Eastern Caribbean Currency Union: Selected Issues Paper

International Monetary Fund. Western Hemisphere Dept.

doi:10.5089/9781498300100.002.A001

Jump to Content Jump to Main Navigation

Eastern Caribbean Currency Union: Selected Issues Paper

Author: International Monetary Fund. Western Hemisphere Dept.

Publication Date:: 22 Feb 2019

eISBN:: 9781498300100

Language:: English

Keywords:: ISCR; CR; ECCU country; country authorities; private sector; depreciation rate; resilient investment; investment decision; insurance framework; costs of ND; resilient investment cost; Fiscal sustainability; cost neutral

Download PDF (2.4 MB)

Selected Issues Paper

Abstract

Selected Issues Paper

Building Ex-Ante Resilience to Natural Disasters¹

A. Introduction

1. Natural disasters (NDs) recurrently affect the Easter Caribbean Currency Union (ECCU), resulting in human loss, destruction of infrastructure, and fiscal costs. Natural disasters put pressure on government’s finances in the near and long term. In the near term, pressures arise from unanticipated needs for immediate social protection and rehabilitation expenditures, at a time when revenues typically decline. In the long term, the costs of ND contribute to the ratcheting up of public debt (Acevedo, 2014). For developing small states such as in the ECCU, which are subject to larger and more frequent disasters that affect the entire economy, NDs can have a large impact on the economy and on government finances (text chart).

2. Ex-ante buffers and insurance coverage in ECCU countries is insufficient. The private sector is in general uninsured or underinsured for ND, especially the most vulnerable segments of the population, typically the majority and most exposed. Because of insufficient market-based insurance, governments become the de-facto ultimate insurer, especially for extreme ND events. This means that governments are typically called to cover not only the costs of destruction of public infrastructure, but also a significant share of private losses. Government are also asked to provide for social support, adding to the fiscal pressures. The Caribbean Catastrophe Risk Insurance Facility (CCRIF), to which all ECCU countries have access, have been a valuable instrument, but most countries risk ceding remains below needs mainly because of the perceived high cost and competing developmental needs under fiscal sustainability challenges, and the imperfect correlation between parametric triggers for disbursement and damages.

3. Limited fiscal space constrains resilient investment. Resilient investment is limited, with countries’ efforts and resources allocated mainly to ex-post recovery and reconstruction. Resilient investment is costlier than non-resilient, resulting in difficult spending allocation trade-offs. Political economy factors can also play against government incentives for investment in resilience, as benefits may not be clear to voters in the short term.

4. The results in this paper underscore the importance of a shift from ex-post recovery to a focus on ex-ante resilience building. The region currently invests little in resilience building, with policy response and financing occurring after counties have been hit, resulting in high population exposure and significant asset loss. Recurrence and relatively high frequency of NDs in the ECCU also imply that these costs are incurred multiple times, before structures are fully amortized.

5. Ex-ante resilient investment and insurance are key to the welfare and financial sustainability of the ECCU, given high intensity and recurrence of NDs. Government insurance can provide financing after a ND for immediate relief and rehabilitation. In addition, insurance can strengthen fiscal sustainability because the cost of insurance premia implies internalization of expected costs of NDs’ damages, reducing the need for debt financing after NDs with insurance payouts. Acevedo (2014) finds that tropical storms and hurricanes have a negative effect on growth and a permanent effect on debt accumulation for a subsample of Caribbean countries. The recurrence of NDs implies that their fiscal costs can derail existing efforts to improve fiscal sustainability if this expected source of financial stress is not addressed. Resilient investment can limit asset loss and better support output recovery after NDs. Chapter 2 in this Selected Issues Paper (SIP) shows that including costs and returns of ex-ante resilient investment in robust fiscal frameworks for ECCU countries can accommodate the costs while supporting fiscal sustainability.

6. This paper presents a quantification of the long-term benefits of ex-ante resilient investment and insurance needs against NDs. The paper proceeds in two sections. Section B presents cost-benefit analysis of resilient investment based on a dynamic stochastic general equilibrium model tailored to small states and calibrated to all ECCU economies. Section C quantifies government insurance coverage needs and costs using an empirical stochastic model that simulates NDs fiscal costs. The insurance needs are framed within the World Bank insurance layering framework. Section D concludes.

B. Resilient Investment: Model Simulations

7. This section present cost-benefit analysis of investment in resilient structures in ECCU countries. Resilient investment is costly. It requires additional spending for a given level of investment in physical terms. Benefits, on the other hand, result from lower destruction and capital replacement costs and reduced output loss after NDs. In addition, resilient investment can have a multiplicative effect on output. The decline in losses and destruction implies higher expected returns to private investment when public infrastructure is resilient. This can induce an increase in private investment and capital stock. Moreover, higher investment and capital increase labor productivity. In countries affected by out-migration such as in the ECCU, this can imply an increase in employment by inducing inward migration or reducing out-migration (text chart). Ultimately, higher investment and employment reinforce each other with positive feedback, resulting in a potentially large multiplicative effect on output. This section seeks to quantify these output gains.

Key Model Assumptions

8. The analysis is based on a dynamic stochastic general equilibrium model tailored to capture key features of small states affected by NDs. The model includes four sectors (Appendix I includes a more detailed presentation):

Private Investor Household. It invests physical capital and hires labor to produce a single tradable good competitively. It makes rational investment decisions to maximize the value of household consumption intertemporally. Output decisions depend on factor costs and productivity and also on the stock of public infrastructure invested by the government. Private capital can be destroyed by NDs. Private investors can also hold foreign assets, allowing externally financed investment -important in small island states with large tourism sectors financed with foreign direct investment (FDI). Investor households pay taxes on investment returns and consumption and receive government transfers.
Private Worker household. It supplies labor in the domestic economy to private investors. It can also migrate to work outside the economy and send remittances to the household. It displays hand-to-mouth consumption behavior (no savings), implying it allocates labor to maximize concurrent consumption. It pays labor and consumption taxes to the government and receives government transfers.
Government Worker household. It works in the public sector. It displays hand-to-mouth consumption behavior (no savings), consuming its income in the concurrent period. It pays labor and consumption taxes to the government and receives government transfers.
Government. It collects tax revenues on consumption, capital returns, and wages. Nontax revenues are also collected -capturing mainly grants and Citizenship-by-Investment programs in ECCU countries. Expenditures include public wages, purchase of the tradable good, transfers to all households, interest on public debt, and public investment. Public investment can be of two types: resilient to NDs, and non-resilient. Resilient investment is not damaged by ND shocks, but it is costlier. Both types of capital are assumed to be perfect substitutes in production -their contribution to output is the same.²

9. The model’s aggregate production function illustrates the interaction among the participating sectors and their contribution to output, ultimately informing the role of resilient investment. The production function takes the form

Y_{t} = A_{t} θ_{t} {K_{t - 1}^{g}}^{α g} {K_{t - 1}}^{α k} L_{t}^{d^{β}}

where Y_t is output; A_t is total factor productivity (TFP); θ_t ∈ (0,1] captures efficiency loss in periods tin which the economy has been hit by a ND; $K_{t - 1}^{g}$ is public capital stock determined by government investment; K_t-1 is private capital stock; and $L_{t}^{d}$ is private household labor allocated in the domestic economy. Changes to public capital, private capital, and labor as households react to government’s decisions on resilient investment underpin the model’s predictions on output and other key economic indicators. αg < 1; αk < 1; αk + β = 1. A is assumed to remain constant.

10. Non-resilient public capital and private capital are subject to a random depreciation shocks in periods with NDs. NDs are assumed to be randomly distributed in line with intensity and frequency in the data. Capital stocks evolve according with the following laws of motion:

\begin{array}{l} K_{t}^{r} = K_{t - 1}^{r} (1 - δ) + I_{t}^{r} \\ K_{t}^{n} = (1 - D_{t}) K_{t - 1}^{n} (1 - δ) + D_{t} \int K_{t - 1}^{n} (1 - δ - δ_{t}^{D}) f (δ_{t}^{D}) ⅆ δ_{t}^{D} + I_{t}^{n} \\ K_{t} = (1 - D_{t}) K_{t - 1} (1 - δ) + D_{t} \int K_{t - 1} (1 - δ - δ_{t}^{D}) f (δ_{t}^{D}) ⅆ δ_{t}^{D} + I_{t} \end{array}

where $K_{t}^{r}$ is resilient public capital; $K_{t}^{n}$ is non-resilient public capital; $K_{t}^{g} = K_{t}^{n} + K_{t}^{r}$ is a dummy variable that takes the value D_t = 1 in periods in which the economy is hit by a ND, D_t = 0 otherwise. D_t has a binomial distribution with two possible outcomes, “disaster” and “no-disaster”, with annual probability of a ND P(D) = P to be set in the calibration -therefore 1/P is the frequency of NDs. The probability density function f(δ^D) governs the distribution of capital destruction shocks $δ_{t}^{D}$ by NDs. Notice that resilient capital is not affected by ND shocks, while non-resilient public capital and private capital are both destroyed by NDs.³

11. The government is assumed to follow a passive fiscal policy, with revenue and expenditure instruments set exogenously as policy variables. K^g = Kⁿ + K^r is determined by government investment, which in the model are set as an exogenous policy decision according to public investment allocations I^g = Iⁿ + I^r. I^r is assumed to be costlier than Iⁿ, with a price p^r > 1 -tradable output price is normalized to be equal to 1. The government is not assumed to follow any specific fiscal policy rule nor optimization decision process. It is assumed to follow a passive tax revenue and expenditure stance, maintaining recurrent and capital expenditures constant in real terms. This assumption allows the treatment of government revenue and expenditure parameters as policy variables-including tax rates, level and composition of public investment.

12. The government can borrow externally. Public debt evolves according to the identity

B_{t} = (1 + r) B_{t - 1} + G_{t} - R_{t}

where B_t is the sock of public debt; r is the interest rate on public debt. Primary expenditures G_t and revenues R_t are determined are by

G_{t} = (1 + τ^{c}) C_{t}^{g} + (1 + τ^{L}) w^{g} L_{t}^{g} + T_{t} + I_{t}^{n} + p^{r} I_{t}^{r}

R_{t} = τ^{c} C_{t} + τ^{L} (L_{t}^{d} + L_{t}^{g}) + τ^{K} r_{t}^{K} + N T_{t}

where $C_{t}^{g}$ is government consumption; $L_{t}^{g}$ is public sector employment; $T_{t} = T_{t}^{g} + T_{t}^{L} + T_{t}^{k}$ is transfers to government worker, private worker households, and investors, respectively; and $I_{t}^{n} + p^{r} I_{t}^{r}$ is public investment. With T denoting tax rates, government revenues are determined by consumption taxes $τ^{c} C_{t}; C_{t} = C_{t}^{g} + C_{t}^{w} + C_{t}^{k} + C_{t}^{w g}$ consumption by the government, private workers, private investors, and government workers; labor taxes on domestic private and public sectors $τ^{L} (L_{t}^{d} + L_{t}^{g})$ ; taxes on capital returns $τ^{K} r_{t}^{K}$ ; and non-tax revenues NT_t -which in the ECCU includes mainly donor grants and Citizenship-by-Investment (CBI) revenues.

13. The model assumptions imply that private investment, and thus private capital stock and output, are increasing in the share of resilient public investment. This can be illustrated by inspection of the expected output equation

E_{t - 1} Y_{t} = (1 - P) A_{t} {(K_{t - 1}^{r} + K_{t - 1}^{n})}^{α g} {K_{t - 1}}^{α k} L_{t}^{d^{β}} + P A_{t} θ_{t} \int [K_{t - 1}^{r} + K_{t - 1}^{n} (1 - δ_{t}^{D})]^{α g} [K_{t - 1} (1 - δ_{t}^{D})]^{α k} L_{t}^{d^{β}} f (δ_{t}^{D}) ⅆ δ_{t}^{D}

where E_t-1is the expectations operator as of period t – 1. With probability 1 – P there is no ND and output can be produced using the full amount of capital invested prior to period t. With probability P there is a ND which destroys non-resilient public capital and private capital by a share $δ_{t}^{D}$ . The expected output equation implies that expected marginal productivities of private capital and labor are decreasing in the share of non-resilient public capital, as determined by $δ_{t}^{D}$ in the second term. In other words, private capital and labor employment are high when resilient public capital is also high. This is because expected private output loss is smaller for given NDs’ frequency 1/P and intensity distribution $δ_{t}^{D}$ . As a result, expected private investment returns and thus private capital and output are higher with more resilient public investment. A change in the share of resilient vs. non-resilient capital has a positive permanent or “structural” impact on output.

14. Resilient public capital and the resulting increase in private investment improves labor marginal productivity, inducing upward wage pressure and inward migration. Labor employed in the domestic economy $L_{t}^{d}$ can take different values depending on labor migration decisions and on the occurrence of a ND in each period. Domestic labor supply adjusts until domestic wages and domestic labor marginal productivity are equal to the international wage (labor opportunity cost), that is, when private sector labor is indifferent between working domestically vs. abroad (see Appendix I).⁴ The increase in the share of resilient public capital with the corresponding increase in private capital explained above put upward pressure on marginal labor productivity and domestic wages, inducing inward migration -or reducing outward migration. This also reduces remittance flows to the private worker household -remittances are assumed to be a share of foreign wage income. Notice that higher private investment and inward labor migration reinforce each other’s marginal productivities with positive feedback, resulting in a multiplicative effect on output.

15. Higher output, consumption, and labor improve government revenues, allowing a cost-benefit analysis of the fiscal impact of costlier resilient investment per country. In principle, costlier resilient investment would have a “direct” negative impact on government finances and debt dynamics. However, the endogenous response of the economy also increases tax revenues, the “indirect effect”, underpinned by the improvements in output, consumption, labor, and the stock of private capital. The balance on the economic benefit of the additional cost of resilient investment would therefore depend on how the direct and indirect effects balance out, as per the model calibration for each country.

16. An important remark is that resilient investment costs are effective immediately, while benefits materialize only gradually in the long-term. This implies that a government policy shift towards an increase in resilient investment will worsen fiscal performance in an initial phase, while the stock of resilient capital is gradually built. This implies that resilient investment will, ceteris paribus, increase public debt and financing needs before the output and revenue benefits materialize.

Calibration to ECCU Countries

17. The model is calibrated to replicate key moments of ECCU economies. Public capital stock is calibrated based on public investment rates and depreciation rates. Depreciation rates are set consistent with standard parametrization in the literature. The NDs’ depreciation rate shocks are calibrated based on CCRIF estimated average annual losses (AAL) (text charts), scaled in units of public and private capital, respectively.⁵ The stock of private capital is obtained from steady-state endogenous investment decisions of private investors, plus foreign capital stock based on annual FDI flows in line with historical data. Domestic labor in private and public worker households is calibrated based on labor shares as per each country’s social security data. Labor migration share, defined as the share of the total labor force working abroad, is set according to United Nations data (right text chart).

18. The government sectors are calibrated to match fiscal data. Model tax rates are set to match implicit tax rates in the data –revenue as a share of the corresponding tax base. Revenues per tax and non-tax category are calibrated to match shares to GDP. Expenditures by category are also calibrated to match ratios to GDP in the data. Public workers’ wages are calibrated to match the wage bill, with public employment based on national social security statistics. Public investment is also calibrated to match capital expenditures to GDP ratios.

19. All government revenue and expenditure ratios are set to match long-term levels, with the aim to capture a structural fiscal position excluding transitory factors. The calibrations are based on country-specific averages of historical data and projected trends in the World Economic Outlook database. This is done to remove transitory factors that would distort long-term equilibrium calculations in the policy experiments, resulting in structural equilibria in the calibrations. This remark is important because it implies that transitory dynamics in the simulations, including public debt dynamics, need not match country projections which incorporate anticipated developments, transitory factors, and economic policy shocks. Appendix II presents calibrated parameters.

20. A critical parameter is the price of resilient investment structures, which is set 25 percent more expensive than non-resilient. In the model, this is included by setting the price of resilient capital to 1.25 (the price of the tradable good produced is normalized at 1). This parameter has been set in line with estimates in Ex-Post Damage Assessment Reports by The World Bank, which include estimates of replacement cost of destroyed non-resilient structures and estimated cost of rebuilding with resilience.

Results

21. To evaluate resilience, an experiment is run that consists of increasing the share of resilient public investment to 80 percent of total public investment, while keeping total investment constant in physical terms. First, it is assumed that countries adopt sufficient fiscal consolidation to reach the regional public debt target of 60 percent of GDP by 2030. This ensures sustainable debt dynamics in all countries as a starting condition, thereby enabling the isolation of financing needs that belong to resilience costs exclusively. It is assumed a gradual fiscal consolidation over the initial 5-year period of the simulation, of amount needed to reach the debt target. Second, it is assumed that countries’ initial level of resilient capital is zero –all public investment is non-resilient to NDs. This assumption determines an initial equilibrium under no resilience.⁶ Third, it is assumed that the government increases investment in resilient structures to 80 percent of total investment permanently, while keeping the amount of physical investment constant in real terms. This policy shift gradually changes the composition of the public capital stock until it reaches a share of 80 percent of the total stock (text chart). Given the perfect substitutability assumption between resilient and non-resilient public capital, the shift in public investment composition does not increase output absent any reaction from other sectors. In other words, any changes to output and all underlying endogenous variables capture the endogenous economic response to resilient investment, a pure “resilience effect”.

22. The shift to resilient public investment reduces expected losses from NDs, increasing private investment and capital stock, employment, and output. Resilient public capital reduces private investors’ expected output losses in the event of a ND. As a result, expected returns to private investment are higher relative to non-resilience, resulting in a higher capital stock. The simulations indicate that an increase in resilience from 0 to 80 percent would increase output in all ECCU countries, in the range of 3–11 percent (text chart). Higher private investment of 4–13 percent increases the stock of private capital and the returns to labor and wages, inducing inward labor migration with a reduction of the labor force working abroad of 5–25 percent and higher domestic employment.⁷ Investment and inward labor migration reinforce each other with positive feedback, increasing output. Variations across countries are mainly explained by the size of public investment and capital stock; share of migrant labor; exposure to NDs (frequency and intensity); government size in the economy; share of public capital in total capital; and tax policy mix (i.e. direct vs. indirect taxation; investment returns vs. labor taxation).

23. Countries also benefit in the near term with reduced asset loss and output after NDs. Resilient investment contains output decline after NDs by reducing capital destruction and labor out migration. Also, replacement capital needs and costs are lower. According to the model simulations, these two sources imply additional gains equivalent to 0.7–2.7 percent of GDP on average per year, in addition to the structural increase in the level of GDP (text chart).

24. Fiscal performance improves in the long term, with resilient investment returns more-than-compensating costs. The long-term increase in tax revenues underpinned by higher output, labor, and consumption more-than-offsets the higher cost of resilient investments. As a result, overall fiscal balances improve in the range of 0–3 percentage points of GDP (text chart) with the increase in resilient investment to 80 percent.⁸

25. These output benefits, however, accrue in the long-term, while fiscal positions deteriorate in an initial phase. If ECCU economies start from a state of non-resilience, benefits from a shift to resilient public investment may take a long time, possibly over 40 years before the share of resilient capital becomes dominant. This implies that the growth and tax benefit accrue at a slow pace. The simulations in this paper indicate that if countries were to increase the share of resilient public investment to 80 percent of total investment, growth acceleration during the transition will be in the range of 0.1–0.4 percent per year on average (text chart). However small, these growth accelerations would compound over time.

26. Fiscal performance deteriorates in an initial phase. Costlier resilient investment worsens the fiscal balance initially, because higher tax revenues from resilience takes time to materialize. For example, if a country has a public investment rate of 5 percent of GDP and increases resilient investment to 80 percent while keeping constant total physical investment, the fiscal balance deteriorates by 1 percent of GDP (5 x 0.8 x 1.25 + 5 x 0.2 x 1 = 6). The simulations indicate that the additional cost of resilience would increase public debt by 4–20 percentage points of GDP in the ECCU countries by 2030 above the regional target (text chart). The gap to be filled would be about 0.4–1.5 percent of GDP per year above historical levels to reach the regional debt target, or about US$ 60 million for the region annually. These financing gaps, however, should be interpreted as a financing floor. Public investment may need to be increased above historical norms to accelerate resilience building, especially in some cases with low public investment such as Antigua and Barbuda and Saint Vincent and the Grenadines, or countries under reconstruction that have been recently affected by NDs such as Dominica. In the simulations, only about half of the public capital stocks would be resilient by 2030 at the current investment rates (text chart).

C. Quantifying Insurance Needs: A Layering Framework

27. As part of the strategy to build resilience ex-ante in the ECCU, this section presents estimates of insurance coverage needs and costs in ECCU countries. The results are based on a Monte Carlo experiment including stochastic simulations of output and fiscal revenues and expenditures as these are affected by ND shocks. NDs are identified as the tail of the distribution of fiscal deteriorations after other sources of large shocks have been controlled for in the model estimates. The results are used to estimate insurance coverage needs within a layering framework, in line with World Bank recommended practice. Two sets of simulations are produced, before and after resilient investment –the latter using the results in Section B for 80 percent resilient capital. The first set of simulations allows the quantification of insurance needs and costs in the near term, when resilient investment is low. The second set of simulations recalculates insurance needs and costs after incorporating the benefits of building resilience ex-ante. This exercise informs of plausible fiscal savings from lower insurance costs once ECCU economies are resilient.

28. The three layers included in the simulations are as follows⁹:

Layer 1. Saving Fund (SF) for self-insurance against relatively small but more frequent NDs. This would be the first line of defense, but insufficient for large NDs. It is the least costly –the interest rate on public debt as the opportunity cost. It requires annual budget savings to remain sustainable in expected terms.
Layer 2. Caribbean Catastrophe Risk Insurance Facility (CCRIF). It would provide additional funding when financing needs after NDs are above the SF’s depletion point, adding coverage for large disasters.
Layer 3. State-contingent debt financing instruments could be used for extreme NDs. These instruments are typically the costliest, mainly because of low potential issuance scale (ECCU states are small) and high fixed cost of damage valuation. Issuance should therefore be limited to extreme events when needs are high and parametric triggers are likely to activate.

29. The instruments in the layers used in the simulations are specifically chosen with consideration of fiscal sustainability challenges prevalent in the region. They imply internalization of NDs expected costs in the form of savings for self-insurance and insurance premia on a recurrent basis. This cost internalization de facto works as a disciplinary device by preventing expenditure of these resources in other allocations, and effectively reducing the need for debt issuance to recover from NDs. This choice of instruments, however, remains illustrative and other options should be considered to balance befits and costs, in line with country-specific considerations. A key option is World Bank’s Catastrophe Deferred Drawdown Option (CAT DDO), which could be second layer to reduce the need for costlier insurance, or to reduce the size of self-insurance needs, but implies debt issuance.

30. The simulations assume that the SFs are initially financed with revenues from the Citizenship by Investment (CBI) programs. In recent years, there has been a substantial windfall in budget revenues from CBI Programs in ECCU countries. These are significant and thus relevant from a macroeconomic perspective (text chart). Using CBI revenues would effectively reduce debt issuance after NDs. It will also avoid its allocation to recurrent expenditures, which is difficult to revert when CBI revenues decline, thus reinforcing fiscal sustainability.¹⁰

Insurance Simulation Methodology – Summary Presentation

31. The methodology for the quantification of insurance needs and costs is based on a Monte-Carlo experiment for each country. It can be summarized in the following steps: (see Appendix II for a more technical presentation):

Step 1: Estimation of fiscal models for each ECCU country. VAR(2) models are estimated for each ECCU country. The vector of endogenous variables includes GDP; non-grant revenue; grants; current primary expenditure; and capital expenditure. All variables are expressed in real terms and as deviations from trend. The vector of controls includes the U.S. real effective exchange rate; oil price; US cycle; and a September 11 2001 dummy –for the terrorist attack in the U.S. that severely disrupted tourism revenues in the ECCU. The control variables are selected to remove other competing sources of large shocks to output and fiscal performance from estimated residuals. This allows the interpretation of the tail of the estimated residuals’ distribution as NDs shocks –the only remaining large shock.
Step 2: Monte-Carlo experiment. Simulate 1000 off-sample stochastic projections with random shocks drawn from the historical distribution. Given the VAR approach, the simulated series display same volatility, persistence, and co-movement as in the historical data.
Step 3: Identification of NDs in the simulations. NDs are identified as the tail X percent fiscal deteriorations in simulations. This is calibrated by setting a Probability of NDs of 1/X (NDs occur every 1/X years on average. X is set in line with NDs data.
Step 4: Define SF’s inflow-outflow rules. This includes a calibration of the SF stock amount, annual budget savings into the SF, and payouts to the budget after NDs. The SFs’ size and annual savings are calibrated to cover the fiscal costs of NDs in 95 percent of the cases, thus resulting insufficient in the largest 5 percent NDs. The annual savings and size are calibrated to ensure the financial sustainability of the SF –no increase or decrease in expected terms.
Step 5: CCRIF coverage and state-contingent debt issuance. The simulations assume that countries purchase parametric insurance targeting coverage of 99 percent of NDs expected fiscal costs. CCRIF’s attachment point (“deductible”) is set at the 10-year estimated loss according to each county’s estimated loss function. The coverage limit is set at the 100-year loss.¹¹ If the coverage limit is insufficient to reach the 99 percent coverage target, it is assumed issuance of state contingent debt in the form of a CAT bond. The simulations include CAT bonds with 3-year maturity and 500 basis points spread over Libor.

32. CAT bonds do not increase net debt and pose no liquidity risk. Bond issuance proceeds are typically held in an Special Purpose Vehicle (SPV) that can only be accessed to service debt when the call option is triggered by a ND.¹² The net fiscal cost of the CAT bond is therefore the difference between the SPV investments’ return and the interest rate on public debt. Notice also that the debt service/liquidity risk of gross debt issuance of a typical bond is less applicable to the CAT bond because the CAT bond provides debt service relief, and it remains a liquid investment while unused.

Results: Quantitative Assessment of Insurance Needs

33. This section presents two sets of results: insurance needs and costs with low and high resilient investment. First, the insurance needs and costs with low levels of resilience, which is relevant in the near term. Second, the long term needs after resilient investment has been completed. The latter illustrates plausible declines in insurance costs –one of the benefits of resilient investment.

Insurance with Low Resilience

34. The simulations indicate that for the ECCU covering 99 percent of the fiscal costs of NDs requires coverage of 13–31 percent of GDP. Under the illustrative coverage assumptions in the simulations, SFs for self- insurance amount to 6–12 percent of GDP. All countries require maximum CCRIF access, with coverage estimated in the range of 2–17 percent of GDP. As this remains insufficient reach 99 percent coverage, all countries issue CAT bonds in the range of 2–5 percent of GDP (text chart). These thresholds and coverage levels are illustrative. Countries should choose coverage and instrument composition according to preferences towards risk aversion, fiscal space, capacity constraints, or other idiosyncratic considerations. Institutional capacity to safeguard the integrity of the SFs should also be taken into consideration.

35. Annual fiscal costs of the insurance layers and targets above are in the range of 0.5–1.8 percent of GDP. The simulations indicate higher costs for Dominica, Antigua and Barbuda and St. Kitts and Nevis, reflecting higher estimated Average Annual Losses of NDs in those countries. The cost composition also points at the relatively expensive nature of insurance instruments. SFs are more cost effective relative to the significant level of coverage targeted (text chart). In the simulations, insurance costs have multipliers of 1.5–2.0 (ratio of insurance premia to expected payouts). Notice that multipliers above 1 imply that insurance worsen fiscal sustainability in expected terms.

36. The high insurance costs could be difficult to accommodate given fiscal sustainability challenges in the ECCU. Increasing the share of coverage with SFs could reduce costs, but it could prove challenging to maintain in practice given competing developmental needs and political pressures for spending. It is important to remark that targeting a lower coverage level does not reduce the expected costs of NDs, it only implies no internalization of these costs and need for additional debt issuance ex-post.

37. Insurance needs will open an additional financing gap. The international community, including climate funds, can reduce incentives for underinsurance with concessional financing to cover insurance costs, as part of a comprehensive ex-ante resilience strategy. Concessional financing could help equalize insurance costs with expected average annual losses of NDs –effectively resulting in fiscal multipliers of 1. This implies an incentive-compatible strategy for governments and donors. From the governments’ perspective, it incentivizes the purchase of appropriate levels of insurance coverage without worsening long-term fiscal sustainability in expected terms. From the international community’ perspective, it ensures the allocation of government resources in line with their fiduciary mandates –and could therefore result in an increase in donor grant flows. Under the simulation assumptions above, making insurance fiscally-neutral in expected terms implies grants in the range of 0.2–1.1 percent of GDP per year –equivalent to about US$40 million per year for all ECCU countries (text chart).

Insurance with High Resilience

38. Investment in resilient structures would reduce the cost of insurance in the long term. The results above on insurance coverage and costs are based on historical data, therefore capturing low resilience. However, if countries pursue resilient investment, the expected costs of reconstruction and the disruption of economic activity after NDs would gradually decline. This implies a decline in expected losses and tax revenue with NDs, and thus lower insurance needs. By incorporating the results in Section B on the long-term benefits of resilient public investment on output, tax revenues, and fiscal costs and savings from resilient investments in the Monte Carlo experiment, insurance needs decline to about ¼ relative to no resilience (text chart). To ensure results’ comparability, the assumptions of 80 percent resilience in Section B and coverage levels of 95 percent of NDs fiscal costs with a SF and up to 99 percent with insurance are maintained. In this scenario, the 99 percent coverage is reached with a higher share of more cost-effective SFs, and without need for costlier CAT bond issuance.

D. Conclusions: Putting it all Together

39. Recurrence and intensity of NDs affecting ECCU countries requires a comprehensive ex-ante resilience strategy. The quantitative exercises in this paper indicate that it can yield significant long-term benefits. First, investment in resilient physical infrastructure increases the level of output and tax revenues on a permanent basis. Second, it implies savings from reduced reconstruction costs and output decline after NDs. Third, when combined with a financial insurance layering strategy, it helps address fiscal sustainability concerns from NDs’ shocks. Fourth, resilience and insurance coverage reinforce each other with positive feedback: over time, resilient investment reduces insurance needs, while insurance protects fiscal space for costlier resilient structures.

40. In the illustrative example simulation of 80 percent resilient investment as a share of total investment, output levels can increase by 3–11 percent in the long term. This is because the private sector internalizes higher returns to private investment and employment, including through a decline in labor out migration. The results also indicate that, despite its higher cost, resilient investment improves fiscal performance in the long term with the increase in tax revenues, underpinned by the increase in output, employment, investment, and consumption, assuming government spending remains constant in real terms.

41. The long-term benefits, however, imply up-front costs that deteriorate public finances in the near term, requiring a fiscal effort before resilience benefits materialize. Costlier resilient structures increase governments’ capital spending. For example, in the simulations with an increase in resilient structures to 80 percent of total public capital, capital expenditures would increase by 0.6–1.5 percent of GDP for the same levels of investment in physical terms. Insurance costs are also significant, implying additional fiscal cost of 0.5- 1.9 percent of GDP in the near term, before substantial physical resilience is achieved.

43. Insurance costs, however, need not worsen fiscal sustainability if supported by grants with appropriate incentives. Annual insurance costs are largely recovered when payouts proceed after NDs:

The net cost of maintaining SFs is small. It is determined by the spread between the interest rates on public debt (which in most cases is low given prevalence of concessional official financing) and returns on SFs’ investments. Annual fiscal savings needed to achieve the financial sustainability of the SFs with low probability of depletion support fiscal sustainability by ensuring NDs’ fiscal costs are appropriately internalized in the fiscal frameworks.
CCRIF and state-contingent debt are needed to ensure coverage of large but less frequent NDs. However, high premia relative to expected payouts imply worsening of fiscal sustainability in expected terms –insurance costs are about twice expected payouts for high coverage options. Donor grants could play an important role by making insurance cost neutral, for example, in the form of an insurance subsidy that covers excess insurance costs above expected payouts.

43. The layers’ triggers could be calibrated to achieve an efficient cost-minimizing insurance framework. Countries with particularly large CBI deposits in reserve could increase self-insurance and thus reduce the need of costlier options. Also, CCRIF’s attachment point (“deductible”) and coverage limit could be calibrated to ensure payouts are triggered when SFs are near depletion in expected terms. This might require tuning of CCRIF’s coverage options to match insurance instruments in other layers’ exhaustion and triggering points.¹³

44. A framework to support fiscal sustainability is a necessary precondition for a consistent resilience financing strategy. Given limited fiscal space, the timing of resource allocation is key. The recurrence and potentially devastating impact of NDs implies SFs should be created as soon as feasible, in light of low initial levels of physical and financial resilience. CBI resources could finance startup costs. Increasing insurance coverage will aid fiscal sustainability by ensuring internalization of NDs’ fiscal costs and mitigating the need for debt issuance. Meanwhile, ECCU countries should also pursue resilient public investment in all eligible projects given substantial returns. The amount of investment, however, would need to remain consistent with fiscal sustainability, a necessary condition given the long-term nature of resilience building, and mindful of capacity constrains (i.e. availability of specialized labor, financial spillovers to the private sector, administrative and execution constraints). Over time, as resilience improves, insurance needs would need to be reassessed to internalize the savings from a decline in expected damages.

45. After fiscal consolidation, there will be financing gap to be filled. Resilient investment and insurance are costly, putting pressure on government finances in an initial phase, before the benefits of resilience materialize in meaningful amounts from a macroeconomic perspective. However, in light of insufficient fiscal space, most ECCU countries would find it difficult to afford the costs of resilient investment and insurance without support from the international community, including climate funds. For example, if all ECCU countries adopt fiscal consolidation to reach the regional debt target of 60 percent of GDP by 2030, the additional costs of resilient investment and insurance would add financing gaps of 0.4–1.5 and 0.2–1.1 percent of GDP per year on average through 2019–2030, respectively (text chart).

46. Early specification and communication of fiscal consolidation plans is necessary to ensure resilient investment can be sustained over time without financing disruption. This is critical given the long-term nature of resilience building. It is also key to facilitate donor grant financing eligibility. The simulations indicate that concessional financing would need to increase by about US$100 million for the region if they were to fill the financing gaps above, for public investment remaining at historical levels in physical terms. Supporting the fiscal plans with strong institutions, including in the form of fiscal rules, would signal commitment to fiscal sustainability and increase the chances of concessional financing, in light of the fiduciary responsibilities and due-diligence requirements of donor funds.

47. Concessional financing could also contribute to fiscal sustainability with appropriate contractual design. Donor grants’ disbursement conditions can be specified to achieve incentive-compatible cost sharing of resilient investment and NDs’ insurance costs. The simulations in this paper provide one such example with concessional financing, including grants, for resilient investment assumed to be disbursed after a credible fiscal consolidation framework is in place with specific fiscal consolidation targets, and with concessional financing to cover insurance premia to make it cost-neutrality –i.e. countries pay share of premia to reach an insurance multiplier of 1 or less.

References

Acevedo, S. 2014. “Debt, Growth and Natural Disasters: A Caribbean Trilogy” IMF Working Paper 14/125.
- Search Google Scholar
- Export Citation
Barro, R. 2006. “Rare Disasters and Asset Markets in the Twentieth Century.” Quarterly Journal of Economics, 121: 823–899.
- Search Google Scholar
- Export Citation
Barro, R. 2009. “Rare Disasters, Asset Prices and Welfare Costs.” American Economic Review, 99(1): 243–264.
- Search Google Scholar
- Export Citation
Cavallo, E. A., and Noy, I., 2011. “Natural Disasters and the Economy: A Survey.” International Review of Environmental and Resource Economics, 5: 63102.
- Search Google Scholar
- Export Citation
Cavallo, E., Galiani, S., Niy, I., and Pantano, J., 2013. “Catastrophic Natural Disasters and Economic Growth.” Review of Economics and Statistics, 95(5): 1549{1561.
- Search Google Scholar
- Export Citation
Cummins, J.D. 2008. \CAT Bonds and Other Risk-Linked Securities: State of the Market and Recent developments.” Risk Management and Insurance Review, 11(1): 23–47.
- Search Google Scholar
- Export Citation
Cummins, J.D. 2012. “CAT Bonds and Other Risk-Linked Securities: Product Design and Evolution of the Market.” In Extreme Events and Insurance: 2011 Annus Horribilis, ed. Christophe Courbage and Walter R. Stahel, 39–61. The Geneva Association.
- Search Google Scholar
- Export Citation
Cummins, J. D., and O. Mahul. 2009. Catastrophe risk financing in developing countries: principles for public intervention. World Bank Publications.
- Search Google Scholar
- Export Citation
El-Ashram, A., Gold, J. Xu,X., 2015. “Too Much of a Good Thing? Prudent Management of Inflows under Economic Citizenship Programs.” IMF Working Paper 15/93.
- Search Google Scholar
- Export Citation
Froot, Kenneth A. 2001. “The Market for Catastrophe Risk: a Clinical Examination.” Journal of Financial Economics, 60(2): 529–571.
- Search Google Scholar
- Export Citation
Lee, Jin-Ping, and Min-Teh Yu. 2007. “Valuation of Catastrophe Reinsurance with Catastrophe Bonds.” Insurance: Mathematics and Economics, 41(2): 264{278.
- Search Google Scholar
- Export Citation
Noy, Ilan. 2009. “The Macroeconomic Consequences of Disasters.” Journal of Development Economics, 88(2): 221–231.
- Search Google Scholar
- Export Citation
Rasmussen, T. N. 2004. “ Macroeconomic Implications of Natural Disasters in the Caribbean .” IMF Working Paper 04/224.
- Search Google Scholar
- Export Citation
World Bank, 2017. “Sovereign Climate and Disaster Risk Pooling”. International Bank for Reconstruction and Development, Chapter 1. Section 5.
- Search Google Scholar
- Export Citation

Appendix I. Model Assumptions

The economy produces a single tradable good with the following aggregate production function

Y_{t} = A_{t} θ_{t} {K_{t - 1}^{g}}^{α g} {K_{t - 1}}^{α k} L_{t}^{d^{β}}

where Y_t is output; A_t is total factor productivity; θ_t ε (0,1] captures efficiency loss in periods t in which the economy has been hit by a ND; $K_{t - 1}^{g}$ is public capital stock determined by government investment; K_t-1 is private capital stock; and $L_{t}^{d}$ is private household labor allocated in the domestic economy. αg < 1; αk<1; αk + β = 1. The different sectors contribute to output by supplying factors of production. The government supplies public capital; private investor households supply private capital; and private worker households supply labor.

In each period t, the sequence of events and decisions is as follows. First, the ND shock materializes. With the remaining undestroyed private and public capital, private investors hire labor from private worker households. Private worker households decide labor allocation in the domestic economy or to migrate and earn the foreign wage and send remittances to the household for consumption. Labor markets clear, output is produced and consumption and investment decisions proceed. The government collects taxes and makes transfers to households. Any gap between government revenues and expenditures is financed with foreign debt.

Private Investor Household

Private investors are assumed to be rational and to maximize expected welfare. It is assumed the representative private investor household chooses a consumption sequence ${c_{t}^{k}}_{t = 0}^{\infty}$ , makes investment decisions ${i_{t}}_{t = 0}^{\infty}$ , and hires labor $l_{t}^{d}$ to maximize discounted utility

\begin{matrix} U = Σ_{t = 0}^{\infty} β^{t} \log (c_{t}^{k}); β \in (0, 1) & (1) \end{matrix}

subject to the constraints

\begin{matrix} k_{t} + q_{t} = (1 - D_{t}) (1 - δ) k_{t - 1} + D_{t} \int k_{t - 1} (1 - δ - δ_{t}^{D}) f (δ_{t}^{D}) ⅆ δ_{t}^{D} + i_{t} + (1 + r^{*}) q_{t - 1} & (2) \end{matrix}

\begin{matrix} i_{t} = y_{t}^{e} - (1 + τ^{c}) C_{t}^{k} - w_{t}^{d} l_{t}^{d} - (1 + τ^{k}) r_{t}^{k} k_{t - 1} + T^{k} & (3) \end{matrix}

\begin{matrix} y_{t}^{e} = (1 - P) y_{t} + P y_{t}^{'} & (4) \end{matrix}

\begin{matrix} y_{t} = A_{t} {(K_{t - 1}^{r} + K_{t - 1}^{n})}^{α g} {k_{t - 1}}^{α k} l_{t}^{d^{β}} & i f n o N D \\ y_{t}^{'} = A_{t} θ_{t} \int [K_{t - 1}^{r} + K_{t - 1}^{n} (1 - δ_{t}^{D})]^{α g} [k_{t - 1} (1 - δ_{t}^{D})]^{α k} l_{t}^{d^{β}} f (δ_{t}^{D}) ⅆ δ_{t}^{D} & i f N D & (5) \end{matrix}

\begin{matrix} k_{0} > 0; q_{0} > 0 & (6) \end{matrix}

where $K_{t}^{r}$ is resilient public capital; $K_{t}^{n}$ is non-resilient public capital; with $K_{t}^{g} = K_{t}^{n} + K_{t}^{r}$ is a dummy variable that takes the value D_t = 1 in periods in which the economy is hit by a ND. It is assumed D_t has a binomial distribution with two possible outcomes (disaster and no-disaster), with probability of the ND state P(D) = P. The probability density function f(δ^D) governs the distribution of capital destruction shocks $δ_{t}^{D}$ by NDs. Resilient capital is not affected by ND shocks, while non- resilient public capital and private capital are both destroyed by NDs in the proportion $δ_{t}^{D}$ .¹ (2) is the private assets accumulation identity, with first term determining available capital in periods with no ND, second term with ND, and third term representing investment in international risk free assets q_t-1 yielding return r^*. (3) is the intra-temporal budget constraint: private investment it is the amount of realized output y_t and government transfers T^k remaining after investor’ consumption $c_{t}^{k}$ and retribution to production factors, and tax payments. Notice that investors rent/lend capital in the domestic economy or invest for own production, at the domestic interest rate r_t. (4) is expected output. (5) is the production technology, with output y_t depending on the occurrence of a ND at the beginning of t. k₀ is capital endowment and q₀ is the initial stock of international assets.

Competitive market and representative producer assumptions implies that in equilibrium the number of producing units is irrelevant, allowing the aggregation of firm-specific capital stock k_t into aggregate private capital K_t.

Private Sector Worker Household

Labor for domestic production is supplied by a representative household endowed with labor that can opt to migrate to work abroad and send remittances. It is assumed worker households exhibit hand-to-mouth behavior, thereby making no savings or investment decisions. Their only optimizing behavior is the allocation of labor endowment to maximize intra-temporal household consumption. The private sector worker problem can then be written as the maximization of worker household consumption $c_{t}^{w}$ in each period t subject to the constraints

\begin{matrix} c_{t}^{w} = [(1 - τ^{L}) w_{t}^{d} l_{t}^{d} + φ w^{f} l_{t}^{f}] (1 - τ^{c}) & a n d & l_{t}^{d} + l_{t}^{f} = l \end{matrix}

The household labor endowment lcan be allocated to the domestic labor market $l_{t}^{d} \leq l$ and earn the net wage $(1 - τ^{L}) w_{t}^{d}$ , where τ^L is the tax rate on labor income, or work in foreign labor markets by allocating $l_{t}^{f} \leq l$ earn the foreign wage w^f, and remit the share φ ∈ [0,1] to the household. It is assumed labor is mobile internationally, implying that in equilibrium

w_{t}^{d} = φ \frac{w^{f}}{(1 - τ^{L})} = w^{d} .

Government Worker Household

Public sector workers are assumed to display hand-to-mouth behavior. It is assumed the public sector workers cannot migrate and can only work in the public sector earning wage w^g. This is a simplifying assumption that capture the ECCU empirical observation that public wages are set high relative to other sectors at a level that prevents workers in the public sector from choosing to work in the private sector or to migrate -except for the typically small portion of high skilled public workers. Under these assumptions, the entire labor endowment of the worker household l^g is allocated to the government sector and used for domestic consumption intra-temporally, net of taxes,

(1 - τ^{c}) [(1 - τ^{L}) w^{g} l^{g} + T^{g}] = c_{t}^{w g} .

Government

The government supplies public capital, of types resilient and non-resilient to NDs, financed with taxes, nontax revenues, and debt issuance. The government does not follow any specific fiscal policy rule nor optimization decision process. It is assumed to have a passive fiscal policy stance: tax rates and expenditures are maintained constant. This assumption allows treatment of the parameters in the government budget constraint as policy variables –in particular, level and composition of public investment, as needed to simulate a change in investment decisions towards resilient structures.

Public debt is issued to external creditors. This captures the empirical observation that most public debt financing in ECCU countries is external, given narrow and illiquid domestic markets for sovereign debt instruments. Public debt evolves according to the identity

B_{t} = (1 + r) B_{t - 1} + G_{t} - R_{t}

where B_t is the sock of public debt; r is the interest rate, assumed to remain constant. Primary expenditures G_t and revenues R_t are determined by

\begin{array}{l} G_{t} = C_{t}^{g} + w^{g} L_{t}^{g} + T_{t} + l_{t}^{n} + p^{r} I_{t}^{r} \\ R_{t} = τ^{c} C_{t} + τ^{L} (w_{t}^{k} L_{t}^{g} + L_{t}^{g} w^{g}) + τ^{k} r_{t}^{k} + N T_{t} \end{array}

where $C_{t}^{g}$ is government consumption; $L_{t}^{g}$ is public sector employment; $T_{t} = T_{t}^{g} + T_{t}^{L} + T_{t}^{k}$ is transfers to government worker and private worker households, respectively; and $I_{t}^{n} + p^{r} I_{t}^{r}$ is expenditure in public investment The composition of the stock of public capital Kg = Kⁿ + K^r is thus a policy decision, which depends on public investment allocation I^g = Iⁿ + I^r. I^r is assumed to be costlier than In, with a price p^r > 1 (tradable output price normalized to be equal to 1). This implies that a change in the composition of public investment towards the resilient type increases capital expenditures, even if the amount of physical investment remains unchanged. With τ denoting tax rates, government revenues are determined by consumption taxes τ^cCt (aggregate consumption of investors, private and public workers, and the government), labor taxes on domestic private and public sectors’ labor $τ^{L} (L_{t}^{d} + L_{t}^{g})$ ; taxes on capital returns $r^{k} r_{t}^{k}$ ; and non-tax revenues NT_t -mainly donor grants and CBI revenues.

As indicated above, NDs shocks also destroy a share of public capital. It is assumed that non- resilient public capital is subject to depreciation shocks in periods with NDs, which are assumed to be randomly distributed to capture NDs’ intensity and frequency. The evolution of public capital stock follows the following laws of motion

\begin{array}{l} K_{t}^{r} = K_{t - 1}^{r} (1 - δ) + I_{t}^{r} \\ K_{t}^{n} = (1 - D_{t}) K_{t - 1}^{n} (1 - δ) + D_{t} \int K_{t - 1}^{n} (1 - δ - δ_{t}^{D}) f (δ_{t}^{D}) ⅆ δ_{t}^{D} + I_{t}^{n} . \end{array}

Expected aggregate output is increasing in the stock of resilient public investment. Economy-wide expected output is thus given by

E_{t - 1} Y_{t} = (1 - P) A_{t} {(K_{t - 1}^{r} + K_{t - 1}^{n})}^{α g} {K_{t - 1}}^{α k} L_{t}^{d^{β}} + P A_{t} θ_{t} \int [K_{t - 1}^{r} + K_{t - 1}^{n} (1 - δ_{t}^{D})]^{α g} [K_{t - 1} (1 - δ_{t}^{D})]^{α k} f (δ_{t}^{D}) ⅆ δ_{t}^{D}

where E_t-1 is the expectations operator as of period t-1. K_t-1 and $L_{t}^{d}$ are aggregate investor household capital and equilibrium domestic labor utilization, respectively. With probability 1 – P there is no ND and output can be produced with the full amount of capital invested before period t, and with probability P there is a ND which destroys non-resilient public capital and private capital by a share $δ_{t}^{D}$ . As labor is not predetermined, the share of labor employed domestically $L_{t}^{d}$ takes different values depending on the occurrence of a ND, which adjusts with labor migration.

Inspection of the expected output equation indicates that marginal productivities of private capital and labor are decreasing in the share of non-resilient public capital. This is determined by the presence of the $(1 - δ_{t}^{D})$ affecting $K_{t - 1}^{n}$ in the second term. Therefore, private capital and employment are high when resilient public capital is also high. As resilient public capital is resistant to ND shocks, expected private output loss is smaller for given ND frequency 1/P and intensity distribution $f (δ_{t}^{D})$ . As a result, expected private investment returns and thus private capital and output are higher with more resilient public investment, for given stock of total public investment.

Appendix II. Parameter Calibration

Source: Fund staff calculations based on data from authorities’ data.

	ATG	DMA	GRD	LCA	KNA	VCT
Real Sector
Annual probability of natural disaster	0.150	0.200	0.200	0.200	0.250	0.200
Semi-elasticity private capital	0.350	0.350	0.350	0.350	0.350	0.350
Semi-elasticity public capital	0.083	0.100	0.097	0.088	0.052	0.061
Labor force	53.082	45.064	76.467	72.401	42.420	46.625
Labor force, private sector	32.000	23.000	45.600	45.800	18.000	24.300
Labor force, working abroad (migration)	13.082	15.064	24.867	15.101	12.420	12.725
Foreign remittances / GDP	0.006	0.100	0.030	0.015	0.013	0.045
Share of labor migration wage transferred to domestic household	0.014	0.075	0.049	0.034	0.020	0.055
Labor force, public sector	8.000	7.000	6.000	11.500	12.000	9.600
Emigration rate, share of labor force abroad	0.327	0.502	0.482	0.264	0.414	0.375
Private capital depreciation rate	0.044	0.044	0.044	0.044	0.044	0.044
Non-resilient public capital depreciation rate	0.056	0.056	0.056	0.056	0.056	0.056
Resilient public capital depreciation rate	0.056	0.056	0.056	0.056	0.056	0.056
Capital depreciation in average natural disaster (disaster damage)	0.082	0.052	0.025	0.021	0.070	0.026
Government Sector
Implicit interest rate public debt	0.040	0.035	0.048	0.048	0.027	0.043
Government revenues / GDP	0.208	0.305	0.252	0.223	0.256	0.294
Government Operations: taxes on goods and services + taxes on imports / GDP	0.074	0.175	0.078	0.156	0.147	0.180
Government Operations: taxes on property + corporate taxes / GDP	0.015	0.000	0.017	0.020	0.007	0.020
Government Operations: taxes on wages (taxes on income – corporate taxes) / GDP	0.010	0.050	0.112	0.023	0.052	0.070
Government Operations: non-tax revenues (grants, CBI, other) / GDP	0.109	0.080	0.044	0.026	0.050	0.024
Primary expenditures / GDP	0.227	0.295	0.205	0.245	0.281	0.287
Public sector employment, share of labor force	0.151	0.155	0.078	0.159	0.189	0.206
Government Operations: compensation employees	0.080	0.100	0.085	0.085	0.115	0.132
Total public investment expenditure share to GDP	0.030	0.075	0.060	0.047	0.048	0.039
Government Operations: public investment expenditure share to GDP, non resilient	0.030	0.075	0.060	0.047	0.048	0.039
Government Operations: public investment expenditure share to GDP, resilient	0.000	0.000	0.000	0.000	0.000	0.000
Share of resilient public capital in total public capital	0.000	0.000	0.000	0.000	0.000	0.000
Price resilient public capital relative to non resilient	1.250	1.250	1.250	1.250	1.250	1.250
Government Operations: Transfers and subsidies, including pensions	0.062	0.050	0.020	0.067	0.058	0.080
Government Operations: pensions and social transfers to HH	0.040	0.031	0.014	0.043	0.032	0.046
Government Operations: subsidies to productive sector	0.012	0.010	0.004	0.013	0.012	0.016
Government Operations: pensions and social transfers to public workers	0.010	0.009	0.002	0.011	0.014	0.018
From fiscal accounts: Government consumption (purchases of goods and services)	0.055	0.070	0.040	0.046	0.060	0.036
Public debt to GDP ratio	0.868	0.750	0.710	0.710	0.620	0.780
Target public debt to GDP ratio	0.600	0.600	0.600	0.600	0.600	0.600
Implicit tax rate: actual collection concumption tax / total consumption	0.066	0.150	0.098	0.150	0.150	0.150
Implicit tax rate: actual collection corporate tax / corporate net income	0.015	0.000	0.017	0.020	0.007	0.020
Implicit tax rate: actual collection payroll tax / wages paid	0.127	0.499	1.319	0.268	0.452	0.530
Ratio government non-resilient investment to total public investment	1.000	1.000	1.000	1.000	1.000	1.000
Ratio government resilient investment to total public investment	0.000	0.000	0.000	0.000	0.000	0.000
Depreciation rate non-resilient capital	0.056	0.056	0.056	0.056	0.056	0.056

Source: Fund staff calculations based on data from authorities’ data.

	ATG	DMA	GRD	LCA	KNA	VCT
Real Sector
Annual probability of natural disaster	0.150	0.200	0.200	0.200	0.250	0.200
Semi-elasticity private capital	0.350	0.350	0.350	0.350	0.350	0.350
Semi-elasticity public capital	0.083	0.100	0.097	0.088	0.052	0.061
Labor force	53.082	45.064	76.467	72.401	42.420	46.625
Labor force, private sector	32.000	23.000	45.600	45.800	18.000	24.300
Labor force, working abroad (migration)	13.082	15.064	24.867	15.101	12.420	12.725
Foreign remittances / GDP	0.006	0.100	0.030	0.015	0.013	0.045
Share of labor migration wage transferred to domestic household	0.014	0.075	0.049	0.034	0.020	0.055
Labor force, public sector	8.000	7.000	6.000	11.500	12.000	9.600
Emigration rate, share of labor force abroad	0.327	0.502	0.482	0.264	0.414	0.375
Private capital depreciation rate	0.044	0.044	0.044	0.044	0.044	0.044
Non-resilient public capital depreciation rate	0.056	0.056	0.056	0.056	0.056	0.056
Resilient public capital depreciation rate	0.056	0.056	0.056	0.056	0.056	0.056
Capital depreciation in average natural disaster (disaster damage)	0.082	0.052	0.025	0.021	0.070	0.026
Government Sector
Implicit interest rate public debt	0.040	0.035	0.048	0.048	0.027	0.043
Government revenues / GDP	0.208	0.305	0.252	0.223	0.256	0.294
Government Operations: taxes on goods and services + taxes on imports / GDP	0.074	0.175	0.078	0.156	0.147	0.180
Government Operations: taxes on property + corporate taxes / GDP	0.015	0.000	0.017	0.020	0.007	0.020
Government Operations: taxes on wages (taxes on income – corporate taxes) / GDP	0.010	0.050	0.112	0.023	0.052	0.070
Government Operations: non-tax revenues (grants, CBI, other) / GDP	0.109	0.080	0.044	0.026	0.050	0.024
Primary expenditures / GDP	0.227	0.295	0.205	0.245	0.281	0.287
Public sector employment, share of labor force	0.151	0.155	0.078	0.159	0.189	0.206
Government Operations: compensation employees	0.080	0.100	0.085	0.085	0.115	0.132
Total public investment expenditure share to GDP	0.030	0.075	0.060	0.047	0.048	0.039
Government Operations: public investment expenditure share to GDP, non resilient	0.030	0.075	0.060	0.047	0.048	0.039
Government Operations: public investment expenditure share to GDP, resilient	0.000	0.000	0.000	0.000	0.000	0.000
Share of resilient public capital in total public capital	0.000	0.000	0.000	0.000	0.000	0.000
Price resilient public capital relative to non resilient	1.250	1.250	1.250	1.250	1.250	1.250
Government Operations: Transfers and subsidies, including pensions	0.062	0.050	0.020	0.067	0.058	0.080
Government Operations: pensions and social transfers to HH	0.040	0.031	0.014	0.043	0.032	0.046
Government Operations: subsidies to productive sector	0.012	0.010	0.004	0.013	0.012	0.016
Government Operations: pensions and social transfers to public workers	0.010	0.009	0.002	0.011	0.014	0.018
From fiscal accounts: Government consumption (purchases of goods and services)	0.055	0.070	0.040	0.046	0.060	0.036
Public debt to GDP ratio	0.868	0.750	0.710	0.710	0.620	0.780
Target public debt to GDP ratio	0.600	0.600	0.600	0.600	0.600	0.600
Implicit tax rate: actual collection concumption tax / total consumption	0.066	0.150	0.098	0.150	0.150	0.150
Implicit tax rate: actual collection corporate tax / corporate net income	0.015	0.000	0.017	0.020	0.007	0.020
Implicit tax rate: actual collection payroll tax / wages paid	0.127	0.499	1.319	0.268	0.452	0.530
Ratio government non-resilient investment to total public investment	1.000	1.000	1.000	1.000	1.000	1.000
Ratio government resilient investment to total public investment	0.000	0.000	0.000	0.000	0.000	0.000
Depreciation rate non-resilient capital	0.056	0.056	0.056	0.056	0.056	0.056

Source: Fund staff calculations based on data from authorities’ data.

Appendix III. Insurance Simulation Methodology

The starting point is to estimate an empirical model for each economy that captures the effect of ND on output and government finances. To this end, an unrestricted Vector Auto-regression Model (VAR) is estimated for each country including fiscal determinants of public debt dynamics, including vectors Y_t and X_t of endogenous and exogenous variables, respectively,

Y_{t} = γ_{0} + Σ_{k = 1}^{p} γ_{k} Y_{t - k} + Σ_{j = 1}^{n} β_{j} X_{j} + ϵ_{t} .

The endogenous variables in the VAR estimates include the cyclical components of GDP; government revenues excluding grants; grants; current primary expenditures; and capital expenditures. These are expressed as a share of each indicators’ trend¹,

{\hat{y}}_{i t} = y_{i t} / y_{i t}^{t r e n d}; {\hat{y}}_{i t} \in Y_{t} .

The exogenous control variables X_j account for non-ND sources of major shocks in ECCU countries, and thus estimated residuals ε_t~N(0,Ω) are orthogonal to these. The vector of control variables X_j includes the U.S. real effective exchange (to capture competitiveness pressures given that the EC dollar is pegged to the U.S. dollar); the oil price (all countries are highly dependent on oil imports); the cyclical component of the U.S. output (the main source of tourist revenues); and a dummy for the September 11, 2001 terrorist attack in the United States that disrupted air travel and tourism exports. This allows the identification of ND shocks as the only potentially large shock remaining -ε_t thus includes ND and “small-shocks.² The variance-covariance matrix Ω characterizes the joint statistical properties of the contemporaneous disturbances of the endogenous variables. γ_k and β_j are vectors of coefficients.

The second step is to run a Monte-Carlo experiment. This involves generating many simulations using the estimated model above. Each simulation is shocked with a sequence of random vectors ${\hat{ϵ}}_{t + 1}, ..., {\hat{ϵ}}_{T}$ such that ∀ τ ∈ [t + 1, T], ε_τ = Wν_τ, where ν_τ~N(0,1), and W is such that Ω = W’W where W is the Choleski factorization of Ω. The estimated VAR is then used to generate 2000 forecasts Y_t for each country with the randomly-generated shocks ε_τ,τ = t + 1, …,T. In this way, the VAR produces joint dynamic responses of all variables in Y_t.³ Each simulation is a projection consisting of a sequence of the five endogenous variables in the model, each affected by a sequence of simulated random shocks. In this way, the simulations mimic historical patterns in terms of the volatility, persistence, and co-movement of the endogenous series, as observed in the sample data. The results are then used to compute probability density functions for each of the five endogenous variables in each year projected, for the period 2017–2030.

These simulations can then be used to calculate public debt dynamics for each random simulation with the debt accumulation identity

D_{t + l + 1} = (1 + i_{t + l + 1}) D_{t + l} - P B_{t + l} + (I_{t + l}^{S F} - O_{t + l}^{S F}); D_{t + l} > 0; l = t + 1, ..., T;

where D_t+l+1 is the stock of public debt in year t+l+1, and PB_t is the primary balance obtained from the revenue and primary expenditure endogenous variables in the simulations. The implicit interest rate i_t is calculated as the ratio of interest expenditures in year t divided by public debt stock in t-1. $I_{t + l}^{S F} - O_{t + l}^{S F}$ are the below-the-line inflows and outflows from the SF for NDs. Depending on the sequence of events (occurrence or non-occurrence of a ND in any given year in the simulations), different debt paths are thus possible, as these flows vis-à-vis the budget replace debt issuance. Notice that the debt stock projections are not affected in expected terms, provided savings into the fund are utilized in the long-term and across simulations in a given period.⁴

Given that these projections are obtained as deviations from trend, they are then calculated as a percent of GDP. To that end, a deterministic trend is projected for each variable, assuming each and all trends grow at the same constant rate starting from the end point of the estimated trend in the sample period, that is

y_{i t + l} = {\hat{y}}_{i t + l} y_{i t + l}^{t r e n d} w i t h y_{i t + l}^{t r e n d} = y_{y t} {(1 + g)}^{l - t}; l = t + 1, ..., T;

where g is the potential growth rate assumption. After all endogenous variables are expressed in real-term levels, they can be expressed in percent of GDP by dividing each of the fiscal indicator projections by the GDP projection in each of the vector simulations. In order to ensure revenue and expenditure indicators as a percent of GDP are consistent with the data, the starting points of the trend projections are set at constant prices of the last year of the sample.⁵

The third step is to identify the occurrence of NDs in each simulation, as needed to inform the triggering of financing flows vis-à-vis the SFs. To this end, the simulations include an algorithm that identifies as a ND the largest d percent fiscal deteriorations. The fiscal deteriorations are computed as the sum of the year-on-year changes of (i) non-grant revenue NG_t (with a negative sign as tax revenues decline along with output as affected by the ND); (ii) grant revenues Gr_t (which typically increase after ND as donor partners increase their supports) (iii) current primary expenditure G_t (as more social assistance and goods and services are needed after NDs); and (iv) capital expenditure K_t (on account of additional expenditures for rehabilitation and reconstruction). The fiscal aggregate random variable z_tεZ used to identify simulated NDs, can be written as

z_{t} = Δ G r_{t} - Δ N G_{t} + Δ G C E_{t} + Δ G K E_{t}

Simulated NDs are identified as the largest random draw realizations of the random variable z_t. The algorithm computes the distribution of this sum in every simulation year t+l, l=t+1,…T, across the 2000 simulations, and identifies as a ND all the random realizations that fall in the highest d percent tail of the of the probability density function of this sum. In this way if, statistically in a given simulation, non-grant revenues decline significantly, and grant revenues, current primary expenditures, and capital expenditures increase significantly (a typical pattern after a ND), then that random simulation draw is labeled as a ND by the algorithm. As mentioned above, this identification rests on the assumption that all other major sources of large shocks have been controlled for in the VARs, and thus every remaining negative large shock is a ND.

The fourth step is to specify SFs financing flows vis-à-vis the budget. The simulations assume that in years with no ND (as identified by the algorithm explained above), the budget contributes savings to the SF.⁶ These budget contributions to the SF are modeled as a fixed parameter θ as a percent of the previous’ year GDP, with θ calibrated to achieve the financial sustainability of the Fund with a sufficiently low probability of depletion or, in other words, to ensure the SF stock is stable in expected terms⁷.

In the event a ND occurs, as identified by the algorithm, a financing inflow to the budget from the SF takes place. This budget financing is computed as the sum of four components:

+ Gap of non-grant revenues below trend. Captures the decline in tax and non-tax revenues that typically take place after NDs as a result of a typical decline in economic activity and tax compliance.
- Gap of grant revenues above trend. Grants tend to be higher after NDs as a result of an increase in donor support, reducing the need for financing flows from the Fund.
+ Gap of current primary expenditure above trend. Captures higher expenditures in social support and rehabilitation of infrastructure after NDs. An additional fixed amount as a percent of GDP is deducted to capture below-trend spending reprioritization. The reprioritization below trend is denoted $ρ^{G} G_{t + l}^{t r e n d}; ρ^{G} \in (- 1, 1)$ .
+ Gap of capital expenditure above trend. Captures the higher public investment that typically follows NDs with the reconstruction spending. An additional fixed amount as a percent of GDP is deducted to capture below-trend spending reprioritization. The reprioritization below trend is denoted $ρ^{K} K_{t + l}^{t r e n d}; ρ^{K} \in (- 1, 1)$ .

Denote the random variable obtained from the sum of the four components above as

σ_{t} = - (N G_{t} - N G_{t}^{t r e n d}) - (G r_{t} - G r_{t}^{t r e n d}) + (G_{t} - G_{t}^{t r e n d} - ρ^{G} G_{t}^{t r e n d}) + (K_{t + l} - K_{t}^{t r e n d} - ρ^{K} K_{t}^{t r e n d}) .

The contributions to the budget continue until the year in which each indicator returns to a level that is below the value in the year prior to the ND -the SF therefore finances the “hump”.

Allowing for expenditure re-prioritization in the simulations is key for a realistic assessment of SFs’ size. In practice, a significant share of the fiscal space for social support and reconstruction after ND is obtained by way of reallocation and re-prioritization: some pre-ND allocations are postponed or cancelled. The text chart illustrates re-prioritization in the case of Dominica, after it was affected by Tropical storm Erika in August 2015. This implies that the reconstruction expenditures do not require an equivalent increase in public investment relative to the original investment levels without a ND. This is the reason for the reprioritization terms capturing expenditures that are postponed or cancelled, as explained above for current primary and capital spending.

The SF stock in each simulation evolves following the difference of saving inflows $I_{t + l}^{S F}$ and outflows $O_{t + l}^{S F}$ vis-à-vis the SF,

S F_{t + 1} = S F_{t} (1 + r_{t}) + I_{t + 1}^{S F} - O_{t + 1}^{S F} = S F_{t} (1 + r_{t}) + (1 - N D_{t}) θ G D P_{t - 1} - N D_{t} σ_{t},

with θGDP_t+l-1 = 0 in years in which the algorithm identifies a ND (ND_t = 1), with probability $P [d] = P [σ_{t} \geq \bar{σ}] = p$ and σ_t+l in year with no ND as identified by the algorithm, with probability 1-P[d] (ND_t = 0) . As explained above, $\bar{σ}$ is a parameter specified in the calibration when setting the annual probability p of a ND, as per the frequency observed historically. For example, if a ND occurs every 5 years on average, then the calibration requires to set p = 0.2, which then informs the value of the threshold $\bar{σ}$ as per the estimated probability density function in the distribution of σ_t. r_t is the rate of return of assets in the SF. If in a simulation the SF is depleted, SF_t= 0, the simulations assume that the remaining financing is covered with public debt issuance.

The modeling of the SF also includes an assumption for the initial stock value, the start-up cost. This initial amount of assets SF₀ > 0 affects the probability of depletion over a time horizon. As the proposal assumes that the start-up cost of establishing a SF is funded with existing CBI assets, it has not been added to the debt stock at the beginning of the simulations.

Insurance Calibration

The simulation parameters are calibrated consistent with staff’s macroeconomic frameworks for each ECCU country. The macroeconomic and fiscal parameters for calibration include potential GDP growth rate; the implicit interest rate on public debt; fiscal consolidation targets in percent of GDP. Appendix I shows the specific parametric calibrations used in the simulations for each country.

The parameters affecting the SF are calibrated to cover 95 percent of the fiscal deteriorations after NDs. The calibration ensures the SF long-term financial sustainability with a low probability of depletion. For example, in the case of Dominica, the probability of a ND was set at P[d]=0.2, broadly consistent with the historical frequency of ND occurring every 5 years on average. The initial size of the Fund stock is set at 10 percent of GDP, as needed to obtain a probability of depletion within the next ten years of 0.08. Notice this probability of depletion determines insurance coverage, and it is to be chosen also consistent with risk tolerance of the authorities, although a sufficiently low probability of depletion is preferable to ensure the financial sustainability of the SF. In the simulations it is assumed that the probability of depletion is 5 percent -only in the most extreme 5 percent NDs’ fiscal costs the SF savings are insufficient and is depleted. Budget saving flows into the SF in years without a ND of 1.5 percent of previous-year’s GDP are needed to obtain a sustainable SF stock of assets in expected terms.

The remaining parameters for calibration specify the amount of SF financing to the budget after a ND, including for spending re-prioritization. To this end, “base” levels of capital expenditures and current primary expenditures are calibrated, with “base” defined as the level of spending that would prevail in a year in which there is no spending associated with the occurrence of a ND. This is captured by the parameters ρ^G and ρ^K, which need to be set carefully to ensure realistic expenditure reprioritization.

CCRIF and state-contingent debt are then calibrated to cover fiscal deteriorations post NDs to top up self-insurance coverage. These instruments are introduced as follows:

CCRIF. It is assumed the CCRIF insurance premium P^CCRIF is determined by the standard insurance formula

p^CCRIF = aAAL + b SD(AAL)

where a and b are fixed parameters calibrated to match observed premia, consistent with insurance multipliers in the range 1.5–2.0. AAL and SD(AAU) are NDs average annual losses and standard deviation, respectively, set at values reported by CCRIF based on estimated loss functions’ damages for tropical cyclones and earthquakes. In each country’s simulation it is assumed that the attachment point and maximum coverage are set at the estimated damages along the expected loss function for 20 and 100 year estimated damages, respectively. Maximum coverage under CCRIF is thus the difference between maximum coverage and attachment point. CCRIF payouts are triggered according with the ND algorithm explained above. The payout amount is determined as a proportion of simulated losses. In addition, in light of imperfect correlation between parametric triggers under CCRIF and actual damages, it is assumed that CCRIF payouts are discounted by a factor of 0.5 for 1/20 year losses based on insurance multipliers of near 1 for low coverage (that is, payouts turn out to be ½ of losses on average for the smaller NDs for which CCRIF is triggered). For larger and less frequent NDs, it is assumed that the correlation increases at a constant rate until convergence to a value of 1 for 1/100 year loss (the payout is proportional to the loss covered). These assumptions result CCRIF disbursements that are increasing in the simulated intensity of NDs. The correlation between triggers and payouts of 0.5 is set according to CCRIF estimates of insurance multipliers of about 1 for relatively smaller and more frequent NDs (used for 1/20 year NDs), and about 2 for large NDs (1/100 year NDs).

State-contingent debt. Governments issue catastrophe (CAT) bonds for debt service relief in case of a ND. The proceeds are invested in safe asset, with returns equivalent to Libor, and held in a Special Purpose Vehicle (SPV). Governments hold a call option on the principal of the SPV with triggers specified in the bond contract. If ND occurs, governments can withdraw funds from SPV to pay claims, and interest and principal payments are forgiven. If ND does not occur, investors receive principal and interest. It is assumed CAT bonds are issued with 3-year maturity and interest rate equivalent to Libor plus 500 basis points. The amount of CAT bond issuance is assumed to be determined by the residual coverage needs to top up self-insurance and maximum CCRIF coverage. If in the simulations maximum insurance under CCRIF remains insufficient to reach 99 percent overall coverage of NDs’ fiscal costs, the simulations assume that countries issue US$ 10 to 100 million in CAT bonds to reach the 99 percent target.

The simulations with resilience repeat the exercise under no resilience adjusting the fiscal model parameters with the results in Section B. This is done by incorporating the results in Section B to the long-term benefits of resilient public investment on output, tax revenues, and fiscal costs of resilience to the Monte Carlo experiment. To ensure comparability, the experiment maintains the illustrative 80 percent resilience assumption of Section B, and also the coverage levels of 95 percent coverage of NDs fiscal costs with a SF and up to 99 percent with insurance. The adjustments to the Monte Carlo experiment are as follows:

One-time increase in potential GDP (trend) by the growth rate in Section B for 80 percent resilience.
One-time increase in tax revenue trend by the estimated amount in Section B for 80 percent resilience.
Primary current expenditures remain unchanged in real terms, that is, at the same levels simulated without resilience.
Capital expenditures are increased by multiplying the simulated levels without resilience by a factor 0.8 x 1.25 to capture the additional cost of resilience, at a price 25 percent more expensive than non-resilient investment, applied to 80 percent of spending.
The algorithm used to identify NDs is the same as with no resilience. The simulated NDs frequency and intensity are therefore of exactly same distribution as in the simulations without resilience.
Outflows of the SF to the budget in each simulated ND is adjusted by applying a factor of 0.2, which captures the fact that fiscal costs are 80 percent lower than without resilience. The SF initial stock and annual saving inflows are adjusted down as appropriate to obtain a probability of depletion of 95 percent with financial sustainability.
CCRIF attachment point and coverage limit assumptions are same as with no resilience. CCRIF is purchased to achieve coverage of 99 percent of the NDs’ fiscal costs –the disbursements from the SF plus the CCRIF payout cover the full increase of the NDs annual fiscal costs in 99 percent of the events, and in the remaining 1 percent are insufficient.
CAT bonds are issued if CCRIF is not enough to cover 99 percent of NDs’ fiscal costs if the coverage with the SFs and CCRIF are insufficient. Assumptions are the same as without resilience.

Prepared by Alejandro Guerson (WHD).

This implies, for example, that a road or bridge resilient to natural disasters provides the same service as one that is non-resilient. Depreciation rates of both types of capital are assumed to be the same, except when affected by a natural disaster. In this case, non-resilient capital suffers a depreciation shock.

The assumption that resilient capital is completely unaffected by the natural disaster seems extreme considering real-life events, where even resilient structures can be damaged albeit by a lesser extent. This assumption simplifies the model solution, and it is not critical for the generality of the results. For example, the model could be calibrated so that the share of resilient vs. non-resilient capital matches the expected combined destruction of the aggregate capital stock, including of resilient structures.

The model assumes that public sector workers cannot migrate or work in the private sector. This is a simplifying assumption to capture the fact that public sector wages are significantly higher than in most other sectors in ECCU countries and are not set competitively. As public workers do not contribute to output in the model, any assumption with regards their migration decisions are not relevant to the results.

Given no availability of AAL estimates for Grenada, the St. Lucia natural disaster depreciation rate calibration is used as a proxy, which is the ECCU country with closest income per capita.

This is a simplifying assumption to capture a low initial level of resilience. Some ECCU countries have started with some resilient public investment, but the process is still at an early stage.

Average out-migration rates are about 40 percent on average, and near 80 percent for high-skilled workers, as reported by data estimates from the United Nations.

In Grenada these forces balance out resulting in no long-term improvement in fiscal performance.

The layering framework is in line with World Bank best practice recommendations. See for example “Sovereign Climate and Disaster Risk Pooling”, 2017, International Bank for Reconstruction and Development.

In the case of St. Vincent and the Grenadines, with no CBI program, a SF for ND may require debt issuance or a period of increased fiscal savings.

The attachment point and maximum coverage are approximated values based on discussions with governments.

Proceeds of CAT bond issuance are assumed to be invested in risk free liquid assets yielding international rates. As a result, only the net cost of CAT bond issuance, the insurance spread, is included in the simulations.

The cost-effective calibration of triggering points across insurance layers can be difficult given imperfect correlation of CCRIF and state contingent bonds’ parametric triggers with natural disaster damages, and significant standard deviation of expected damages and losses.

The assumption that resilient capital is completely unaffected by the natural disaster seems extreme considering real-life events, where even resilient structures can be damaged albeit by a lesser extent. This assumption simplifies the model solution, and it is not critical for the results. For example, the model could be calibrated so that the share of resilient vs. non-resilient capital matches the expected combined destruction of the aggregate capital stock, including of resilient structures.

The cyclical components of GDP are estimated using the Hodrick-Prescott filter on 1990–2016 annual data. All variables expressed in real terms using the GDP deflator.

The sample data used in the estimation spans 1990–2016.

Notice that the results are not sensitive to the ordering of the variables in the VAR, as the stochastic simulation results are shaped per the variance-covariance matrix of reduced-form errors n, which is unique.

Forward iteration on the debt accumulation identity shows that, in expected terms, $I_{t + l}^{S F} - O_{t + l}^{S F}$ if SF’s inflows and outflows are calibrated to be zero in expected terms, as set in the simulations by construction.

It is therefore implicitly assumed that the deflators of GDP and the remaining fiscal variables change at the same rate in the projections.

If the simulations result in a fiscal deficit, then there would be a need to issue public debt to finance the required contribution to the Fund.

This ensures that the saving rate is of an amount commensurate with the fiscal costs of ND. If saving inflows into the SF are set too high, then the size of the SF would tend to increase in expected terms, accumulating excess assets inefficiently. If, on the other hand, the saving flow is set too low, the SF would unsustainably decline in expected terms towards depletion.

References

Bova E., Carcenac N., and M. Guerguil (2014) “Fiscal Rules and Pro-cyclicality of Fiscal Policy in the Developing World,” IMF Working paper, 14/122.
- Search Google Scholar
- Export Citation
CDB (2014) Public-Private Partnerships in the Caribbean: Building on Early Lessons, Caribbean Development Bank
- Search Google Scholar
- Export Citation
Greenidge K., Roland C., Thomas C., and L. Drakes (2012) “Threshold Effects of Sovereign Debt: Evidence from the Caribbean,” IMF Working Paper, 12/157.
- Search Google Scholar
- Export Citation
IMF (2009) “Fiscal Rules—Anchoring Expectations for Sustainable Public Finances,” IMF Policy Paper (Washington, DC: International Monetary Fund)
- Search Google Scholar
- Export Citation
IMF (2012) Fiscal Rules in Response to the Crisis—Toward the “Next-Generation” Rules. A New Dataset Schaechter A., Kinda T., Budina N. and A. Weber
- Search Google Scholar
- Export Citation
IMF (2016) “Analyzing and Managing Fiscal Risks: Best Practices,” IMF Policy Paper (Washington, DC: International Monetary Fund)
- Search Google Scholar
- Export Citation
IMF (2017a) “Fiscal Rules for the ECCU,” IMF Country Report 17/151 (Washington, DC: International Monetary Fund)
- Search Google Scholar
- Export Citation
IMF 2017b) Lledo V., Dudine P., Eyraud L., and Peralta, 2017 “How to Select Fiscal Rules? A Primer,” IMF How-To Note, December.
- Search Google Scholar
- Export Citation
IMF (2017c) Baum, Eyraud, Hodge, Jarmuzek, Kim, Mbaye, and Ture, 2017, “How to Calibrate Fiscal Rules? A Primer,” IMF How-to Note, December.
- Search Google Scholar
- Export Citation
IMF (2017d) “Sixth Review under the Extended Credit Facility (ECF) Arrangement and Financing Assurances Review” IMF Country report No 17/131.
- Search Google Scholar
- Export Citation
IMF (2018a) SDN 18/04 “Second-Generation Fiscal Rules: Balancing Credibility, Flexibility, and Simplicity.”
- Search Google Scholar
- Export Citation
IMF (2018b) “Staff Guidance Note on the Fund’s Engagement with Small Developing States,” January.
- Search Google Scholar
- Export Citation
IMF (2018c) “Jamaica: Staff Report for the 2018 Article IV Consultation Third Review Under the Stand-By Arrangement and Request for Modification of Performance Criteria,” IMF Country report No 18/103.
- Search Google Scholar
- Export Citation
IMF (2018d) “Grenada: Staff Report for the 2018 Article IV Consultation,” IMF Country report No 18/236.
- Search Google Scholar
- Export Citation
IMF (2018e) “How to Manage the Fiscal Costs of Natural Disasters,” IMF How to Note, 18/03.
- Search Google Scholar
- Export Citation
Lledó, V.D., S. Yoon, X. Fang, S. Mbaye, & Y. Kim (2017) Fiscal Rules at a Glance (Washington, DC: International Monetary Fund)
- Search Google Scholar
- Export Citation
Mitchell W., James R., and A.M. Wickham (2018) “Managing the Government Wage Bill and Civil Service Reform in ECCU Member Countries,” IMF Working paper, forthcoming.
- Search Google Scholar
- Export Citation
World Bank (2018) “FROM KNOWN UNKNOWNS TO BLACK SWANS” How to Manage Risk in Latin America and the Caribbean” Semiannual Report – Office of the Regional Chief Economist, October 2018
- Search Google Scholar
- Export Citation

Appendix Table: Analysis of ECCU-6 Fiscal Policies over the Cycle

Note: The table shows regression coefficients between cyclical components of real spending and real GDP, estimated through the HP filter with a smoolthing parameter lamda=100 (see Bova et al. 2014). Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Source: IMF staff estimates.

	1998–2017
	Cyclical Component of Real Spending
	Antigua and Barbuda	Dominica	Grenada	St Kitts and Nevis	St Lucia	St Vincent and the	ECCU
Cyclical Component of real GDP	0.350***	0.641**	0.105	0.268**	0.0164	0.261*	0.227***
Cyclical Component of real GDP	(0.095)	(0.231)	(0.110)	(0.096)	(0.136)	(0.126)	(0.067)
Constant	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Constant	(0.020)	(0.008)	(0.009)	(0.006)	(0.011)	(0.005)	(0.027)
Observations	20	20	20	20	20	20	20
R-squared	0.432	0.299	0.049	0.301	0.001	0.192	0.392
	1998–2007
	Cyclical Component of Real Spending
	Antigua and Barbuda	Dominica	Grenada	St Kitts and Nevis	St Lucia	St Vincent and the	ECCU
Cyclical Component of real GDP	0.280**	1.000***	0.039	0.479**	0.038	0.264**	0.174
Cyclical Component of real GDP	(0.087)	(0.267)	(0.247)	(0.153)	(0.185)	(0.093)	(0.100)
Constant	-0.003	0.003	0.005	-0.003	-0.004	-0.004	-0.012
Constant	(0.018)	(0.008)	(0.018)	(0.008)	(0.015)	(0.004)	(0.036)
Observations	10	10	10	10	10	10	10
R-squared	0.566	0.636	0.003	0.552	0.005	0.502	0.276
	2008–2017
	Cyclical Component of Real Spending
	Antigua and Barbuda	Dominica	Grenada	St Kitts and Nevis	St Lucia	St Vincent and the	ECCU
Cyclical Component of real GDP	0.420**	0.417	0.128	0.153	-0.0348	0.342	0.267**
Cyclical Component of real GDP	-0.176	-0.397	-0.088	-0.122	-0.230	-0.327	-0.097
Constant	0.006	0.003	-0.004	0.001	0.005	0.005	0.015
Constant	(0.038)	(0.014)	(0.008)	(0.008)	(0.017)	(0.009)	(0.043)
Observations	10	10	10	10	10	10	10
R-squared	0.415	0.121	0.209	0.164	0.003	0.120	0.485

	1998–2017
	Cyclical Component of Real Spending
	Antigua and Barbuda	Dominica	Grenada	St Kitts and Nevis	St Lucia	St Vincent and the	ECCU
Cyclical Component of real GDP	0.350***	0.641**	0.105	0.268**	0.0164	0.261*	0.227***
Cyclical Component of real GDP	(0.095)	(0.231)	(0.110)	(0.096)	(0.136)	(0.126)	(0.067)
Constant	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Constant	(0.020)	(0.008)	(0.009)	(0.006)	(0.011)	(0.005)	(0.027)
Observations	20	20	20	20	20	20	20
R-squared	0.432	0.299	0.049	0.301	0.001	0.192	0.392
	1998–2007
	Cyclical Component of Real Spending
	Antigua and Barbuda	Dominica	Grenada	St Kitts and Nevis	St Lucia	St Vincent and the	ECCU
Cyclical Component of real GDP	0.280**	1.000***	0.039	0.479**	0.038	0.264**	0.174
Cyclical Component of real GDP	(0.087)	(0.267)	(0.247)	(0.153)	(0.185)	(0.093)	(0.100)
Constant	-0.003	0.003	0.005	-0.003	-0.004	-0.004	-0.012
Constant	(0.018)	(0.008)	(0.018)	(0.008)	(0.015)	(0.004)	(0.036)
Observations	10	10	10	10	10	10	10
R-squared	0.566	0.636	0.003	0.552	0.005	0.502	0.276
	2008–2017
	Cyclical Component of Real Spending
	Antigua and Barbuda	Dominica	Grenada	St Kitts and Nevis	St Lucia	St Vincent and the	ECCU
Cyclical Component of real GDP	0.420**	0.417	0.128	0.153	-0.0348	0.342	0.267**
Cyclical Component of real GDP	-0.176	-0.397	-0.088	-0.122	-0.230	-0.327	-0.097
Constant	0.006	0.003	-0.004	0.001	0.005	0.005	0.015
Constant	(0.038)	(0.014)	(0.008)	(0.008)	(0.017)	(0.009)	(0.043)
Observations	10	10	10	10	10	10	10
R-squared	0.415	0.121	0.209	0.164	0.003	0.120	0.485

Appendix I. Tools Used to Evaluate the Fiscal Framework Design

IMF staff uses a range of tools to inform the revision of existing fiscal responsibility frameworks and assess the performance of alternative designs.¹

Counterfactual analysis. This approach amounts to “rewriting history” through a retrospective scenario whereby a multi-year numerical target is assumed to be introduced at some point in the past. The method analyzes how the government behavior and the economic indicators would have changed under different targets and what would be today’s outcomes. For instance, Andrle and others (2015) compare the performance of the FRFs based on the expenditure and structural balance rules (or targets) if they had been adopted in Italy and France since the early 2000s. For the ECCU countries, a counterfactual analysis was used to assess and compare the effects of the operational budget balance-based target and expenditure-based targets (see charts in paragraph 30), assuming that such frameworks would have been introduced in 2002 in all ECCU countries. For tractability, various simplifying assumptions needed to be used.²
Scenario analysis. This approach is forward-looking and simulates the effect of rules over the forecasting horizon. It was initially developed by Debrun and others (2008) for Israel and further expanded in IMF (2009). The effects of rules are simulated under various scenarios, including a baseline (which could be the IMF World Economic Outlook projections) and shock scenarios. For the ECCU countries, this type of scenario analysis was used to assess different public debt trajectories that include key recommended elements of fiscal frameworks such as fiscal adjustment and incorporation of resilient investment and insurance buffers (see chart in paragraph 49).
Stochastic simulations. The forward-looking performance of fiscal frameworks or rules can also be assessed in response to stochastic shocks. Instead of simulating ad hoc deterministic scenarios (like in the previous approach), the shocks are drawn from a distribution representing the past behavior of the data. This approach builds on the framework developed in Celasun and others (2007) and was applied to the United Kingdom (IMF, 2010a). Based on repeated simulations of random macroeconomic shocks, fan charts are derived representing the frequency distribution of the budgetary aggregates for each fiscal rule and year of projection. In the ECCU countries, an element of stochastic simulations was used in this paper to assess the needed buffer for a safe level of debt (see paragraph 28), as a partial exercise that is useful for informing the design of the fiscal frameworks.
Model-based selection. The most elaborate approach to assess the design of fiscal frameworks is to use a multi-country macroeconomic model (for instance, a medium-scale DSGE), which incorporates the intertemporal decisions of households and firms as well as the general equilibrium effects of fiscal frameworks, including on expectations. The general idea is to apply shocks to the model and analyze how the economy responds in the presence of alternative frameworks. Simulations can be conducted around the steady state of the model (as in IMF, 2009; and Andrle and others, 2015) or around a baseline forecast (as in IMF, 2012). Shocks are calibrated in an ad-hoc way or, preferably, on past data. For instance, IMF (2009) presents GIMF simulations performed for three stylized economies representing a small open advanced economy, a large open advanced economy, and a small open commodity-exporting economy. The shocks considered are a domestic demand shock, an exogenous fall in supply (productivity shock), and, for the commodity-exporting economy, an exogenous change in external demand for the commodity. Various rules are assessed by plotting the path of GDP, inflation, debt, deficit, tax revenues and interest expenditure in deviations from the steady state over a 15-year horizon. For ECCU countries, Chapter 1 in this SIP is an example of such model-based approach for some aspects that can inform the design of the fiscal framework.

Appendix II. Analysis of ECCU-6 Fiscal Performance

The ECCU countries’ heterogeneity can be tracked through an evolution of key measurable factors impacting debt sustainability in each country. These factors include (i) fiscal deficit positions (gross and net of CBI inflows) relative to those consistent with long-term debt-stabilizing levels (these show as lines in the respective country-specific charts below); (ii) economic growth performance both in terms of long-term trends and short-term fluctuations (with the latter shown as bars in the respective charts) and (iii) large natural disasters (shown in red color among the growth bars for the years in which the disasters occurred). These indicators are discussed as part of a holistic narrative for country-level achievements and shortcomings in regaining debt sustainability.¹

Countries on track to reach debt sustainability. Over the past two decades, both Grenada and St. Vincent and the Grenadines largely avoided an average fiscal deficit bias relative to debt-stabilizing levels, with periods of “underperformance” periodically alternating with those of “overperformance.” Additionally, both countries saw their fiscal position improve recently, with fiscal balances currently being above debt-stabilizing levels. Historically, however, both countries had to deal with fiscal pressures and shocks. There was a significant and protracted deterioration of Grenada’s fiscal position in the aftermath of 2004–06 natural disasters, ultimately necessitating a correction implemented during 2014–17 and supported by the new FRF. St. Vincent and the Grenadines had the good luck of avoiding large natural disasters so far this century while maintaining a prudent fiscal position on average despite a down-trend in economic growth.
Countries with “moderate” debt sustainability problems. Unlike the above two countries, both St. Lucia and St. Kitts and Nevis exhibited a noticeable deficit bias, as debt-stabilizing balances were only (and barely) achieved in their best fiscal years, but not on average (in the case of St. Kitts and Nevis, this refers to the fiscal balance net of CBI inflows). The deficit bias owes to different reasons in the two countries. In St. Lucia, the 2010 natural disaster coincided with a sustained deterioration of the country’s fiscal position, but, unlike in Grenada, the subsequent deficit correction that started in 2013 has been insufficient. St. Kitts and Nevis avoided large natural disasters this century, but has not strengthened underlying fiscal position in the face of surging CBI inflows. Still, in both countries as of 2017 the underlying fiscal positions were close to debt-stabilizing levels, pointing to relatively moderate adjustment need for achieving fiscal sustainability.²
Countries with large debt sustainability problems. Both Antigua and Barbuda and Dominica exhibited a pronounced average underlying deficit bias over 2001–17, and both countries were hit by multiple natural disasters recently that added to the fiscal imbalances. These factors played out differently by country. In Dominica, the fiscal position was close to the debt-stabilizing balance through 2008, but the global financial crisis and, subsequently, truly devastating disasters of 2015 and 2017 triggered a major deterioration in fiscal sustainability. By contrast, Antigua and Barbuda’s deficits were substantially and persistently higher than those ensuring debt-stabilizing levels in the earlier part of the period. The imbalances moderated in the second part of the period, despite the natural disasters. Still, the underlying fiscal balances were not sufficient to reach debt stabilizing levels, let alone exceed them to correct the previous imbalances.

Appendix III. Institutional Factors to Underpin the ECCU FRFs

The FRFs should rely on institutions, including legal frameworks and PFM procedures, to help comply with the targets and do so in an effective way. This would permit, for each country, to strike a right balance between the binding medium-term elements of the FRFs and scope for policy discretion that is needed to respond to shocks and other unforeseen events. In the ECCU, the key areas of institutional support for the FRFs concern:

Level of legislation. A formal legal architecture can make policy commitments more binding over a longer-term horizon and increase the costs of non-compliance. Thus, most countries enshrine fiscal responsibility in laws or statutory norms, and a few in their Constitutions (see IMF 2012).¹ A higher level of legislation would, other things equal, signal a longer-lived and more broad-based commitment to fiscal responsibility while constraining passage of legislation that is inconsistent with this goal.² For example, Grenada passed a stand-alone FRL while aligning with it elements of other fiscal legislation. However, the higher legal level by itself is not a magic bullet and would not work if enforcement and accountability procedures are weak.
Statistics, data, and accounting. There is a critical mass of indicators essential to monitor the fiscal targets. In this regard, priorities are reliable data on: (i) NFPS public debt for tracking the FRF’s anchor; (ii) fiscal deficits and their key components (e.g., interest bill) at different government levels, as well as CBI inflows for tracking the operational targets(s); (iii) fiscal risks such as contingent obligations, including from public private partnerships; (iv) any other fiscal and economic indicators that are essential for country-specific FRFs, including any supplementary targets such as the wage bill or public investment.³ ECCU countries have substantial data gaps, including on debt and deficits, that need to be fixed by improving data and harmonizing methodologies.
Fiscal projections. Reliable budget forecasts help implement the targets as they would minimize unwarranted deviations from the framework that could hurt its credibility. The ECCU countries’ own budget forecasts in recent years have been generally characterized by expenditure overruns on current spending and under-execution of capital spending. For ECCU countries, the key ingredients to solid forecasting would include: (i) macroeconomic and fiscal projections that are exclusively based on technicians’ inputs; and (ii) inclusion of the average long-term fiscal and real-sector impact of natural disasters (which are yet to be incorporated in most ECCU countries).
Budget, cash, and debt management. The budget management processes should be synchronized with the FRFs and further upgraded to support policy counter-cyclicality. This would include updating medium term fiscal frameworks in the leadup to each annual budget with planned trajectories for the key fiscal targets for rolling multi-year periods. Additional documents could be produced alongside the budget, including a medium-term debt strategy and a fiscal risk statement. The procedures of budget execution, in the absence of large adverse shocks, should favor stability of budget appropriations. This would allow the operation of automatic stabilizers on the revenue side and could be facilitated by improved cash management and pre-agreed procedures for saving revenue overperformance. Accordingly, debt management procedures should also be upgraded with a view to reducing the interest cost and financing risks.
Government transparency. The governments’ fiscal objectives need to be supported by fostering public debate and engaging civil society and the media. Its medium-term fiscal strategy documents would help frame such a public debate and evaluate progress. To this effect, the government could develop a framework for ex-ante and ex-post verification of compliance with fiscal targets, explaining deviations and corrections and setting out a credible plan for dealing with deviations. Another pillar of transparency is a comprehensive and timely publication of all budget and medium-term fiscal documents, typically during the budget process.
Independent monitoring. Institutions such as fiscal councils can assess compliance with targets, evaluate projections and risks, improve public awareness, and promote transparency and accountability. The challenges of establishing such institutions in small countries with limited resources may imply that creating a fiscal council could take some time. However, the example of Grenada’s Fiscal Responsibility Oversight Committee (FROC) suggests that a fiscal council can become operational relatively quickly. The councils could initially concentrate on a narrower range of tasks, such as ex-post compliance, while gradually building capacity in other areas. Operationalizing the councils involves establishing a trusted, but also arm’s-length, relationship with the government. This would enable the flow of essential information in both directions, while creating space for the councils’ public outreach and accountability.⁴
Sanctions and accountability. It is not uncommon for the FRLs to contain sanctions for non-compliance. These may take the form of (i) personal sanctions of government officials (involving dismissals or fines); and (ii) large financial penalties for institutions or even countries (e.g., in the supranational frameworks, as is the case for the EU). Experience suggests, however, that “punitive” sanctions have limited effectiveness (see IMF 2018a). A key reason is that self-imposed sanctions are unlikely to be implemented at the national level. A more promising avenue to incentivize compliance would be through reputational costs and increasing awareness of the benefits of compliance, including access to financing and other benefits for compliers.
Public investment management. Public investment is a key bottleneck for the ECCU, reflecting the region’s large investment needs and limited capacity to implement complex, multi-year projects. Assessments of the region generally highlight low quality of infrastructure and the need for it to be more resilient to natural disasters (CDB, 2014). The PIMA-based analysis of ECCU countries documents weaknesses in project management, which needs to be significantly upgraded. Findings from a survey of officials using the PIMA methodology show an overall score of 3.9 for the ECCU average, compared to a top score of 10 and 5.4 for Jamaica (see IMF (2018d)). This suggests that there is considerable potential for improving public investment management.

Prepared by Bogdan Lissovolik (WHD).

The two other, smaller, economies – Anguilla and Montserrat – are very particular cases, being British Overseas Territories, with Montserrat having very low public debt due to the lack of authorization to borrow. As such, these two economies are largely excluded from the analysis of this paper.

The total estimate of debt relief is difficult to measure precisely in the debt stocks given the insufficient data for some of the prior years and given that some of the debt restructurings (e.g., Grenada in 2006) provided debt relief that was not fully and quickly reflected in the debt stocks. The estimate of debt relief provided so far this century is in any case likely to exceed 10 percent of ECCU’s combined GDP.

Greenidge et al. (2012) estimate that gross debt beyond the threshold of 55–56 percent of GDP is associated with lower economic growth.

“Underlying” deficits are defined as those of net of CBI inflows. As these inflows are expected to moderate from the high current levels, headline fiscal balances would not be a reliable gauge of the fiscal stance going forward.

Most of the estimates are significantly different from zero, except for St. Lucia and Grenada.

The exact composition effect of the CBI inflows on the other items in the fiscal position differed by country and in some cases affected regular revenues rather than spending (see below).

St. Vincent and the Grenadines (the only ECCU-6 country without a CBI program) is in the process of building a Savings fund.

One such potential determinant would be “political procyclicality,” whereby fiscal deficit positions are affected by the electoral cycle -- a feature common to many countries in the world, including the Caribbean.

Fiscal responsibility frameworks codified in a law and underpinned by multi-year numerical targets are often referred to as “fiscal rules.” This paper eschews the term “fiscal rules” in favor of “fiscal responsibility frameworks” (FRF) to underscore the increased flexibility afforded by the “second-generation” features of these fiscal frameworks (see IMF 2018a).

Key elements associated with well-designed frameworks are listed in (IMF 2018a) and are discussed in the following section.

The adjustment was accompanied by the collaborative debt restructuring, which reduced the debt stock by around 8 percent of GDP, which was less than a quarter of the cumulative debt reduction of 37 percentage points between 2013 and 2017.

The key difference is the absence of a cap on general expenditure growth in Jamaica (although the country has the same wage bill to GDP cap sub-target as Grenada). Additionally, the deficit target is that on the overall balance and not primary balance, while the medium debt target in Jamaica is to be achieved by a specific date (2025), being “undated” in Grenada.

Both countries introduced their FRFs during concurrent IMF programs, which arguably helped strengthen the pre-conditions and policy implementation. Nevertheless, Grenada has continued to implement its FRF after the program expired, while Bahamas enacted its FRF in 2018 without an IMF program.

In this respect, it could be recognized that, despite encouraging experiences so far in Grenada and Jamaica, their new frameworks have not yet been tested by significant natural disasters. While Grenada’s and Jamaica’s FRFs include provisions to deal with natural disasters, these elements could be further enhanced through a more systematic approach.

As per IMF (2018a), well-designed frameworks typically involve (i) calibrating targets based on economic principles; (ii) broad institutional and economic coverage of targets; (iii) incentivizing saving in good times; (iv) precise escape clauses in case of extreme events, and (v) ensuring institutional support that enhances accountability and transparency.

As of 2018, some of the ECCU countries had already tabled medium-term fiscal frameworks in parliaments, while in others the process was ongoing. Most countries were weighing the merits of adopting formal fiscal responsibility legislation to solidify progress. For example, one country authority considered that, in the absence of extensive punitive sanctions for non-compliance in most existing FRLs, these were not suitable for codification in legal norms but rather could be adopted as government or parliamentary declarations of fiscal responsibility.

As exemplified by the case of St. Vincent and the Grenadines, which restructured its public debt after it reached 64 percent of GDP in 2006.

In ECCU, in addition to the natural disasters, fiscal shocks can stem from external demand, off-budget operations, and “tail risks” such as bank bail-outs.

To be sure, discretionary revenue measures could also be implemented under an expenditure-based targets, but in practice spending-based measures are likely to dominate over time in this framework.

Additionally, Grenada’s stopped overstating public investment due to improvements in fiscal statistics starting from 2016, as some of the current spending transactions can no longer be misclassified as capital spending.

For example, Serbia’s fiscal responsibility law of 2011 envisions exempting public investment that exceeds 3 percent of GDP by up to 2 percentage points.

See IMF (2018e) for broader fiscal policy steps that are recommended to deal with natural disasters.

For example, unlike Jamaica sequencing with implementing PFM-type reforms with respect to the FRL substantially differed from Grenada’s, including for example, the introduction of a fiscal council.

While the fiscal multipliers and hence the growth effects of adjustment are estimated to be very small in the ECCU, such a large adjustment could be challenging to implement for political and social reasons.

This conclusion hinges on an assumption that resilient investment is 25 percent more costly than normal investment.

The scenario assumes implementation of fiscal consolidation strategies broadly in line with advice of recent IMF Article IV consultations for each ECCU country, which can be approximated by a budget-balance-based operational target in a fiscal framework. Staff analysis suggests that these strategies are compatible with a prudent target on a fiscal balance net of CBI inflows.

It is assumed, for illustration, that the bulk of the extra cost is covered by concessional financing with a significant grant element, with some delay in the provision of such financing.

Specifically, the more ND-resilient economy in this scenario would enjoy the advantages of (i) higher long-term growth; (ii) better revenues; (iii) lower interest cost due to the need to issue less debt; and (iv) reduced cost of natural disasters as it is offset by insurance payouts.

For insurance alone, the cost of maintaining the same coverage after 2030 would be only one-fourth of the annual cost of scaling up the insurance until then.

The confidence effects from the FRFs would remain in effect even if headline fiscal indicators are thrown off course in the event of a large natural disaster.

This Appendix draws heavily on IMF, 2017b with respect to the examples of the non-ECCU countries, whose references are cited therein.

For example, in this exercise, it was assumed, for illustrative purposes, that past debt restructurings in the ECCU would have occurred even at lower levels of debt.

The evolution of public debt could be an additional characteristic. However, it is not very informative because of the frequent debt restructurings in the ECCU. In this context, most of the debt drivers are already captured by the discussion of the fiscal deficits relative to debt-stabilizing levels.

St. Lucia’s CBI program has been relatively small and started relatively recently, so its impact on the underlying fiscal deficit position is not considered in this analysis as it is not yet macroeconomically relevant.

For example, in Europe countries that included fiscal responsibility in the Constitution include Germany, Poland, Slovakia, Slovenia, and Spain.

Acts with lesser legal power such as political or coalition agreements are relatively rare (involving about 15 percent of FRFs worldwide). One risk is that they may be considered as more short-term-oriented political or electoral moves and thus not effective in limiting the politicians’ bias toward excessive deficits.

In 2016, Grenada improved its accounting of public investment by removing from it recurrent spending.

In Grenada’s case, the FROC is unconnected to the government and reports directly to parliament.

Other Publishers

Save
Cite
Email this content

Share Link

Copy this link, or click below to email it to a friend
Email this content
or copy the link directly:

https://www.elibrary.imf.org/view/journals/002/2019/063/article-A001-en.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully

Eastern Caribbean Currency Union: Selected Issues Paper

Author: International Monetary Fund. Western Hemisphere Dept.

Volume/Issue:: Volume 2019: Issue 063

Publisher:: International Monetary Fund

ISBN:: 9781498300100

ISSN:: 1934-7685

Pages:: 67

DOI:: https://doi.org/10.5089/9781498300100.002

Keywords
Building Ex-Ante Resilience to Natural Disasters1
References
Appendix I. Model Assumptions
Appendix II. Parameter Calibration
Appendix III. Insurance Simulation Methodology
- Insurance Calibration
References
Appendix Table: Analysis of ECCU-6 Fiscal Policies over the Cycle
Appendix I. Tools Used to Evaluate the Fiscal Framework Design
Appendix II. Analysis of ECCU-6 Fiscal Performance
Appendix III. Institutional Factors to Underpin the ECCU FRFs

View raw image

Damage and frequency of natural disasters in the ECCU

View raw image

Migration Rates in the ECCU

(Share of labor force working abroad, in percent)

View raw image

Estimated Disaster National Loss^1/

(In percent of GDP)

View raw image

Natural Disasters Average Annual Loss^1/

(In percent of GDP)

View raw image

Public Capital Stock in Physical Terms

(In percent of GDP)

View raw image

Potential GDP with increase in resilience to 80 percent

(In percent change relative to no resilience)

View raw image

Long-term GDP Return of Investment in Resilience

(Percent change with increase in resilience to 80 percent)

View raw image

Topics

Countries

Series

About

Resources

Abstract

Building Ex-Ante Resilience to Natural Disasters1

A. Introduction

B. Resilient Investment: Model Simulations

Key Model Assumptions

Calibration to ECCU Countries

Results

C. Quantifying Insurance Needs: A Layering Framework

Insurance Simulation Methodology – Summary Presentation

Results: Quantitative Assessment of Insurance Needs

Insurance with Low Resilience

Insurance with High Resilience

D. Conclusions: Putting it all Together

References

Appendix I. Model Assumptions

The economy produces a single tradable good with the following aggregate production function

Private Investor Household

Private Sector Worker Household

Government Worker Household

Government

Appendix II. Parameter Calibration

Appendix III. Insurance Simulation Methodology

Insurance Calibration

References

Appendix Table: Analysis of ECCU-6 Fiscal Policies over the Cycle

Appendix I. Tools Used to Evaluate the Fiscal Framework Design

Appendix II. Analysis of ECCU-6 Fiscal Performance

Appendix III. Institutional Factors to Underpin the ECCU FRFs

Same Series

Other IMF Content

Other Publishers

Asian Development Bank

Food and Agriculture Organization

Inter-American Development Bank

International Labour Organization

Nordic Council of Ministers

The World Bank

Share Link

Table of Contents

Headings

Figures

Building Ex-Ante Resilience to Natural Disasters¹