losses in absolute terms over time or in different locations, or simply attempting to monitor loss history for a single location. Existing global loss databases are useful for certain kinds of analyses, but require improvement in measurements, accuracy, and consistency. For example, there is an ongoing debate in the literature over whether losses from natural disasters are actually increasing over time (Figure 3.1), or whether the data reflect large, recent singular extreme events (e.g., the Tohoku earthquake and tsunami), changes in asset values, changes in reporting, changes in housing stock, improved awareness, or some combination of these. When national losses are normalized for population and wealth, upward patterns in normalized losses appear to become less significant (Pielke and Landsea, 1998; Brooks and Doswell, 2001; Miller et al., 2008); however, other evidence suggests that even with normalization for population and wealth, losses are increasing significantly, especially in the United States (Gall et al., 2012) (Figure 3.2). Improvements in disaster-data collection will help clarify these fundamental tendencies.
FIGURE 3.1 Natural hazard losses worldwide 1980-2011. Source: Munich RE (2012).