Information about Western Hemisphere Hemisferio Occidental

8 Government Responses to Natural Disasters in the Caribbean

David Robinson, Paul Cashin, and Ratna Sahay
Published Date:
September 2006
  • ShareShare
Information about Western Hemisphere Hemisferio Occidental
Show Summary Details
Paul Cashin and Pawel Dyczewski 

Economic vulnerability in the Caribbean region is extremely high. A major driver of this vulnerability is the region’s exposure to external shocks generated by natural disasters—over the last three decades, all of the six Eastern Caribbean Currency Union (ECCU) countries that are members of the IMF were ranked among the top 10 in the world in terms of natural disaster events per square mile (see Chapter 7).

Despite this susceptibility to natural disasters, Caribbean risk mitigation activities have been limited and the region’s risk-transfer markets are generally weak.1,2 As a result, natural disasters have a large negative impact on economic activity and poverty in the Caribbean. This chapter briefly sets out the major channels through which small, disaster-prone countries can respond to the challenges posed by natural disasters, with an emphasis on options for ameliorating disaster risk in the Caribbean.

Disaster Risk Mitigation

In addition to obtaining needed post-reconstruction funds, at the government level there are three main responses to the economic vulnerability induced by natural disasters.

Risk identification and reduction focus on mitigating the effects of a disaster should one occur. Proper risk identification involves hazard data collection and mapping, and vulnerability and risk assessments. Similarly, risk reduction activities can tackle revealed vulnerabilities through broad programs of disaster mitigation and preparedness, such as strengthening and relocating structures, retrofitting existing structures, and implementing and enforcing land use codes and building standards.3 Such activities are important, as they can sharply (and permanently) reduce disaster risk exposure and help lower the cost of risk transfer mechanisms by reducing the underlying structural risk of physical assets. Organizations such as the U.S. Agency for International Development (USAID), the European Union, the Caribbean Development Bank, and the World Bank have funded a series of disaster mitigation initiatives in the region in recent decades.

The second government response is self-insurance, which involves establishing economy-wide insurance through the intertemporal transfer of national resources. A typical example might involve the accretion of a precautionary saving fund (based on actuarial probabilities) to draw down upon in the event of a disaster.4 For many developing countries, the first response to a natural disaster involves the diversion of revenue sources from development expenditure to disaster relief and reconstruction, as has been the experience in Grenada in 2004 following Hurricane Ivan. Other forms of self-insurance involve domestic borrowing and tapping external saving (through borrowing and remittance flows).5

The third government response to the economic vulnerability induced by natural disasters is risk transfer, which involves the transfer of resources across states of nature. Risk transfer mechanisms can, in principle, provide a valuable channel for providing capital for rapid rehabilitation and reconstruction of public and private assets. There are several types of risk transfer mechanisms: (1) external assistance, through sovereign debt relief and official development assistance; (2) market insurance and reinsurance, which can provide replacement coverage for public and private assets beyond the capacity of self-insurance; (3) insurance risk pooling, whereby geographical or cross-industry pooling lowers the high cost of disaster risk insurance emanating from the correlation of disaster risks; (4) capital-market-based risk transfer instruments, such as catastrophe bonds, catastrophe options, or weather-related derivatives; (5) contingent lines of credit, whereby sums of credit are made available to insurers and banks in the event of a large disaster, on the basis of the payment of an annual commitment fee; and (6) changes in the composition and structure of public borrowing, which could promote risk-sharing between debtors and creditors.

Some developing countries have undertaken precautionary measures such as establishing disaster funds to draw upon in the event of a natural disaster.6 These funds are based on the principle that as self-insurers, governments should garner sufficient funds to cope with disasters. As a useful rule of thumb, such calamity funds should concentrate on absorbing catastrophic risks that cannot be readily transferred—in particular, disaster-related damage suffered by farmers, the underinsured or noninsured, and the poor. While budgets in Caribbean countries do make provision for disasters, it is typically not to provide resources for disaster funds, but rather for current expenditure on emergency relief and disaster response activities.

As a self-insurance strategy, diversification of labor income through emigration and remittances is a useful means to insure against covariant risk arising from natural disasters. The Caribbean has the highest emigration rate in the world—about 12 percent of the labor force of the region migrated to countries in the Organization for Economic Cooperation and Development during 1970–2000(Mishra, 2006). The Caribbean is also the largest recipient of remittances (in proportion to GDP) in the world. Remittances were about 10 percent of regional GDP in 2002, an amount much larger than overseas development assistance and foreign direct investment flows. It has been pointed out in the literature that remittances play a critical role in insurance and in reducing consumption vulnerability arising from shocks.7

Implications and Weaknesses of the Caribbean Natural Disaster Management Approach

The catastrophe insurance industry faces higher risks and has less developed means of risk assessment than other types of insurance, resulting in higher and more variable premiums. Generally, insurance premiums transfer risk across time and space using well defined techniques for risk assessment. Insurance premiums are calculated based on three main factors: the probability distributions of adverse events, the structural vulnerability of the insured assets, and the value of such assets.8 Unfortunately, because catastrophic events are by definition rather rare and very severe, there is a limited actuarial base available for calculating probability distributions and intensities of their future occurrence. As a result, catastrophe insurance must compensate for this uncertainty factor with higher premiums. Hence, the premiums for catastrophe insurance are proportionally higher than the probability of the insured events. When catastrophic events happen with greater than anticipated frequency (such as hurricanes and earthquakes in the mid–1990s), insurance companies (particularly reinsurance companies) respond with increased premiums, especially for vulnerable regions. For instance, Caribbean countries faced 200 to 300 percent higher insurance premiums in 1992 after Hurricane Andrew in Florida and in 1994 after the Northridge earthquake in California.

National and regional disaster contingency funds are too small to meet financing needs following a disaster. High reliance on donor emergency assistance throughout the Caribbean adversely affects the creation of any sizeable contingency funds. The Eastern Caribbean Central Bank has created a fiscal reserve fund for all member countries in economic difficulties (including those caused by disasters), with contributions sourced from each country’s share of central bank profits. However, the fund’s small size is insufficient for major disaster relief. Another contingency fund for the region is the Disaster Mitigation Facility for the Caribbean (DMFC). In 2001, with support from the USAID Office of Foreign Disaster Assistance, the Caribbean Development Bank established the DMFC, marking an important step in promoting and coordinating risk management within the region. Activities of the DMFC include support for strengthened building standards and enforcement mechanisms, and assistance to member countries in developing national risk management policies and plans.

Traditional insurance markets in the Caribbean are characterized by relatively concentrated coverage, high prices, and low risk transfer. According to a World Bank study (Pollner, 2001), the proportion of residential and commercial properties in the Caribbean covered by traditional insurance is significantly higher (at around 2.3 percent of GDP) than in other developing countries. Relative to population size, the Caribbean enjoys one of the highest densities of insurance companies—one company per 14,000 inhabitants versus one per 107,000 inhabitants for the United States. This translates into an overcrowded industry, with an average insurance company writing barely over US$1 million in premiums, less than 1 percent of a comparative U.S. insurance firm. In such a market, most Caribbean insurance companies act as mere agencies of large reinsurance companies (transferring some 70 percent of premiums and risks to reinsurers in Europe and the United States), rather than genuine underwriters. The cost of reinsurance in the Caribbean is high, largely due to the small capitalization of local insurers and the high exposure of the region to disasters. Further complicating this situation is the insurance industry’s almost exclusive focus on medium-sized to large dwellings and private businesses. Most low-income households, small businesses, and public infrastructure in the region remain uninsured. It is estimated that between 25 and 40 percent of the dwellings in the region are uninsured (and largely uninsurable)—the great majority of them belonging to the above-mentioned groups.9

While the regulatory framework for insurers in the Caribbean is generally adequate, the specifics of the region’s insurance market necessitate reforms aiding market consolidation. National governments regulate insurance markets in the Caribbean, and most countries have adopted European insurance regulations, including registration and capital requirements. However, due to fragmentation of the Caribbean insurance market, the expense ratios of insurance companies are high by international standards. The resulting situation prevents adequate buildup of capital, and makes local insurance companies more vulnerable. Furthermore, the tax systems of some Caribbean countries discourage insurers from setting up specific reserve provisions prior to catastrophic events. Another limitation of insurance market fragmentation is the unwillingness of local companies to insure special risk categories, such as power utilities or major hotels and tourist resorts. Regulations increasing minimum capital requirements and tightening solvency ratios could contribute to industry consolidation and wider insurance coverage in the region.

Ex post financing of damage following natural disasters by international financial institutions and bilateral donor assistance has provided the largest pool of funds for rehabilitation and reconstruction in disasteraffected Caribbean countries. The World Bank has supported natural disaster reconstruction projects across the region, and in recent years has expanded its investments in disaster mitigation projects in members of the Organization of Eastern Caribbean States (OECS)10 Similarly, since 1962, the IMF has provided assistance on nonconcessional terms to 26 member countries afflicted by 29 separate natural disasters (Table 8.1) through Emergency Assistance for Natural Disasters (ENDA) initiative. Since January 2005, IMF members eligible for the Poverty and Growth Reduction Facility (PRGF) have been able to access this facility at concessional rates (an interest rate of one-half of 1 percent a year), with the interest subsidies financed by grant contributions from bilateral donors.11 A higher amount of resources can be accessed under other IMF facilities, such as Stand-By Arrangements and the PRGF, but they are slower to disburse, subject to conditionality, and not geared toward the specific financing problems induced by natural disasters.

Table 8.1.International Monetary Fund Emergency Assistance for Natural Disasters, 1962–2005
CountryYearEventIn millions of

U.S. dollars
In percent

of quota
Egypt1962Crop failure24.026.7
Dominican Republic1979Hurricane22.231.8
St. Lucia1980Hurricane2.350.0
St. Vincent & the Grenadines1980Hurricane0.525.0
Yemen, Arab Republic1983Earthquake10.750.3
Solomon Islands1986Cyclone1.525.0
Dominican Republic1998Hurricane55.925.0
St. Kitts & Nevis1998Hurricane2.325.0
Malawi2002Food shortage23.025.0
Sri Lanka2005Tsunami158.425.0
Source: International Monetary Fund.
Source: International Monetary Fund.

While most Caribbean countries have looked to international financial institutions and bilateral donor agencies for assistance to recover from disasters, it is rather uncommon for such flows to fully offset the losses incurred. This shortfall has arisen largely due to stagnation in the level of resources available from international donors.12 Overseas development assistance flows from developed countries to developing countries have fallen in real terms since the early 1980s, with average flows to ECCU countries having declined from 11 percent of ECCU GDP in the 1980s to less than 5 percent over 2000–03. Of this shrinking amount, an increasing proportion has been allocated to postdisaster reconstruction.

External assistance from the international community has typically been made available without any conditions on undertaking disaster mitigation measures, thereby creating moral hazard problems. As a result, development institutions (both multilateral and bilateral) have served, in effect, as reinsurers of last resort. In addition, given that international financial institutions and bilateral donors find it difficult to commit themselves not to provide such assistance, moral hazard arises because there is then little incentive for disaster-affected countries to undertake disaster mitigation investments, or to self-insure against future disasters. A useful option for international financial institutions would be to provide disasterrelated lines of credit, with access contingent on the ex ante undertaking of disaster mitigating activities.

In spite of the existence of risk-transfer mechanisms, very few developing country governments use them to reduce the resource gap (defined as the difference between funds available and needed for postdisaster reconstruction expenditure). The size of the resource gap rises with the probability of adverse events—the gap for 1-in–20-year events (5 percent probability of occurrence) is typically smaller than that for 1-in–100-year events (1 percent probability). Unfortunately, little reconstruction funding is provided by traditional insurance coverage in Caribbean countries. In addition, sovereigns rarely purchase disaster insurance, and typically do not insure public assets, while catastrophe bonds and weather derivatives issued by developing country sovereigns are nonexistent. At present, the major channels for resource flows for postdisaster expenditure typically involve external aid flows, reallocation of budget expenditures, increased domestic credit (chiefly through local commercial banks), redirection of existing loans from international financial institutions, and additional external commercial credit or credit from international financial institutions.

Scope for Improved Disaster Management

Transferring catastrophic risks to the capital market has been effectively used in some countries as means of spreading risk, stabilizing the insurance market, and increasing insurance coverage. Given the capital constraints of reinsurance companies (and the resulting fluctuations in catastrophe insurance rates), as well as domestic risk aversion of local insurers, there is scope for seeking additional risk-bearing capacity in capital markets. Securitization of catastrophe risk into marketable financial securities is one such solution, and several instruments have been actively trading in developed country markets (United States, Europe, and Japan) since the mid–1990s. A prime example of such a security is catastrophe bonds, which pay investors high yields but are subject to default on all or part of principal and interest if a catastrophic event occurs during the life of the bond. The insured invests the principal in a risk-free asset and is allowed to withdraw only when the specified catastrophic event occurs.13 Other examples of such securities include exchange-traded catastrophe options (the purchaser of such an option can demand payment if an insurance claims index exceeds a prespecified level); catastrophe equity puts (an option which allows the insurer to sell equity shares after a disaster); catastrophe swaps (whereby an insurance portfolio with potential payment liability is swapped for a security with cash-flow payment obligations; and weather derivatives (contracts that provide payments on the occurrence of specified weather events).14

Capital market-based instruments are potentially relevant to the Caribbean, but have yet to be used. They could address current market failures—mainly enabling public utilities and the government sector to obtain some form of insurance against catastrophic damage to public infrastructure. However, a major barrier to using these noninsurance hedges is their cost—in particular, the transaction cost of using these instruments (particularly for single transactions) makes catastrophe bonds (for example) significantly more expensive than traditional insurance as a means of transferring risk (Swiss Reinsurance Company, 1999). As such, catastrophe bonds and other securities are likely to find their greatest applicability in relation to large risk-transfer transactions that are beyond the capacity of insurance and reinsurance markets to bear. However, this is not the environment found in the Caribbean, where the demand for insurance has been more than adequately covered by insurers, who then reinsure the bulk of their risk. This state of affairs suggests that traditional insurance will continue to be the dominant option for Caribbean countries seeking to undertake additional ex ante transference of disaster risk.

Catastrophe insurance pooling might address many problems in the Caribbean insurance market. Instead of transferring insurance risk abroad or to capital markets, Caribbean governments would have the option of insurance pooling for the entire region. Insurance pooling can be much more efficient than individual country insurance. The primary reason is that insurance premiums depend not only on the probability of a given event but also on the uncertainty attached to that probability. Uncertainty of the catastrophic event for a single Caribbean country is much higher than the uncertainty for a group of countries. According to Pollner (2001):

  • “Pooling not only institutionalizes the coverage via insurance of catastrophic risks for both the private and public sectors, but also allows more standardization in the rating and pricing of such risks. Pooling also provides more leverage to cover risks with limited capital available. By retaining some part of the risk that is bearable, this also helps stabilize the availability of such insurance funding and its pricing. This is accomplished via more efficient accumulation of catastrophe reserves which can help buffer some of the global market risks related to natural disasters.” (p. 79)

Such pooling of insurance funds has been successfully implemented in several countries, including the United States, Japan, and New Zealand. Box 8.1 examines how pooling has been employed in Turkey.

It has been suggested in recent years that the high cost of insurance in the Caribbean could be reduced through disaster mitigation initiatives and regional pooling of insurance coverage (designed to diversify risk). The recommendation to establish a regional catastrophe insurance pool came out of a Caribbean Community and Common Market (CARICOM) Working Party on Insurance in the late 1990s. However, this attempt to establish a regional insurance pool for OECS countries (under the auspices of the World Bank) failed to take root, as several countries opted out of the discussions (in part due to the lack of grant funds to complete technical work assessing the actuarial viability of the project). In the wake of the disastrous 2004 hurricane season in the Caribbean, a second attempt is now underway, again under World Bank stewardship, to establish a risk pooling mechanism for CARICOM countries. At the heart of this proposal is the notion that risk pooling across different risk zones in the Caribbean has advantages in that it would lower the minimum net capital requirement and allow for more efficient reinsurance arrangements. Extending such pooling to other small island economies in the world would further diversify the risk, thereby lowering premiums.

Box 8.1.Turkish Catastrophe Insurance Pool

The Turkish Catastrophe Insurance Pool (TCIP) is a recent example of risk pooling for developing countries. The TCIP is relevant to the Caribbean region because it combines risk pooling with the introduction of appropriate incentives for loss mitigation. Specifically, the TCIP is an earthquake risk insurance pooling program that is mandatory for owners of urban residential property. The pool provides cover up to approximately US$50,000 for each dwelling, with a premium that varies across the country depending upon seismicity of the area and the type and quality of housing construction. The government exercises oversight to ensure that insurance pools are managed responsibly. Exposure of the insurance pool is managed by the TCIP’s own reserves, with higher layers of exposure covered by the global reinsurance market and the World Bank.

The financial support provided by the World Bank in the form of liquidity readily available to insured homeowners affected by future events, along with the involvement of private insurers and reinsurers, have contributed to the success of the TCIP since its creation in 2000. That success is measured by the program’s high penetration ratios (about 17 percent of households, the highest among similar pooling programs of national catastrophe insurance for homeowners). The TCIP has also produced greater insurance capacity for Turkey and Turkish risk, and has promoted a broader and more efficient (re)insurance market for such risk.

Changing the composition and structure of public borrowing could also assist in the international transfer of the economic risk of a natural disaster. Compared with advanced countries, developing countries find it difficult to issue long-term debt in their own currencies. Sovereign borrowers also lack equity-like instruments to ensure that investors share in the gains and losses of the sovereign’s economic performance. However, the risk-sharing benefits of equity can be mimicked by the issuance of financial instruments with payment terms indexed to real variables that are (1) partly within the control of national authorities, such as national GDP, or (2) exogenously-determined variables such as real commodity prices or the occurrence of a natural disaster. Such real indexation would provide insurance-like benefits by reducing both the likelihood of debt crises and (by acting as an automatic fiscal stabilizer) the need for procyclical fiscal policies. There is a role for international financial institutions to play in encouraging the creation of markets for the issuance of such real-indexed bonds through such activities as coordinating contacts between debt managers and international investors, and boosting the independence of national statistical agencies (Borenzstein and Mauro, 2002).

Issues of Relevance to Developing Countries

Consequences of Postdisaster Assistance

The provision of postdisaster financial aid has been the traditional strategy for dealing with Caribbean natural disasters, with international donors becoming the de facto insurer of last resort. However, Caribbean governments keen to implement precautionary disaster prevention and mitigation measures are required to undertake current expenditures to reduce future risk, and as with any expenditures, these have opportunity costs. At present, postdisaster assistance is highly subsidized, yet by purchasing insurance and disaster-risk transfer mechanisms, poor disaster-affected countries will have to bear many of the costs presently borne by international donors. Unless international donors can credibly commit to not provide postdisaster assistance, there is little incentive for countries to undertake risk-transfer strategies or engage in risk-reduction and mitigation efforts. This is the “Samaritan’s dilemma,” whereby households, farmers, and businesses rationally believe that governments will be under political pressure to recognize uninsured losses; in turn, governments rationally believe that external donors will provide postdisaster assistance. In both cases, there is likely to be rational underinvestment in ex ante disaster mitigation activities.

Demand for Risk-Transfer Mechanisms

The demand for risk-transfer mechanisms will be determined by a country’s willingness to accept the risk of lower future income due to disasters. This willingness is a function of the probabilistic size of the risk, the cost of insurance, and the cost of other risk-transfer mechanisms (including subsidized postdisaster assistance). The degree of risk aversion in Caribbean countries will be a key determinant of the desirability of risk transfer mechanisms.15 Moreover, the greater the postdisaster access to external savings, the lower will be the demand for risk-transfer mechanisms.

What Is Government Risk from Natural Hazards?

How does one measure government risk from natural hazards? In seeking to identify the risk that is being transferred, the responsibility of a developing country government for losses from natural disasters is often poorly defined. Is government risk to be limited to the rehabilitation of public assets (shifting the risk of government-owned assets)? Or is public risk to be extended to losses by household, farmers, and businesses (resolution of market failures in the provision of risk-transfer options to nongovernment),16 or even extended further to work programs and public assistance to the poor? An important barrier to the adoption of risk-transfer mechanisms in developing countries is both the inability to calculate government risk from natural hazards and the tendency of developing country governments to assume private sector risks (World Bank, 2002).

However, for the transfer of catastrophe risks to operate, the risks being hedged against need to be precisely quantified. Typically there is a four-stage process in quantifying the hedging of government risk: first, there must be catastrophe modeling in order to quantify the expected annual loss; second, where the provision of government-sponsored insurance is involved, additional ambiguities involving moral hazard and adverse selection must be taken into account; third, as disasters reduce income and destroy personal assets, it must be considered that the poor may have claims on public resources in times of crisis; and finally, the cost of any risk-transfer mechanism needs to be compared to existing sources of internal and external financing to cover disaster risk. To properly shift risk, the risk of loss itself needs to be defined. In developing countries, the risk of loss to government assets and activities of government may be so ambiguous that risk-shifting is not a viable option for the components of government risk.


The Caribbean ranks as one of the most disaster prone regions in the world. The macroeconomic impact of natural disasters often severely reduces the well-being of these small island economies, with a disproportionate impact on the poorest segments of the population and the mostly-uninsured public infrastructure.

Rather than preparedness for highly likely disaster events, Caribbean governments have emphasized responses after the fact. Given limited coverage for natural disaster risks provided by local insurance markets, and the dearth of incentives for governments and households to undertake risk mitigation investments, Caribbean governments have typically emphasized ex post responses to natural disasters through the receipt of donor-based emergency external assistance and the diversion of expenditures within domestic budgets. However, while the frequency of natural disasters and the value of assets at risk continue to rise, the capacity of donors to fund disaster assistance continues to be constrained. The end result is a growing gap between the need for, and availability of, resources for disaster reconstruction and relief.

Several tiers of nonmutually exclusive disaster risk management approaches are appropriate for Caribbean countries facing the world’s highest risk of natural disaster. Broadly, Caribbean countries should continue to finance postdisaster expenditures with their traditional financing instruments, supplemented by innovations in insurance and risk-transfer instruments. As always, the appropriate mix of financing options will need to place strong weight on least-cost financing alternatives.

The mix of financing options for postdisaster expenditure can usefully be arrayed as a graduated response to increasing layers (or levels) of natural disaster risk:

  • First, undertaking proper vulnerability assessments and fostering actions to mitigate disaster risk and enhance postdisaster response is a key means to reduce immediate catastrophe risk. Indeed, the lack of knowledge by Caribbean governments of the risk of disasters has hindered the ability of policymakers to plan for them. Precautionary actions would include adopting and enforcing strong building codes, enhancing disaster management agencies, and ensuring effective supervision of national insurance companies.
  • Second, lower-level risk layers could be covered by the establishment of ex ante funding approaches, including the creation of taxpayer-funded national disaster contingency funds, emigration and remittance flows, and intertemporal consumption smoothing through the provision of traditional insurance for key public assets. While self-insurance will not provide the full cost of disaster reconstruction, it is important that sufficient funds be available to governments to meet the immediate short-term costs of disaster relief and rehabilitation. Continuation of self-insurance by national governments, through the exercise of their taxing and borrowing powers to provide financing for disaster reconstruction and relief, is also important. However, Caribbean governments also need to consider the sustainability of public debt stocks in any decision to incur additional domestic and external debt to finance postdisaster expenditures.
  • Third, for higher-risk layers, greater recourse could be made to risk-transfer mechanisms such as regional insurance pools for catastrophe insurance of public and private assets. Where insurance markets are underdeveloped (as in the Caribbean), this may involve spreading risk by establishing a regional catastrophe insurance pool, potentially supported by reinsurance and catastrophe bonds, and by requiring mandatory insurance policies and stringent risk-mitigation initiatives.
  • Fourth, for extremely high-risk layers, provision could be made for access to contingent lines of credit
  • Fifth, funding of postdisaster expenditures is important. Such funding includes the continuing provision by international financial institutions and bilateral donors of concessional loans and grants to finance postdisaster mitigation and reconstruction costs, focusing on disaster relief and the rehabilitation of low-income households. Such funds should be made at least partly contingent on undertaking ex ante risk mitigation activities, so as not to encourage excessive moral hazard (Gurenko and Lester, 2004).

While international capital market instruments (such as catastrophe bonds and weather derivatives) are promising risk-transfer mechanisms, without subsidies from international financial institutions or donors they are likely to be beyond the reach of most developing countries. A more practical approach would be to continue to tap local insurance markets until they are saturated. Indeed, proposals by the World Bank to establish the Caribbean Catastrophe Insurance Pool combine both approaches, involving government-supported regional insurance pools and publicly-issued catastrophe bonds.

Importantly, the opportunity cost of greater use by developing countries of risk-transfer mechanisms needs to be considered. Indeed, it may be optimal from a developing country government perspective to engage in further borrowing, and to seek debt forgiveness and donor flows, as the main responses to natural disasters. That is, the public sector in most Caribbean countries does not typically insure its assets against catastrophic events, and this behavior may well be optimal given the existence of donor support and prevailing conditions in regional insurance markets.17

Risk-transfer mechanisms can also play a vital role in promoting risk mitigation. A key objective should be to transform the balance of catastrophe risk management in the Caribbean away from ex post and ad hoc responses and toward ex ante risk mitigation activities. As noted by the IMF (2003), the willingness of donors to fund ex post disaster relief and reconstruction is finite. Financing gaps between limited donor resources and the growing need for postdisaster funding will continue to widen unless disaster-prone Caribbean countries undertake more disaster risk identification and mitigation activities, supplemented by greater recourse to risk-transfer mechanisms.


    Auffret, P.,2003, “High Consumption Volatility: The Impact of Natural Disasters,”World Bank Policy Research Working Paper 2962(Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    Benson, C., and E.Clay,2003, Economic and Financial Impacts of Natural Disasters: An Assessment of Their Effects and Options for Mitigation(London: Overseas Development Institute).

    • Search Google Scholar
    • Export Citation

    Borenzstein, E., and P.Mauro,2002, “Reviving the Case for GDP-Linked Bonds,”IMF Policy Discussion Paper 02/10(Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Cashin, P.,2004, “Key Features of Caribbean Business Cycles,”IMF Working Paper 04/136(Washington: International Monetary Fund).

    Centre for Research on the Epidemiology of Disasters (CRED), 2005, “Emergency Disasters Database (EM-DAT).”Available via the Internet:

    • Search Google Scholar
    • Export Citation

    Froot, K.,1999, “The Market for Catastrophic Risk: A Clinical Examination,”NBER Working Paper No. 7286(Cambridge, MA: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation

    Gurenko, L., and R.Lester,2004, “Rapid Onset Natural Disasters: The Role of Financing in Effective Risk Management,”World Bank Policy Research Working Paper No. 3278(Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund, 2003, “Fund Assistance for Countries Facing Exogenous Shocks.”Available via the Internet:

    • Search Google Scholar
    • Export Citation

    Mishra, P.,2006, “Emigration and the Brain Drain: Evidence from the Caribbean,”IMF Working Paper 06/25(Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Pollner, J.,2001, “Managing Catastrophic Disaster Risks Using Alternative Risk Financing and Pooled Insurance Structures,”World Bank Technical Paper No. 495(Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    Rasmussen, T.,2004, “Macroeconomic Implications of Natural Disasters in the Caribbean,”IMF Working Paper 04/224(Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Swiss Reinsurance Company, 1999, “Alternative Risk Transfer for Corporations: A Passing Fashion or Risk Management for the 21st Century?”(Zurich: Swiss Reinsurance Company).

    • Search Google Scholar
    • Export Citation

    World Bank, 2002, “Catastrophes and Development,”Disaster Risk Management Working Papers Series 4(Washington: World Bank).


A disaster is the realization of risk (the potential for significant loss), requiring the presence of a hazard, and the vulnerability of physical and human capital to that hazard. While the Caribbean is the most disaster-prone region in the world, there are significant differences in disaster (typically hurricane) exposure within the region (Rasmussen, 2004). Traditionally, the Leeward Islands (St. Kitts and Nevis, Anguilla, Montserrat, and Antigua and Barbuda) are more exposed to hurricanes than the northern Windward Islands (St. Lucia, Dominica) and Barbados, which in turn are more exposed than the southern Windward Islands (Grenada, St. Vincent and the Grenadines) and Trinida and Tobago.


Cashin (2004) finds that Caribbean output volatility is about twice that of the United States, while Auffret (2003) attributes much of this excessive volatility to exogenous natural disaster shocks.


While economic diversification and the restructuring of economies away from disasterprone activities are traditional means of risk mitigation, they represent a much greater challenge in small island economies with economic agents subject to covariant disaster risk.


However, contingency funds maintained in liquid accounts offer a lower rate of return than that typically earned on alternative investments of such funds, and there may be political difficulties in maintaining annual commitments and protecting accumulated funds (Benson and Clay, 2003).


Reconstruction financed by borrowing does increase national public debt, but typically does not increase a country’s ability to service its debt. In addition, most of the six ECCU countries have public debt stocks that are extremely high. Reliance on debt-financed disaster reconstruction is not an optimal policy for highly indebted disaster-prone countries.


In 1996, Mexico established a Fund for Natural Disasters (FONDEN) that is an annual budgetary allocation designed to meet postdisaster expenditures. The fund is designed to finance the repair of uninsured infrastructure, restore the productivity of affected lowincome farmers, and support disaster relief activities (particularly in rural areas). However, FONDEN was insufficiently capitalized to accomplish its multiple obligations, and was recapitalized with the assistance of the World Bank in 2002. In addition, it is important to bear in mind that public commitment to extend disaster coverage to private assets reduces the incentive for economic agents to purchase risk-transfer instruments.


Evidence from econometric analysis using panel data from 1980–2002 for 13 Caribbean countries indicates that while there is weak support for a contemporaneous insurance motive, the insurance effect does occur with a lag of two years. A 1 percent decrease in real GDP is associated with an increase in remittances of about 3 percent, following a two-year lag. There is also evidence that countries with higher remittances have lower volatility of real private consumption, a result that is consistent with the insurance motive (Mishra, 2006).


Typically, only certain public assets are insured in most Caribbean countries, such as key public buildings, as well as some hospitals and airports. Catastrophe insurance is more common in insuring hotels and private tourism infrastructure. While there is no compulsory insurance coverage in the eastern Caribbean, catastrophe coverage (involving all natural hazards) is typically required to secure a mortgage.


Auffret (2003) confirms that Caribbean catastrophe insurance premiums represented about 1.5 percent of GDP over 1970–99, while average (insured and uninsured) losses were only about 0.5 percent of GDP. Both would be equal under actuarially fair pricing, and confirm that the price of catastrophe insurance in the Caribbean can be characterized as high.


Worldwide, the World Bank funded postdisaster reconstruction projects in the 1980s and 1990s totaling more than US$14 billion. The projects focused on repairs to transportation infrastructure, energy systems, and essential social services.


As of April 30, 2005, four countries—Grenada, Malawi, Maldives, and Sri Lanka—had outstanding purchases under the IMF’s ENDA policy. Both Grenada and Malawi have accessed ENDA at subsidized rates of interest.


A study of Dominica in the year following the 1995 hurricane season (during which it was hit with two hurricanes and a tropical storm) revealed that pledged grants and loans totaled about 40 percent of the total cost of storm damage. Similarly, pledges received following Hurricane Ivan’s devastation of Grenada in 2004 amounted to about 20 percent of storm damage.


In capturing the financial risk of catastrophic events and transferring them to capital markets, catastrophe bonds pay out if a defined event (such as a category 4 hurricane on the Saffir-Simpson scale) occurs. Catastrophe bonds have traditionally been issued by insurance or reinsurance companies to assist in transferring underwriting risk. However, while such bonds are of potential relevance to governments in developing countries, as yet no such country (including those in the Caribbean) has used these bonds to transfer catastrophe risk.


Weather-indexed securities have not been as successful as originally envisaged, even in developed countries. A major stumbling block appears to be that of “basis risk”—indexes such as the Saffir-Simpson scale or quantity of rainfall are often poorly correlated with the extent of individual losses.


Catastrophe insurance is expensive, with premia several times larger than the actuarially determined expected loss, chiefly due to a large risk premium arising from the variance of catastrophic losses (Froot, 1999). Over the past two decades, less than 1 percent of losses from catastrophes were insured in poor countries (Rasmussen, 2004).


Crop insurance is not typically available in the Caribbean, which makes poor farmers especially vulnerable to natural disasters. Banana growers are the exception. Growers’ cooperatives have banded together to provide crop insurance to farmers in Dominica, St. Lucia, Grenada, and St. Vincent and the Grenadines affected by windstorm damage through the Windward Islands Crop Insurance scheme. About one-fifth of losses (including those arising from disasters) are covered. The size of the fund has been hampered by limited reserves, traditional adverse selection and moral hazard constraints, costly monitoring of small farmers, and large covariant risk in insuring crop yields.


From the perspective of Caribbean governments, the opportunity costs of risk-transfer mechanisms include the following: (1) creation of a catastrophic risk insurance program will limit discretion to provide disaster relief subsidies and lessen the ability of countries to access postdisaster external assistance; (2) accumulating funds in national disaster funds will divert scarce national savings from other productive uses; and (3) creation of a regional disaster insurance pool may result in the loss of reinsurance commissions to local insurers that have a relationship with international reinsurers.

    Other Resources Citing This Publication