates an average cumulative fiscal cost of one severe natural disaster that strikes PICs of 13.8-20.6 percent of GDP.
- Considering a median severe natural disaster interval of 13.8 years, the annual fiscal cost is estimated to be around 1-1.5 percent of GDP.
- Samoa, Vanuatu, and Solomon Islands have the highest annual fiscal cost, reflecting high frequency of disasters and less time to rebuild buffers.
- The fiscal costs shown in Table 4 should probably be regarded as lower bound estimates.

### Data limitations / Caveats
- Estimation uses pooled data to estimate a “typical” disaster fiscal cost in the region; figures are indicative rather than precise.
- Intervals between disasters are based on data since 1980; no allowance made for impact of climate change which may shorten intervals between disasters.
- Fiscal costs of smaller natural disasters are excluded; estimated costs focus only on relatively severe natural disasters.
- EM-DAT is the only source used for natural disaster data; for some disasters EM-DAT damage or population affected numbers differ from other sources (e.g., World Bank), so damage might be under- or over-estimated in certain cases.
- Both timing of disasters (stochastic) and magnitude of impact can vary greatly; even saving at rates indicated would not guarantee coverage for all major disasters—sometimes savings would be more than sufficient, and other times not.

### Determining the appropriate size of fiscal buffers for natural disasters
- The appropriate size of fiscal buffers should account for:
  - (i) external financing (loans and grants, including insurance payments),
  - (ii) domestic financing,
  - (iii) contingency credit or provision in the annual budget and budget reallocation.
- The fiscal cost borne by government can be reduced by involvement of international donors, non-governmental organizations, local business community, and remittances from relatives abroad.
- Fiscal buffer formula presented:
  - Fiscal buffer for natural disasters = Fiscal cost of natural disasters – External financing – (Domestic Financing + Contingent provisions / credits)

### A. External financing (observed experience and ranges)
- Humanitarian aid after severe disasters in PICs ranged from US$ 2.8 million to US$ 43.7 million, depending on severity.
- Insurance payouts under the Pacific DRFI Program: Vanuatu received US$ 1.9 million (about 0.25 percent of GDP) after Tropical Cyclone Pam; Tonga received US$ 3.5 million (about 0.8 percent of GDP) after Tropical Cyclone Gita.
- IMF provided emergency loans to Vanuatu of US$ 23.8 million (about 3.0 percent of GDP) in June 2015 with no interest payment.
- ADB and WB expected to provide Vanuatu around US$ 40 million (about 3.8 percent of GDP) for reconstruction during 2017–2022, with grant elements more than 50 percent.
- For Fiji, ADB and WB each provided US$ 50 million emergency assistance loans.
- Total external support (humanitarian aid in disaster year plus loan disbursement from disaster year onwards) is expected to range from 1.7 to 18.5 percent of GDP, depending on type and severity of disasters.
- Policymakers should be conservative in assuming post-disaster external budget support because much assistance may go directly to the private sector.

### B. Domestic financing (observed experience and constraints)
- Domestic borrowing scope is very limited in PICs and cost of borrowing is generally high; domestic borrowing risks crowding out private sector recovery financing.
- Vanuatu did not borrow domestically after Cyclone Pam for recovery and reconstruction.
- Fiji increased domestic borrowing partly to fund recovery; in FY16-17 infrastructure bonds amounting to almost 2 percent of GDP were sold, largely to the Fiji National Provident Fund (FNPF), commercial banks, and insurance companies.
- Central bank financing is a potential source but should be strictly limited due to risks of fiscal dominance and higher inflation.
- National Provident Funds (NPFs) can support private sector rebuilding and have been used as disaster relief instruments:
  - Fiji allowed FNPF members to withdraw up to 30 percent of their pension account balance after Cyclone Winston.
  - Vanuatu’s NPF allowed withdrawals up to 20 percent of retirement savings, with total drawdowns amounting to VUV 1.7 million (2.1 percent of GDP) for three months after Cyclone Pam.
- Early pension withdrawals can help household recovery but, without measures to strengthen NPF financial position (e.g., adjusting contribution rates), contingent government liabilities for under-saving for retirement could materialize.
- Conclusion: governments should not plan to draw heavily on domestic borrowing as a significant part of public post-disaster financing.

### C. Contingent credit / provision and budget reallocation
- Some PICs have contingency provisions in annual budgets:
  - Solomon Islands sets aside SI$ 26 million (about 0.2 percent of GDP and 0.7 percent of total expenditure).
  - Vanuatu reserves VUV 50 million (about 0.05 percent of GDP and 0.2 percent of total expenditure, excluding grants-funded expenditure) in the 2018 budget.
- These contingency funds are very limited relative to disaster costs and cannot be considered a significant source of disaster financing.
- Governments may use own resources via supplementary budgets and reallocation:
  - Vanuatu spent VUV 402 million (about 0.5 percent of GDP) for Cyclone Pam.
  - Budget reallocation allows immediate relief spending but may complicate budget execution and fiscal policy and can be disruptive or costly if investment projects are halted and later restarted.
- Budget reallocation is effectively a means of limiting additional spending rather than providing new financing.

### Case examples (selected numbers)
- Cyclone Pam (Vanuatu, March 2015):
  - Overall damages more than 60 percent of GDP; affected almost 72 percent of the population (more than 188,000 inhabitants).
  - International grants: around US$ 40.2 (4.6 percent of GDP) from bilateral and multilateral donors; around 85 percent disbursed as of end-2017.
  - IMF financial assistance: US$ 23.8 million (3.0 percent of GDP) in June 2015.
  - IDA expected US$ 50 million (about 4.7 percent of GDP) during 2017–2022 (50 percent loans, 50 percent grants).
  - ADB expected about US$ 29 million (about 3 percent of GDP) during 2017–2019 (50 percent loans, 50 percent grants).
  - Government used VUV 401.6 million (about 0.5 percent of GDP) for relief and recovery.
  - Table 4 aggregates Grants total US$ 40.2 (4.6%); Loans total US$ 63.3 (6.8%); Combined totals shown in context (Total 79.7 / 8.5% and Total 63.3 / 6.8% as presented).
- Cyclone Winston (Fiji, early 2016):
  - Damage and losses: US$ 1.42 million (about 31 percent of GDP) and affected more than 60 percent of population.
  - DRF estimated recovery costs at F$ 730 million for fiscal 2016–2018.
  - Government allocated F$ 136 million from own sources and received F$ 22 million from donors, leaving a financing gap of F$ 575 million.
  - Fiji received about US$ 42.1 million humanitarian aid; US$ 21.5 million from UN Flash Appeal.
  - Major bilateral donors included Australia (US$ 27 million), New Zealand (US$ 10.3 million), EU (US$ 4.8 million); about 75 percent recovery support in-kind.
  - Fiji signed US$ 50 million loan agreements with both World Bank and ADB, totaling US$ 100 million.
  - FNPF allowed members to withdraw up to F$ 1000 (US$ 465) plus additional F$ 5,000 (US$ 2,325) with proof of damaged housing; within two months FNPF approved 170,000 withdrawal applications and disbursed about F$ 250 million cash.
  - Projected costs of recovery by priority (in F$ millions): Priority 1: 183.94 (Government 72.14, Donors 0, Gap 111.8); Priority 2: 169.65 (Government 36.17, Donors 9.89, Gap 123.7); Priority 3: 353.39 (Government 25.79, Donors 12.04, Gap 315.57); Priority 4: 23.88 (Government 2, Donors 0, Gap 23.88); Total 730.87 (Government 135.99, Donors 21.93, Gap 574.95). In percent of GDP: Total 7.41 (Government 1.38, Donors 0.22, Gap 5.83).

### Policy implications and recommendations
- PICs are highly vulnerable to natural disasters; frequency of severe disasters has increased sharply in the past decade and climate change may make this the new normal.
- Fiscal costs of disasters (mainly higher expenditures) are substantial, yet fiscal risks from natural disasters are not systematically accounted for in fiscal planning and budgetary policies of PICs.
- In absence of significant self-insurance, PIC governments are highly dependent on foreign aid and budgetary reallocations; uncertainty of external assistance and disruption of budget plans may inhibit full recovery and slow long-term growth.
- PIC governments should start to build natural disaster risk into fiscal and budgetary planning, including budgetary funding of natural disaster buffers as self-insurance in the medium- and long-term.
- The appropriate size and pace of buffer accumulation should be determined case-by-case, taking into account:
  - estimated frequency and fiscal costs of natural disasters,
  - current level of fiscal buffers,
  - availability of alternative financing options, including external loans and grants.

*Source: wpiea2019152*

### Box 4. Contingency Provision and Exceptional Borrowing in Solomon Islands

### Box 4. Contingency Provision and Exceptional Borrowing in Solomon Islands

### Budget framework and contingency provisions
- The 2018 budget strategy budgets SI$ 26 million (about 0.2 percent of GDP and 0.7 percent of total expenditure, excluding grants-funded expenditure) as a contingency warrant provision for the first time.
- Prior practice: contingent warrants were not budgeted, leading to cuts in other expenditures or revenue increases to fund unforeseen expenditure when shocks occurred.
- The Public Financial Management Act (PFMA) 2013 stipulates that the use of contingent warrants should be tabled in the next sitting of National Parliament and disclosed within 30 days after it is tabled.
- PFMA 2013 allows short-term borrowing to deal with exceptional circumstances like a major economic shock or a natural disaster.
- In the 2017 budget, the government can undertake short-term borrowing up to SI$ 100 million (about 1.0 percent of GDP).
- Concern: given a small domestic market, it is not clear whether the government can secure the funding.

### Interpretation and caveats on disaster-cost estimates
- Figures on fiscal costs of disasters should be interpreted carefully due to:
  - Considerable uncertainty over the magnitude and frequency of natural disasters.
  - An exceptionally large disaster may have a much higher fiscal cost, and the interval between severe natural disasters varies substantially even for a single country.
  - The severity and frequency of severe disasters are expected to rise due to climate change effects in the future.
  - The analysis excludes natural disasters that are below 75th percentile.
- Nonetheless, the estimation can be used as a benchmark to begin assessing appropriate levels of fiscal buffers and budgetary implications.
- Donors can use these estimates to plan more systematically the likely cost of post-disaster assistance to PICs.

### Considerations for the appropriate scale of fiscal buffers (enumerated)
- (i) Investment in resilient infrastructure:
  - Reduces damage from natural disasters and could lessen the need for fiscal buffers.
  - Investing in resilient infrastructure should raise returns to private investment, employment and output, and facilitate continuous provision of public services.
  - Building resilient infrastructure and building financial resilience should normally be considered as complements rather than substitutes.
- (ii) Reliance on external donor finance following a disaster:
  - The extent to which governments can rely on external donor finance is uncertain.
  - Much donor assistance goes directly to the private sector rather than to budget support.
- (iii) Risk transfer mechanisms:
  - Include issuance of catastrophe bonds, insurance of physical assets, and regional risk transfer mechanisms, such as the Pacific Disaster Risk Financing and Insurance Program.
  - Currently, the cost of such insurance in the Pacific is high, and its coverage quite limited, so that payouts meet only a small fraction of needs.
- (iv) Scope for domestic borrowing to finance rebuilding:
  - Domestic borrowing should probably be kept to a minimum so as not to crowd out private sector borrowing to support recovery.
- (v) Budget contingency provisions:
  - Funds should be set aside in the annual budget to prepare for frequent shocks, such as relatively small natural disasters.
  - Authorities with a contingency provision in place should assess the level of the provision and raise it when necessary to take into account actual disbursements for previous disaster events and the internationally common practice of setting aside up to 3 percent of spending.
  - Unused contingency provision should be transferred to a precautionary saving fund, such as a natural disaster fund, which could be used during a disaster in the future.
- (vi) Scope for external borrowing, especially contingent lines of credit:
  - Examples include World Bank’s Cat DDO and ADB’s Pacific Disaster Contingent Savings Facility.
  - Such borrowing is a potential substitute for building independent fiscal buffers.
  - Governance properties surrounding access are somewhat different, and the scale of borrowing will likely be limited by the country’s debt sustainability profile.
  - Consequently, even if the country makes use of such facilities, it may well need to build and replenish its own buffers.

*Box 4. Contingency Provision and Exceptional Borrowing in Solomon Islands*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019152.pdf_