The Economics of Natural and Unnatural Disasters

3 - Expectations and Unexpected Consequences of Public Policy toward Natural and Man-Made Disasters

39 3 Expectations and Unexpected Consequences of Public Policy toward Natural and Man-Made Disasters Anthony M. Yezer George Washington University For the purposes of this chapter, I define disasters very generally as large, sudden, infrequent occurrences that are difficult to forecast and that result in significant economic loss in the form of output, income, property, and life. Particular attention is given to disasters that are geographically concentrated as opposed to events like global depressions. This definition is broad enough to cover such disparate events as a regional recession, earthquake, hurricane, drought, oil spill, or terrorist attack. A general approach to disasters has three advantages. First, the principle of parsimony holds that it is desirable to explain as many phenomena as possible with a single theory. Second, generality allows results developed for one type of disaster event to inform our thinking about the economic effects of other disaster types.1 Third, models that claim to explain the effects of many different types of disasters are much easier to refute than those with few testable implications or with narrow predictive power. Theories that are easily refuted should inspire the strongest beliefs in other theories where there is an absence of successful refutation. Put another way, if someone advances a theory of the effects of Hurricane Katrina and claims that it is uniquely appropriate for the U.S. Gulf Coast, the theory is not likely to be generally useful and, because it is based on a single data point, its ability to account for the effects does not indicate that the findings on which it is based are statistically significant. 40 Yezer Literature on the economic effects of disasters concentrates on measures of direct and indirect effects. Direct effects are losses associated with observable damage to property, production, and persons. Indirect effects are costs of recovery and mitigation efforts. Indirect costs are more difficult to observe but can be and have been well measured. This chapter concerns itself with effects that arise through changes in expectations . Direct observation of expectations is generally either not possible, too expensive, or not precise. Accordingly, expectations models must generate a number of implications that can provide indirect validation of the underlying theory. Three disaster expectations models are examined in this paper. First, and most direct, is the effect of disaster expectations on local property values and economic development. Particular attention is given to the possibility that recent disaster experience changes local disaster expectations . This model implies that economic effects of disaster events are based on the unanticipated component of disaster events, or on the difference between actual and expected disaster losses. Second, the effect of disaster expectations on incentives to develop land is considered. Using models taken from urban economics, it is possible to demonstrate circumstances under which private returns from development of land in hazard-prone areas are less than social returns. Third, disaster expectations of property owners should include not only direct damage to their own assets but also the possibility of asset revaluation due to the external effects of disasters on surrounding property. The findings demonstrate that expectations regarding these external effects make disaster insurance different from other forms of hazard insurance and explain some puzzles about behavior of property owners in disasterprone areas. The next four sections of the chapter discuss these three models of disaster expectations (the direct effect of disaster expectations is analyzed in two sections). The final section summarizes the major findings and develops implications of these models for understanding the likely effects of changes in public policy toward natural and man-made disasters. Public Policy toward Natural and Man-Made Disasters 41 INDIVIDUALAND MARKET RESPONSES TO DISASTER EXPECTATIONS There is ample evidence that disaster expectations are priced into markets. The most obvious example is insurance against disaster events, where pricing is based on sophisticated models of the likelihood that events will occur and the estimate of damage, conditional on the event happening. Those insurance companies that do not price insurance and accumulate reserves using statistical models of disaster expectations do not remain solvent for long and can generally be dismissed as curiosities that have no long-term importance.2 There is a strong argument that competitive pressures force most firms to form and act upon efficient disaster expectations, because they must purchase hazard insurance in order to secure capital investment. However, the case for household responses to disaster expectations is not so obvious. Indeed, there is...