economic entitlements, but not in the severity of the physical characteristics of natural disaster events such as drought or floods. Other lines of research focus on the importance of natural systems and the effects of system change on disaster risk. Extensive research on the impacts of climate change which were assessed by the Intergovernmental Panel on Climate Change (IPCC, 2001), revealed that system change has direct effects on risk and indirect effects through the vulnerability of human populations. A direct effect is increased risk to flooding in coastal settlements, especially those settlements in low-lying coastal areas, in deltas, and on small islands. An indirect effect is the decline of life support functions of coastal ecosystems (e.g., coral reefs and estuaries that support fisheries, recreation, and wildlife habitat) causing increased vulnerability of coastal populations, especially in developing countries, that depend on these ecosystems, which in turn would decrease their capability to cope with future risk (Adger et al., 2005).

Reliability of Data

A major constraint to conducting risk assessments is the absence of reliable data. Comparative assessments of losses at various spatial scales have been made through the use of a wide variety of sources from government compilations, scientific publications, and census information (LA RED, 2002; IPCC, 2001; UNDP, 2004).2 However, systematic record keeping on losses and associated vulnerability indicators is sketchy at best in developed countries and almost nonexistent in developing countries.

Loss of life is the most quantifiable measure, and the most consistently recorded type of disaster loss throughout the world, and frequently constitutes the only loss data available after disasters. Because mortality is considered more reliable than other types of data, it is often viewed as the best indicator for comparative assessments, especially between disasters in developed and developing countries (UNDP, 2004). However, use of deaths as a proxy for disaster risk limits its analysis in relationship to societal development. As noted in Chapter 1, disasters affect people’s lives and livelihoods in many ways other than loss of life. Mortality data do not capture a broader range of other development losses linked to disaster risk trends, and can only point to comparative orders of magnitude in vulnerability and loss. Thus, social, economic, and environmental (built and natural) losses linked to disaster risk should complement analyses based on life losses.


The University of Leuven, Belgium, maintains a central repository of disaster loss data (see, accessed March 24, 2005).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement