Background

Rising natural catastrophic losses are emerging as an economic issue. Comparing the 1990s to the 1960s, the number of catastrophes has increased five-fold, and the damages have increased by a factor of nine (Munich Re 1999). Some of this increase is due to changes in population distribution (more people living in flood-prone regions, for example) and the rising value of infrastructure in harm's way. Evidence is emerging, however, which links windstorms and floods (each accounting for about one-third of total catastrophic damages) to changes in global climate. Warmer surface and ocean temperature results in increased moisture absorption in the atmosphere, and increased moisture absorption leads, in turn, to increased precipitation in the form of floods and windstorm events. Therefore, the worsening catastrophic loss trends of recent decades may be a harbinger of more serious problems to come.

Costs of windstorm are borne primarily by countries in the developed world, costs of flooding by countries in the developing world, and costs of earthquakes are evenly divided between the two regions. In the end, each region bears approximately US$ 35 billion per year in direct costs of natural catastrophes. However, inter-year variation in both the total level of losses and their regional distribution can be large. Of total worldwide economic losses of US$ 65 billion in 1998, close to half were related to a single event in the developing world -- flooding of the Yangtze River. When Hurricane Mitch (the second largest catastrophic event of the year) is figured in, two-thirds of global losses were in the developing world (Swiss Re, 1999). While the interregional distribution is variable, one thing is not: based on income disparities, the per capita burden of catastrophic losses is dramatically higher in developing countries.

Natural catastrophic losses significantly impair development programs by absorbing domestic savings and eroding international development assistance. For example, the Asian Development Bank has estimated that between 1988-98, 5.6% of ADF loans were for disaster rehabilitation. In 1992, nearly 20% of the ADF loans were for rehabilitation assistance related to natural disasters (Arriens and Benson 1999). The World Bank has estimated that in Mexico up to 35% of its lending earmarked for irrigation infrastructure during the past decade has been diverted to pay for the costs of Mexican natural catastrophes. Access to public infrastructure comprises a large component of the wealth of the poorest households and rural infrastructure projects (roads, irrigation, and electrification) have been found to be highly effective in reducing poverty (World Bank 1994). Thus, programs and policies to minimize natural catastrophic losses, to improve public and private responses, and to institute appropriate risk transfer mechanisms will have a high anti-poverty impact.

In July 1999, IIASA instituted a new research initiative in collaboration with the World Bank to explore options for coping with the costs of natural catastrophes in the developing world. The new initiative focuses on two broad issues: estimating the macroeconomic costs of natural catastrophes and understanding institutional structures to implement change in policy.

Estimating macroeconomic costs of natural catastrophes

For most countries in the world, there exist time series data on direct damages caused by natural catastrophes. Typically these data refer to the value of capital (mostly infrastructure and residential structures) damaged. However, surprisingly little work has been done on translating these data into estimates of total direct and indirect economic costs. An initial focus of the research effort focuses on the following proposition: "Given a planned GDP growth path, what additional investment resources must be mobilized in view of the fact that the capital stock is periodically reduced by natural catastrophic losses?" The country that is considered in this exploratory analysis is Argentina.

The analysis is based on RMSM ("Revised Minimum Standard Model"), a World Bank
economic/financial programming model which emphasizes sources-uses accounting consistency. In RMSM, GDP growth is an exogenous assumption. A Harrod-Domar production function is implicitly inverted to solve for required investment expenditure; i.e., assumed GDP growth is combined with an assumed incremental capital-output ratio (ICOR) to solve for required investment. The resources to finance this investment are then calculated in accordance with a set of assumptions on model closure and saving rates.

Natural catastrophes, by destroying capital stock, increase the investment expenditure necessary to achieve a given increase in output because some of the investment is going simply to get the capital stock back to where it was prior to catastrophic damages. The key to our simulation approach was thus adjusting the ICOR upward in each year based on the catastrophic damages drawn from a described loss distribution..

How does this lend itself to modeling macroeconomic impacts of natural catastrophes? The key lies in the concept of "replacement investment." The impact of a downward shock to the capital stock is to raise investment expenditure required to attain targeted GDP growth because some of the expenditure is absorbed by replacing destroyed capital. With this approach, ICOR is a measure of the efficiency of investment expenditure and natural catastrophes reduce that efficiency.

The modeling tool will eventually provide the base for measuring the relative benefits and costs of prevention, mitigation, and financing of natural catastrophe losses.

Institutional Structures to Implement Changed Policy

The development of macroeconomic measures to understand the costs of catastrophes is a tool to evaluate options for dealing with the costs of catastrophes. In addition to understanding the relative financial merits of different policy options, IIASA's research effort also focuses on the institutional framework required in developing countries to support risk-financing options. Without a tradition of private insurance as a vehicle for risk financing, what are the platforms that may be used in these countries to provide risk- financing options? The application of risk financing to catastrophe losses will entail exploring new methodologies for making risk financing viable alternatives for these countries. The role of the government, private sector, and international lending institutions will be need to be explored if viable options for financing risk are to be developed.
Directions for further work

The IIASA and World Bank initiative is a new venture to explore the role that modeling and policy may play in assisting developing countries to deal with the increasing costs of natural disasters. The research work at IIASA should provide the foundation for better understanding the impact of catastrophes on the developing countries and policy options to deal with those impacts.